any consideration to PhysiX factor?

[kylebisme]

Originally posted by: keysplayr2003
I know. But it's because I really believe it. :beer:

You can't be expected to belive anything but as long as you continue to ignore any possible reason to the contrary. I currently use Nvidia cards in favor of ATIs offerings too. However, I have not out of willfully blind alginate to the brand leading me to belive my choice is best for everyone, rather only because I've found them better suited to my own uses.

Now, anyway, my gtx260 showed up along with the ci7 setup to go with it, so I've got some building to do.

 

[chizow]

Its amazing how certain people continue perpetuating the same FUD and disinformation with regards to PhysX, condemning it because its exclusive and proprietary, then attempt to pimp a less capable but also proprietary physics engine like Havok.

BFG10K, I advised you to do more research into PhysX before commenting but it still seems like you're content to comment ignorantly. I don't think you're going to come out smelling like roses on this one either.

But to help dispel some of the garbage being tossed around with regards to PhysX's proprietary status and industry standard compatibility, Derek did a write up about OpenCL today:

From AT - Open, Closed, Proprietary ... Sorting out the Confusion

AMD could make an investment in the CUDA for C language and create either their own compiler (nothing is stopping them). But then you still have the same problem of interoperability as if NVIDIA implemented Brook+. If NVIDIA or AMD want to make their solution work with the other guy, they would need to write a wrapper to translate CAL to PTX or PTX to CAL. Or we could go a different direction and work on building an industry standard virtual ISA for data parallel architectures. But I doubt that effort would ever take off.

So the bottom line is that both AMD and NVIDIA support both proprietary (Brook+ and CUDA for C) and open standard (OpenCL) solutions. There are further differences between Brook+ and CUDA, but the important part is that these proprietary solutions are not ever going to be able to produce one binary that runs on both AMD and NVIDIA hardware both because of the approach used and the fact that AMD and NVIDIA aren't going to work closely enough to make something like that work. At least in the foreseeable future.

OpenCL, on the other hand, offers developers the ability to write an application once, compile it once, and expect it to run on all major GPU hardware. Something that could never happen with ether CUDA or Brook+.

This helps clarify and solidify much of what was already written with regards to the PhysX SDK, API, and driver along with various blurbs about Nvidia stating they planned full compliance with DX11 and OpenCL.

So what's the end-game here? You're going to have "Open" standards co-existing with "Closed" and "Proprietary" standards but all will flow through the Open standards and therefore, be compatible with one another. Even if DX11 or OpenCL provide a standard API, there will still be a need for the SDKs and front-end tools similar to 3D/game engines like UE3, Gamebryo etc. Making any assumptions about PhysX's future based on its proprietary nature is clearly ignorant and shows just how disingenuous AMD is regarding hardware accelerated physics.

 

[BFG10K]

Originally posted by: chizow

BFG10K, I advised you to do more research into PhysX before commenting but it still seems like you're content to comment ignorantly.

This coming from the guy that has never used a 4xxx part yet runs around telling me that I can?t make comparisons. But he can, because he can quote Derek from ?TEH INTARNETTS?.

Uh-uh, your rampant double standards aren?t going fly here. Until you retract your past claims about commenting on something you?ve never used, you?re going to be called out on this every single time you engage in your hypocrisy.

This helps clarify and solidify much of what was already written with regards to the PhysX SDK, API, and driver along with various blurbs about Nvidia stating they planned full compliance with DX11 and OpenCL.

Clarify what, exactly? All of this has been covered already. Perhaps try reading everything you quoted:

But then you still have the same problem of interoperability as if NVIDIA implemented Brook+.

There are further differences between Brook+ and CUDA, but the important part is that these proprietary solutions are not ever going to be able to produce one binary that runs on both AMD and NVIDIA hardware both because of the approach used and the fact that AMD and NVIDIA aren't going to work closely enough to make something like that work.

So that leaves OpenCL, which we?ve already covered on pages 1-2. Where is this universal implementation? Got a link showing us where multiple games are accelerated on multiple boards by the developer writing one piece of OpenCL code?

Now, if the implementation ever does arrive, nVidia?s whole PhysX advantage will nullified anyway, which again has been pointed out repeatedly in this thread.

 

[taltamir]

nvidia actually wants to port physX to AMD cards, and is providing support to ngohq.com in their effort to do so. AMD isn't too happy about it though.

As for brooks+, that competes with CUDA, not with physX... physX runs on top of CUDA, but is not part of it. AMD simply does not have a physics engine that can run on brooks+.
DX11 and openCL both bring CUDA and Brooks+ like functions to the table... but we are talking about programability, NOT a physics toolkit...
what is the difference? Brooks+, DX11, OpenCL, and CUDA means that a developer may write a physics engine that runs directly on the GPU...

physX is a complete written physics engine that developers simply have to interact with. And AFAIK nvidia isn't charging for it at the moment, so its cheaper to use physX than it is to license a physics engine from unreal, or license havock, or another physics engine (or develop one from scratch, that is the most expensive solution). Havock is a full physics engine that runs on the CPU, physX is a full physics engine that runs on the CPU and on the GPU.

 

[chizow]

Originally posted by: BFG10K
This coming from the guy that has never used a 4xxx part yet runs around telling me that I can?t make comparisons. But he can, because he can quote Derek from ?TEH INTARNETTS?.

I never said you can't make comparisons, I just pointed out your perspective was clearly limited to your current experiences, which at the time were limited for over 3 years to Nvidia parts. You confirmed this some months later when it was proven a legacy title (Thief 2) you cried over had the SAME EXACT problems on ATI DX10 hardware as Nvidia hardware. You still haven't admitted to this very day that its an application or API issue and not a hardware or driver issue specific to Nvidia.

I'm also sure you still feel ATI's monthly driver release program is still somehow indicative of driver quality, when the last few months have completely exposed it for what it is, marketing fluff. You're not getting any comprehensive quality or update, you're getting whatever's ready at 4:30pm PST on Friday before cut off day, heh.

Uh-uh, your rampant double standards aren?t going fly here. Until you retract your past claims about commenting on something you?ve never used, you?re going to be called out on this every single time you engage in your hypocrisy.

Why would I need to retract anything I've said? Do I really need to go out and buy a 4850 or 4870 to tell you they're slower than a GTX 280, loud/hot, studders with Far Cry 2, and CCC is still a piece of garbage? I'll take the readily available links, articles, and testimonials as proof positive, thanks.

As for hypocrisy, LMAO. I'm not the one crying about Nvidia drivers with one hand and then waving a few thousand dollars of Nvidia hardware over the last couple years in the other. And that's also considering your last $1000 of Nvidia hardware prematurely died? Gimme a break lol, its clear you're conflicted in your beliefs or you have a hidden agenda. If I'm a hypocrite you're certainly a walking, breathing paradox.

Clarify what, exactly? All of this has been covered already. Perhaps try reading everything you quoted:

But then you still have the same problem of interoperability as if NVIDIA implemented Brook+.

What problem is there? There's nothing coded in Brook+ that Nvidia would need to port to CUDA, and if there was they could simply write a Brook+ to CUDA wrapper or better yet, use the OpenCL to CUDA wrapper they'll need anyways.

So that leaves OpenCL, which we?ve already covered on pages 1-2. Where is this universal implementation? Got a link showing us where multiple games are accelerated on multiple boards by the developer writing one piece of OpenCL code?

OpenCL is the universal implementation, its up to the internal IHVs to make their parts compatible with it, and since the spec was just finalized I'm sure there aren't any games coded for OpenCL. I guess you'll just have to take Nvidia's word for it that when they say CUDA will be fully portable and compatible with OpenCL, it will be. Considering their VP of Embedded Content is the Chair of the OpenCL Group at Khronos I'd say its a pretty safe bet it will be compatible. :laugh:

Now, if the implementation ever does arrive, nVidia?s whole PhysX advantage will nullified anyway, which again has been pointed out repeatedly in this thread.

No it won't, the PhysX front-end SDK will still be licensed to developers and offer features like hardware acceleration and cross-platform portability that no other vendors offer. Going back to some earlier discussions, just because OpenGL became a widely adopted standard, did that make GLQuake or later versions such as Quake3 engine obsolete? Of course not, they went on to become some of the most licensed and successful game engines in history.

 

[BFG10K]

Originally posted by: chizow

I never said you can't make comparisons, I just pointed out your perspective was clearly limited to your current experiences, which at the time were limited for over 3 years to Nvidia parts.

You mean like how your current perspective is zero given you?ve never used a 4xxx part?

Yet you feel you can comment while telling me my comparison ?doesn?t count? because I might?ve swapped the cards too early? LMAO.

Your antics are comical, akin to that of a circus act.

You still haven't admitted to this very day that its an application or API issue and not a hardware or driver issue specific to Nvidia.

The dithering is a driver issue that affects both ATi and nVidia but not Intel. The other issues however are nVidia driver problems. We know this because the guy who coded the fix said so:

It tries to work around a couple of nvidia driver issues by wrapping thief's D3D calls and making small adjustments to them.

There are two main issues causing most of the problems. The first is a problem with rendering pre-transformed verticies from system memory with a z value of exactly one, which was responsible for the invisible stars. My patch just checks for any verticies with z >= 1 and sets it to 0.9999 before passing them on to D3D. My guess is that this is a driver bug, because with a bit of proding I could not only get the entire sky to vanish completely, but I could get the light gem to disapear behind nearby walls, even though it's rendered with the zbuffer disabled.

The second is the lack of dithering support, which presumably was a deliberate design decision by nvidia. I fixed that by tricking thief into using a 32 bit render target, so the lack of dithering isn't noticable.

What part of this are you having trouble understanding? Why do I need to repeat this every time we have this discussion?

I'm also sure you still feel ATI's monthly driver release program is still somehow indicative of driver quality, when the last few months have completely exposed it for what it is, marketing fluff.

To quote you again: I just pointed out your perspective was clearly limited to your current experiences,.

So given your current experiences with 4xxx cards are zero, how exactly do you feel you can claim something is marketing fluff? Because Derek said so, and you know how to use quote tags?

You're not getting any comprehensive quality or update, you're getting whatever's ready

Given they?ve been better overall than nVidia?s since the program started in 2002, I?d say I?m getting a whole lot. But then I?ve been using both vendors since 2001 so I can actually make a comparison, unlike you.

You just regurgitate what Derek says, but only if it suits your agenda. You don?t seem to regurgitate the comments he made that debunked your comments about Steam and market share.

Why would I need to retract anything I've said?

Because according to you, your perspective amounts to zero. Those are your words.

Let me quote it again in case it has slipped your mind: I just pointed out your perspective was clearly limited to your current experiences,

So again given your current experience with ATi cards is zero, that would also mean your perspective is zero. Again, those are your own words.

As for hypocrisy, LMAO. I'm not the one crying about Nvidia drivers with one hand and then waving a few thousand dollars of Nvidia hardware over the last couple years in the other. And that's also considering your last $1000 of Nvidia hardware prematurely died? Gimme a break lol, its clear you're conflicted in your beliefs or you have a hidden agenda. If I'm a hypocrite you're certainly a walking, breathing paradox.

Yep ? I do have an agenda: to spend the most money on my vendor of choice. But that doesn?t mean I?m going to be a PR poster boy and pretend said vendor is perfect.

I notice you have great difficulty understanding the notion that buying a piece of hardware and pointing out its flaws are not mutually exclusive, but that?s something you need to work through as it?s not my problem.

What problem is there?

Ask Derek ? he wrote it.

OpenCL is the universal implementation, its up to the internal IHVs to make their parts compatible with it, and since the spec was just finalized I'm sure there aren't any games coded for OpenCL. I guess you'll just have to take Nvidia's word for it that when they say CUDA will be fully portable and compatible with OpenCL, it will be.

So what if CUDA is fully portable for it? How does that help non-nVidia vendors get PhysX through OpenCL?

No it won't, the PhysX front-end SDK will still be licensed to developers and offer features like hardware acceleration and cross-platform portability that no other vendors offer.

You mean like it is now? Again show me that list of titles I asked for last time and then we can conclude PhysX is viable. Until you provide the list your comments are nothing more than speculation.

Going back to some earlier discussions, just because OpenGL became a widely adopted standard, did that make GLQuake or later versions such as Quake3 engine obsolete?

Why would they be obsolete given they were developed with the exact same standard?

GLQuake and Quake Wars use the same standard and this is nothing like comparing PhysX and OpenCL.

 

[SlowSpyder]

Originally posted by: taltamir
nvidia actually wants to port physX to AMD cards, and is providing support to ngohq.com in their effort to do so. AMD isn't too happy about it though.

As for brooks+, that competes with CUDA, not with physX... physX runs on top of CUDA, but is not part of it. AMD simply does not have a physics engine that can run on brooks+.
DX11 and openCL both bring CUDA and Brooks+ like functions to the table... but we are talking about programability, NOT a physics toolkit...
what is the difference? Brooks+, DX11, OpenCL, and CUDA means that a developer may write a physics engine that runs directly on the GPU...

physX is a complete written physics engine that developers simply have to interact with. And AFAIK nvidia isn't charging for it at the moment, so its cheaper to use physX than it is to license a physics engine from unreal, or license havock, or another physics engine (or develop one from scratch, that is the most expensive solution). Havock is a full physics engine that runs on the CPU, physX is a full physics engine that runs on the CPU and on the GPU.

For whatever it's worth I think this is the first real good Physx argument/point I've read. Assuming that Nvidia isn't charging then I think that'll really give them a good head start as developers will probably want to use it, if nothing else that'd be something you can put on the box.

But with that being said I still would not feel confident predicting that Physx will be 'the' hardware accelerated physics of the future. Out of MS, Intel, AMD, and Nvidia you have just one of thoes companies on board and two of those companies would really probably not like to use it.

 

[chizow]

Originally posted by: BFG10K
You mean like how your current perspective is zero given you?ve never used a 4xxx part?

Yet you feel you can comment while telling me my comparison ?doesn?t count? because I might?ve swapped the cards too early? LMAO.

Your antics are comical, akin to that of a circus act.

My comments were made over a year before you decided to update your frame of reference, so having never used a 4XXX part is irrelevant. You stated you felt ATI's drivers were superior to Nvidia's based largely on your experience with the G80's launch drivers at a time when you had not used an ATI part in ~3 years. Are you contesting this or are you going to just try and cover your tracks with misdirection as you usually do?

The dithering is a driver issue that affects both ATi and nVidia...

So you were wrong that it was an Nvidia-only problem. When asked repeatedly, you insisted the problem did not exist on ATI parts, yet they did.

So given your current experiences with 4xxx cards are zero, how exactly do you feel you can claim something is marketing fluff? Because Derek said so, and you know how to use quote tags?

When a company has to release 3 hot fixes/betas to specifically fix a problem and goes through 3 driver revisions without actually fixing anything, its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline.

Given they?ve been better overall than nVidia?s since the program started in 2002, I?d say I?m getting a whole lot. But then I?ve been using both vendors since 2001 so I can actually make a comparison, unlike you.

I've used both vendors as well thanks, and I've had great experiences with Nvidia drivers. Nvidia has been so good that ATI has given me no reason to spend money on their parts since the 9700pro.

You just regurgitate what Derek says, but only if it suits your agenda. You don?t seem to regurgitate the comments he made that debunked your comments about Steam and market share.

I acknowledged his comments that he felt ATI's drivers were better at the time, but I'd say that opinion has clearly changed. But I still disagreed for his reasoning for dismissing the Steam figures, as they were cumulative to-date. He locked the thread before replying, however a little later Jensen Huang made some of the EXACT same comments referencing the Steam survey at Analyst Day in May:

• Nvidia Analyst Day
  NVIDIA has about 87% of the GPUs running in DX10 boxes while Intel powers only 0.11% (which Jen-sun said "I think that's just an error," and may well have been right). He was on target when he said that in this case "approximately zero seems statistically significant."

and in the comments:

Nice.... by chizow, 266 days ago
Those Steam figures look familiar Derek. I'm surprised JH didn't bring up the Microsoft class action suit as another example of Intel integrated chipsets failing miserably. Nice peak into the current market climate, although there wasn't as much discussion about the future as I had hoped.

Reply
RE: Nice.... by DerekWilson, 266 days ago
heh yeah ... but the steam numbers still say absolutely nothing about the state of the market in early 2007.

they are a good indicator for what is happening now, and i never meant to imply other wise.

i'd love to see something more forward looking as well...

I still disagree with him about the numbers being irrelevant for early 2007 as they are still cumulative however, and I even drilled-down further isolating the exact number of G80 parts based on RAM amount and found the % of G80 parts to R5XX parts, the only DX10 parts available at the time in question made up over 90% of the total DX10 market.

So ya, I'd say if my analysis is in-line with the conclusion of the smartest guy in the room (JH Huang), it should certainly be good enough for a nobody like you.

Because according to you, your perspective amounts to zero. Those are your words.

Let me quote it again in case it has slipped your mind: I just pointed out your perspective was clearly limited to your current experiences,

So again given your current experience with ATi cards is zero, that would also mean your perspective is zero. Again, those are your own words.

Uh, no, I haven't made any idiotic claims like "Nvidia drivers are far superior to ATI's based on my extensive experience with both" when my last ATI part was a 9700pro. This is in contrast to the claims you made after 3 years of exclusive Nvidia experience rofl.

Yep ? I do have an agenda: to spend the most money on my vendor of choice. But that doesn?t mean I?m going to be a PR poster boy and pretend said vendor is perfect.

I notice you have great difficulty understanding the notion that buying a piece of hardware and pointing out its flaws are not mutually exclusive, but that?s something you need to work through as it?s not my problem.

I don't have a problem pointing out a product's flaws, its just clearly obvious your bias prevents you from doing so equally. For example, in your latest act of purchasing hypocrisy, why buy a GTX 216 instead of a 1GB 4870 given price was similar, ATI drivers are superior in your opinion, and the 4870 generally outperformed the GTX 260 in reviews? You don't overclock, warranty obviously does you no good wherever you live, and you clearly have a preference for ATI parts. If there was ever a time to break your Nvidia addiction, surely this would be it. Yet you bought an Nvidia part. But ATI parts run Red Faction better right? LMAO.

Ask Derek ? he wrote it.

He's assuming there's something written in Brook+ worth porting, when there isn't, so there is no problem.

So what if CUDA is fully portable for it? How does that help non-nVidia vendors get PhysX through OpenCL?

That's something they need to work out, the point is that PhysX and CUDA are fully compatible with the industry standard, something you contested which is clearly FUD.

You mean like it is now? Again show me that list of titles I asked for last time and then we can conclude PhysX is viable. Until you provide the list your comments are nothing more than speculation.

Its already been linked to you numerous times. All those titles have licensed versions of the PhysX SDK, even if they do not offer GPU accelerated support. Contrary to what you stated earlier, there have been legacy games that have had GPU PhysX forced on them, UT3, GRAW/2, and Warmonger and there are 2 new titles releasing in weeks that were developed with PhysX from the ground-up.

And before you drivel on about unreleased games, Cryostasis has already been released in Russia, I'm sure the delay is only due to localization. Mirror's Edge releases later this week and given its already out on PS3 and 360 I think its safe to say its going to release and not turn into PhysX vaporware. For the future we have 3 major studios, EA, 2K, and THQ that recently announced PhysX support to add to the likes of Epic and GameBryo that already use PhysX in their titles. So ya, if you want to continue you're going to have to do more than push your weak brand of fear mongering about proprietary standards.

Why would they be obsolete given they were developed with the exact same standard?

GLQuake and Quake Wars use the same standard and this is nothing like comparing PhysX and OpenCL.

That's the point, it won't become obsolete because OpenCL is now the standard, PhysX will build on that standard going forward just as the various Quake engines built on OpenGL. Did people stop using the Quake engine because it was proprietary and they had to pay for it? Of course not.

 

[BFG10K]

Originally posted by: chizow

My comments were made over a year before you decided to update your frame of reference, so having never used a 4XXX part is irrelevant.

Actually it?s very relevant given you have yet to update your frame of reference. You also continue to make comparisons that you forbad me from making before I updated my frame of reference, despite you still not updating yours.

You stated you felt ATI's drivers were superior to Nvidia's based largely on your experience with the G80's launch drivers at a time when you had not used an ATI part in ~3 years. Are you contesting this or are you going to just try and cover your tracks with misdirection as you usually do?

You stated you felt nVidia?s drivers are superior to ATi?s despite not having used an ATi part in ~7 years. Are you contesting this or are you going to just try and cover your tracks with misdirection as you usually do?

So you were wrong that it was an Nvidia-only problem. When asked repeatedly, you insisted the problem did not exist on ATI parts, yet they did.

Nope, when I found the dithering affected ATi parts (in that same thread) I admitted it was an issue. Not only that but I?ve petitioned to get it back @ Rage3D. Of course you wouldn?t know any of this given you spend your days marketing nVidia. That and your ?understanding? of the issue is as woeful as your comments about it.

When a company has to release 3 hot fixes/betas to specifically fix a problem and goes through 3 driver revisions without actually fixing anything, its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline.

Answer the question: why can you make comments based on what you?ve read on the internet but I can?t?

Answer the question.
Answer the question Chizow.

Answer the question: why can you make comparisons of ATi vs nVidia drivers despite you not having used an ATi part for 7 years, while you lambasted me for making such a comparison because I hadn?t used an ATi part for 3 years?

Answer the question.
Answer the question Chizow.

Answer the question: why can you discuss 4xxx parts despite you never having used them, while lambasting me for not using Vista as my primary gaming OS?

Answer the question.
Answer the question Chizow.

I've used both vendors as well thanks, and I've had great experiences with Nvidia drivers. Nvidia has been so good that ATI has given me no reason to spend money on their parts since the 9700pro.

You haven?t used both vendors in 7 years but you?re still making comparisons. I?ve used both vendors? current parts and you still were telling me that I couldn?t make a comparison because ?I swapped the cards too early? or some such nonsense.

Answer the question: why can you make comparisons but I can?t?

Answer the question.
Answer the question Chizow.

I acknowledged his comments that he felt ATI's drivers were better at the time, but I'd say that opinion has clearly changed.

So do you then acknowledge ATi?s drivers were superior at the time? Or do you simply acknowledge his comments when you think he?s changed his mind and it backs your agenda?

Anyway, we aren?t talking about Derek, we?re talking about you given you haven?t touched an ATi part in 7 years.

To quote you again: I just pointed out your perspective was clearly limited to your current experiences,.

Since your current experience with ATi parts is zero, this means your perspective is zero in terms of a comparison and also in terms of commenting about current ATi drivers and parts.

Answer the question: is the above comment true or is it a lie?

Answer the question.
Answer the question Chizow.

Uh, no, I haven't made any idiotic claims like "Nvidia drivers are far superior to ATI's based on my extensive experience with both" when my last ATI part was a 9700pro.

You?re right, you simply made the claim that nVidia?s drivers are superior and left out the last bit telling us that you hadn?t touched ATi parts in 7 years.

Answer the question Chizow: why can you talk about current ATi drivers, having not used an ATi part since 7 years ago, but I can?t do the same despite having current experience with both vendors? latest parts?

Answer the question.
Answer the question Chizow.

I don't have a problem pointing out a product's flaws, its just clearly obvious your bias prevents you from doing so equally.

A bias towards whom? ATi or nVidia?

For example, in your latest act of purchasing hypocrisy, why buy a GTX 216 instead of a 1GB 4870 given price was similar, ATI drivers are superior in your opinion, and the 4870 generally outperformed the GTX 260 in reviews? You don't overclock, warranty obviously does you no good wherever you live, and you clearly have a preference for ATI parts. If there was ever a time to break your Nvidia addiction, surely this would be it. Yet you bought an Nvidia part.

Why buy a GTX260? The fact that you even have to ask that question demonstrates you know even less about the relative strengths of the vendors than I thought you did. Please conduct some research to educate yourself and then we can continue this conversation.

But ATI parts run Red Faction better right?

They sure do

He's assuming there's something written in Brook+ worth porting, when there isn't, so there is no problem.

Uh, no. He?s making the obvious point that proprietary standards don?t mix, standards like PhysX.

That's something they need to work out, the point is that PhysX and CUDA are fully compatible with the industry standard, something you contested which is clearly FUD.

No they aren?t; CUDA and PhysX are proprietary nVidia standards. Adding OpenCL to the mix doesn?t change anything unless you can demonstrate that you?ll achieve hardware accelerated PhysX on all IHVs by developers running OpenCL code.

Its already been linked to you numerous times.

You?re referring to the list from nZone? You?re not being serious, right? LMAO.

All those titles have licensed versions of the PhysX SDK, even if they do not offer GPU accelerated support.

If they don?t support nVidia GPU acceleration then how is PhysX an advantage for nVidia in those titles?

If they?re console games then how is PhysX an advantage for nVidia?

Contrary to what you stated earlier, there have been legacy games that have had GPU PhysX forced on them, UT3, GRAW/2, and Warmonger and there are 2 new titles releasing in weeks that were developed with PhysX from the ground-up.

Nothing was ?forced? into them. The first three had updates to apply cherry-picked PhysX maps which added nothing to the core gameplay, and the last game was designed with hardware PhysX in mind.

And before you drivel on about unreleased games, Cryostasis has already been released in Russia, I'm sure the delay is only due to localization. Mirror's Edge releases later this week and given its already out on PS3 and 360 I think its safe to say its going to release and not turn into PhysX vaporware.

Again, we?ve already covered those two games.

For the future we have 3 major studios, EA, 2K, and THQ that recently announced PhysX support to add to the likes of Epic and GameBryo that already use PhysX in their titles.

Likewise there are at least three major studios that have announced DX10.1 support.

That's the point, it won't become obsolete because OpenCL is now the standard, PhysX will build on that standard going forward just as the various Quake engines built on OpenGL.

It doesn?t matter ? unless OpenCL provides hardware accelerated PhysX on non-nVidia parts then it?s not a standard. This is unlike OpenGL where its code provides hardware acceleration on any part with an OpenGL ICD, and at least four current consumer IHVs have OpenGL ICDs.

Did people stop using the Quake engine because it was proprietary and they had to pay for it? Of course not.

What on Earth are you babbling about? A game engine is not the same as an API. Making such a comparison is akin to saying Far Cry is the same as DirectX.

The Quake engine runs under OpenGL. So does Quake 3. So does Quake Wars. The engine is irrelevant - all you need is an OpenGL ICD and at least four consumer IHVs currently provide one, which means all of them can run those games.

PhysX is a totally different API to OpenCL. That PhysX can be wrapped to OpenCL is no different to claiming Direct3D and Glide are the same thing because there are Glide wrappers that run under Direct3D.

Furthermore, since PhysX only runs on nVidia cards then it doesn?t really matter if an OpenCL wrapper is made because someone has to port the PhysX back-end to the other IHVs to provide hardware acceleration through that API. That anyone could potentially do it is the same as saying anyone could potentially win lotto as long as they buy a ticket.

 

[nRollo]

DP

 

[nRollo]

Originally posted by: BFG10K
Answer the question: why can you make comparisons of ATi vs nVidia drivers despite you not having used an ATi part for 7 years, while you lambasted me for making such a comparison because I hadn?t used an ATi part for 3 years?
.

I do have experience with ATi parts of all generations, and I have experience with PhysX, CUDA, and 3dVision.

At the end of the day, there's nothing about DX10.1 or in the 8X MSAA advantages in some games that could make me even consider non NVIDIA alternatives.

The original question here is whether PhysX needs to be given any consideration- it clearly does.

As the only hardware accelerated physics package available available that has been proven to make dramatic differences in image quality and game experience, it needs to be considered. Nothing ATi parts do can adds the level of immersion PhysX provides, or anything close.

As part of the above proprietary vendor features package, PhysX shines. Care to make a bet with me there will be more PhysX games out at the end of 2009 than DX10.1 games? How much? (not to mention the IQ differences of DX10.1 have thus far been unremarkable so far)

 

[Pantalaimon]

Originally posted by: nRollo

At the end of the day, there's nothing about DX10.1 or in the 8X MSAA advantages in some games that could make me even consider non NVIDIA alternatives.

Somehow I doubt you would consider any other non NVIDIA no matter the situation. You probably would advise people to consider the FX5800 over the 9700 Pro even.

 

[Keysplayr]

Originally posted by: TheSnowman

Originally posted by: keysplayr2003
I know. But it's because I really believe it. :beer:

You can't be expected to belive anything but as long as you continue to ignore any possible reason to the contrary. I currently use Nvidia cards in favor of ATIs offerings too. However, I have not out of willfully blind alginate to the brand leading me to belive my choice is best for everyone, rather only because I've found them better suited to my own uses.

Now, anyway, my gtx260 showed up along with the ci7 setup to go with it, so I've got some building to do.

I don't believe I have ignored anything. I have read, digested, mulled over, and considered most of the comments made by folks who for some reason are dead set against anything to do with PhysX, and come up with, "WTF?". It's damn strange. This is new and exciting stuff and I don't care who you are or what your preferences are, it can't be ignored.

The second bolded area appears to only be your subjective opinion as to why I have arrived to the conclusion that Nvidia has the better choices right now. You say I'm willfully blind because I'm a focus group member. But you ignore the fact that years and years before I was ever part of this group, I felt the same way and expressed myself the very same exact way I do now. You just use my group membership and say I arrived at my conclusions because of it, when all along, I've arrived at this conclusion years ago. I believe what I believe from my own experiences. Past and present. Just as I suppose you do as well. For example, you've ignored every persons post you've read that stated that the 4870 was a better choice over a GTX260. You've read them, mulled them over, and made your choice. I'm not going to condemn you for that, you bought what you thought was better right now for your own reasons. You may not have considered PhysX at all when you decided on your purchase, and thats perfectly fine if it doesn't interest you. But you now have the ability to use it whenever you wish. If a title comes out that makes you go "Ohhhh, I need to get that!" And it's a PhysX title, you're covered. And I'm sure you've considered DX10.1 in your purchase as is evident in the very first post of yours in this thread which is right under my very first post. But you still went with a GTX260. But I wouldn't worry too much about it, as it is becoming clear that Nvidia DX10 GPU's seem to be handling any DX10.1 featured game so far. Chizow had posted about this and linked to some backup.

So, I won't condemn you for your choices, so long as you don't condemn me for mine.

:beer:

 

[Keysplayr]

Originally posted by: Pantalaimon

Originally posted by: nRollo

At the end of the day, there's nothing about DX10.1 or in the 8X MSAA advantages in some games that could make me even consider non NVIDIA alternatives.

Somehow I doubt you would consider any other non NVIDIA no matter the situation. You probably would advise people to consider the FX5800 over the 9700 Pro even.

:::Goes back in time machine:::

The 9700 was an extraordinary achievement. I owned 3 of these cards myself, and a 9500pro. You cannot deny Kudos when Kudos is due. Nvidia ran into various probs with the 5800's. Both companies have hiccups now and again.

:::Back to the present:::

Things are vastly changed from years ago. GPGPU's weren't even a blip on the map. Probably just an idea as these companies work on 2 to 3 gens ahead of current offerings.

 

[SlowSpyder]

Originally posted by: keysplayr2003

Originally posted by: Pantalaimon

Originally posted by: nRollo

At the end of the day, there's nothing about DX10.1 or in the 8X MSAA advantages in some games that could make me even consider non NVIDIA alternatives.

Somehow I doubt you would consider any other non NVIDIA no matter the situation. You probably would advise people to consider the FX5800 over the 9700 Pro even.

:::Goes back in time machine:::

The 9700 was an extraordinary achievement. I owned 3 of these cards myself, and a 9500pro. You cannot deny Kudos when Kudos is due. Nvidia ran into various probs with the 5800's. Both companies have hiccups now and again.

:::Back to the present:::

Things are vastly changed from years ago. GPGPU's weren't even a blip on the map. Probably just an idea as these companies work on 2 to 3 gens ahead of current offerings.

You weren't a focus member back then either. But I do remember way back in those days, I was a lurker, don't think I even had an account here yet, just read these forums just about everyday. I remember thinking to myself, "Who is this Rollo guy, and why does he keep talking about how good the 5800/5900 cards are compared to the Radeon 9700/9800's?"

 

[Keysplayr]

Originally posted by: SlowSpyder

Originally posted by: keysplayr2003

Originally posted by: Pantalaimon

Originally posted by: nRollo

At the end of the day, there's nothing about DX10.1 or in the 8X MSAA advantages in some games that could make me even consider non NVIDIA alternatives.

Somehow I doubt you would consider any other non NVIDIA no matter the situation. You probably would advise people to consider the FX5800 over the 9700 Pro even.

:::Goes back in time machine:::

The 9700 was an extraordinary achievement. I owned 3 of these cards myself, and a 9500pro. You cannot deny Kudos when Kudos is due. Nvidia ran into various probs with the 5800's. Both companies have hiccups now and again.

:::Back to the present:::

Things are vastly changed from years ago. GPGPU's weren't even a blip on the map. Probably just an idea as these companies work on 2 to 3 gens ahead of current offerings.

You weren't a focus member back then either. But I do remember way back in those days, I was a lurker, don't think I even had an account here yet, just read these forums just about everyday. I remember thinking to myself, "Who is this Rollo guy, and why does he keep talking about how good the 5800/5900 cards are compared to the Radeon 9700/9800's?"

It wasn't all bad. Just imagine how good a 2900XT would have looked if the 8800's didn't come to be, and only a 32/40 pipe conventional NV GPU came out like we all expected.

 

[chizow]

Originally posted by: BFG10K
Actually it?s very relevant given you have yet to update your frame of reference. You also continue to make comparisons that you forbad me from making before I updated my frame of reference, despite you still not updating yours.

How is it relevant when your experiences with a 4850 had absolutely no influence whatsoever on your idiotic claims about drivers a year ago based on 3 consecutive years of experience with Nvidia parts? LMAO.

You stated you felt nVidia?s drivers are superior to ATi?s despite not having used an ATi part in ~7 years. Are you contesting this or are you going to just try and cover your tracks with misdirection as you usually do?

Prove it. I've said no such thing. I've maintained both IHVs had issues, you kept insisting Nvidia problems were worst, particularly with legacy titles, and specifically stated ATI's drivers and driver program were better than Nvidia's based on your experiences as if they were anything close to equal! rofl. Shocker, considering your Nvidia frame of reference dwarfed your ATI experiences by at least 3:1. Year and a half later, 8800GTS, 8800 Ultra, 4850, and now GTX 260. 3:1 again, actually 6:1 going back another 3 years. Let me guess, you've experienced more problems with Nvidia parts this time too right? LMAO.

Nope, when I found the dithering affected ATi parts (in that same thread) I admitted it was an issue.

BS, you didn't admit anything until it was shoved down your throat by Ben and Apoppin on these forums. And even then you didn't admit anything, all this after crying about an 8 year old legacy game that only had problems with Nvidia parts, which again proved my point that your opinion was skewed as you had no relevant experience with ATI parts in a period dominated by Nvidia parts. If you did actually have relevant experience, or even did a bit of research, you would've seen ATI parts had the same problems and you could've saved 3-4 pages of foot stomping over some extraordinarily ugly textures in an ugly 8-year old game.

Answer the question: why can you make comments based on what you?ve read on the internet but I can?t?

Answer the question.
Answer the question Chizow.

Answer the question: why can you make comparisons of ATi vs nVidia drivers despite you not having used an ATi part for 7 years, while you lambasted me for making such a comparison because I hadn?t used an ATi part for 3 years?

Answer the question.
Answer the question Chizow.

Answer the question: why can you discuss 4xxx parts despite you never having used them, while lambasting me for not using Vista as my primary gaming OS?

Answer the question.
Answer the question Chizow.

Rofl, OCD much? You made a claim that ATI's monthly driver updates were superior to Nvidia's because you knew when an update was coming and if there was a problem, you would likely get a fix in a month. So again, if a problem in a AAA title has a hot fix released to specifically fix it, but still lingers over 2 additional hot fixes specifically targetting it and is still a problem after 3 WHQL releases, do you think the monthly driver program is working as you described? Or if the similar occurs for a pervasive problem with CF in Vista that lingers on in various permutations still, after 3 driver releases, do you think the monthly driver program is working?

You haven?t used both vendors in 7 years but you?re still making comparisons. I?ve used both vendors? current parts and you still were telling me that I couldn?t make a comparison because ?I swapped the cards too early? or some such nonsense.

Answer the question: why can you make comparisons but I can?t?

Answer the question.
Answer the question Chizow.

Where am I making comparisons? The thread where I explicitly quoted Derek and Anand? Nope, I was just quoting the experiences of people with actual experience with a wide variety of parts from both vendors who clearly felt ATI's drivers were worst than Nvidia's right now. Given they do this for a living and have simultaneous experience with said parts at any given time, their opinion actually means something, unlike your weighted and biased snapshots in time dominated by Nvidia parts.

So do you then acknowledge ATi?s drivers were superior at the time? Or do you simply acknowledge his comments when you think he?s changed his mind and it backs your agenda?

Yep, and I said that in the comments, which you selectively ignored to quote. It also doesn't change the fact his opinion has changed and he and Anand feel ATI's drivers are much worst than Nvidia's in their current state.

Anyway, we aren?t talking about Derek, we?re talking about you given you haven?t touched an ATi part in 7 years.

To quote you again: I just pointed out your perspective was clearly limited to your current experiences,.

Since your current experience with ATi parts is zero, this means your perspective is zero in terms of a comparison and also in terms of commenting about current ATi drivers and parts.

Answer the question: is the above comment true or is it a lie?

Where am I holding out my opinion based on my current experience with ATI parts? LMAO. If I link to a problem on AMD's site detailing CrossFire problems, or link to an essay where Derek/Anand take a poop on ATI's drivers, am I expressing my opinion based on my experiences? No, I'm not.

You?re right, you simply made the claim that nVidia?s drivers are superior and left out the last bit telling us that you hadn?t touched ATi parts in 7 years.

Again, where have I claimed this? I'm very happy with Nvidia's drivers. Not only with out-of-the-box game support, but timely updates and more recently, significant performance boosts.

Answer the question Chizow: why can you talk about current ATi drivers, having not used an ATi part since 7 years ago, but I can?t do the same despite having current experience with both vendors? latest parts?

Answer the question.
Answer the question Chizow.

Because I'm not claiming any opinion based on my experiences when I've had none recently, unlike you. LMAO.

A bias towards whom? ATi or nVidia?

A clear bias towards ATI, just like the rest of this forum. And to think this poll was taken before there was even a reason to own an ATI part.....wonder what it would look like now?

Why buy a GTX260? The fact that you even have to ask that question demonstrates you know even less about the relative strengths of the vendors than I thought you did. Please conduct some research to educate yourself and then we can continue this conversation.

Rofl, once again dodging the question. You clearly prefer ATI drivers, you've claimed they're superior on numerous occasions. Price is similar to a 1GB 4870. 4870 at the time was stronger in reviews. 4870 is generally stronger with high levels of AA, which you clearly prefer. Given your run of bad "luck" with Nvidia parts, reliability/warranty would also favor ATI in your case. TDP is about the same, thermals as well depending on card or with software adjustment. You don't overclock so that's a non-issue. ATI runs all of your favorite 8-year old games without issue. So ya, it should be a slam dunk. Yet you plunked down another few hundred on Nvidia parts lol. A clear case of saying one thing, yet doing the other. You said your agenda is to spend as much as you want on the vendor of your choice, but what exactly are you basing that choice on?

They sure do

So why not buy a 4870 1GB?

Uh, no. He?s making the obvious point that proprietary standards don?t mix, standards like PhysX.

No, he's making the point they don't mix in their native forms, but already answered they can be made to be compatible with a simple wrapper. Or do I need to quote that part again for you? But the real point is there isn't any problem at all if there's nothing written in Brook+ worth porting.

No they aren?t; CUDA and PhysX are proprietary nVidia standards. Adding OpenCL to the mix doesn?t change anything unless you can demonstrate that you?ll achieve hardware accelerated PhysX on all IHVs by developers running OpenCL code.

Once again, it changes everything if you're claiming incompatibility with industry standard as your contrived reason to downplay PhysX. Its not Nvidia's responsibility to ensure PhysX runs on all IHVs hardware if they make their code compatible with the industry standard, OpenCL, as long as it is in fact compatible which it clearly will be. Or are you still contesting this? As Derek already stated in the article, a simple wrapper for Brook+ or a custom compiler might be needed but the impetus is not on Nvidia to write that code to make it happen, nor should it be. Again, this just once again illustrates AMD is more eager to blow sunshine up your ass than do anything about adding PhysX support for their hardware.

You?re referring to the list from nZone? You?re not being serious, right? LMAO.

Yep, I am being serious. You want to dismiss the list with one hand, yet pimp Havok for being widely adopted and therefore, superior with the other. So again, even if that list does not represent GPU accelerated games, it clearly shows PhysX is everything that Havok is, and more, which makes your attempts at pimping another proprietary solution like Havok even more laughable.

If they don?t support nVidia GPU acceleration then how is PhysX an advantage for nVidia in those titles?

If they?re console games then how is PhysX an advantage for nVidia?

It shows PhysX has already been widely adopted, a claim which you've refuted numerous times. Beyond that its simple, it shows different APIs, hardware, and platforms are a non-issue given PhysX's flexibility and that if a developer or publisher were to choose a single physics SDK for all platforms, PhysX clearly offers the best option.

Nothing was ?forced? into them. The first three had updates to apply cherry-picked PhysX maps which added nothing to the core gameplay, and the last game was designed with hardware PhysX in mind.

The PhysX content existed prior to GPU PhysX. All that was needed was a driver that enabled GPU PhysX, at which point you could enable GPU PhysX. This refutes your earlier claims about GPU PhysX not offering any functionality for legacy PhysX titles, which is clearly false.

Again, we?ve already covered those two games.

Rofl, I guess you looked at the calender and noticed its already January 2009! LMAO.

Likewise there are at least three major studios that have announced DX10.1 support.

DX10.1 ya, right. So once again, what features do you enjoy the most about DX10.1? Its amazing how certain people will attempt to diminish PhysX or compare it to DX10.1, but can't even define the benefits of DX10.1.

It doesn?t matter ? unless OpenCL provides hardware accelerated PhysX on non-nVidia parts then it?s not a standard. This is unlike OpenGL where its code provides hardware acceleration on any part with an OpenGL ICD, and at least four current consumer IHVs have OpenGL ICDs.

So again, did Carmack have to write drivers for the various IHVs for OpenGL support, or was the impetus still on the IHVs? Sure he may write vendor-specific extensions for his engine but that's only after they're compliant with the standard. This is no different than with PhysX, you can lead a horse to water but you can't make it drink.

What on Earth are you babbling about? A game engine is not the same as an API. Making such a comparison is akin to saying Far Cry is the same as DirectX.

The Quake engine runs under OpenGL. So does Quake 3. So does Quake Wars. The engine is irrelevant - all you need is an OpenGL ICD and at least four consumer IHVs currently provide one, which means all of them can run those games.

PhysX is a totally different API to OpenCL. That PhysX can be wrapped to OpenCL is no different to claiming Direct3D and Glide are the same thing because there are Glide wrappers that run under Direct3D.

PhysX isn't just an API, its the front-end SDK, API and driver. If the API and driver are portable and can be made compatible with industry standards then the front-end SDK would function very much like any proprietary game engine.

PhysX Flow Chart

If you look at the flow chart near the bottom and replace the PhysX Driver/HAL with OpenCL as your API, then all you would need are compatible OpenCL wrappers for the various hardware solution drivers.

Furthermore, since PhysX only runs on nVidia cards then it doesn?t really matter if an OpenCL wrapper is made because someone has to port the PhysX back-end to the other IHVs to provide hardware acceleration through that API. That anyone could potentially do it is the same as saying anyone could potentially win lotto as long as they buy a ticket.

So who do you think is responsible for porting the PhysX back-end to the other IHVs or initiating interest?

 

[MegaWorks]

Keys what a think is going on here is that our forums member see that you and your group

members are misleading the public here. For me honesty it sucks to see as a Nvidia

focus member, because you tend to be irrational about the topic and you have to develop

some sort of bias. I mean it's your job and I totally understand that, but you also need to

understand that we're not a bunch of chimpanzee here man. I bought the GTX 260 because

my version is faster than the 4870 1GB. I wasn't even considering PhysX, because I think it's

a big PR BS! Anyway man peace and love and good talking to you again.

 

[nRollo]

Originally posted by: SlowSpyder
You weren't a focus member back then either. But I do remember way back in those days, I was a lurker, don't think I even had an account here yet, just read these forums just about everyday. I remember thinking to myself, "Who is this Rollo guy, and why does he keep talking about how good the 5800/5900 cards are compared to the Radeon 9700/9800's?"

What do ancient video cards have to do with PhysX?

 

[BFG10K]

Originally posted by: chizow

How is it relevant when your experiences with a 4850 had absolutely no influence whatsoever on your idiotic claims about drivers a year ago based on 3 consecutive years of experience with Nvidia parts? LMAO.

Don?t twist the issue. You were asked why you were making driver comparisons between ATi and nVidia despite not having used an ATi part for 7 years while blasting me for not having used one for 3 years. We aren?t talking about me, we?re talking about you.

Answer the question Chizow.
Answer the question.

Prove it. I've said no such thing. I've maintained both IHVs had issues,

So are you saying both drivers are equal? Is that what you?re trying to say now? :roll:

From this very thread:

When a company has to release 3 hot fixes/betas to specifically fix a problem and goes through 3 driver revisions without actually fixing anything, its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline.

I've used both vendors as well thanks, and I've had great experiences with Nvidia drivers.

You made a comparison and your comparison is based on what you had read on then internet about ATi because you hadn?t touched an ATi part in 7 years.

But when I made such a comparison by demonstrating a plethora of issues with nVidia?s Vista drivers and how they were cured when people moved to ATi, you engaged in your hand-waving and tried to claim I wasn?t entitled to make such a comparison because I hadn?t used Vista as my primary gaming OS.

So please stop insulting the collective intelligence of this forum by claiming you have not made any driver comparisons. This is an outright lie.

So again I?ll ask why you can make comparisons about something you have read on the internet but I can?t?

Answer the question Chizow.
Answer the question.

BS, you didn't admit anything until it was shoved down your throat by Ben and Apoppin on these forums.

From the same thread:

http://forums.anandtech.com/me...AR_FORUMVIEWTMP=Linear

Originally posted by: BFG10K

Yep, it looks like ATi are dropping the ball with legacy compatibility on their unified parts too.

This is bad news for consumers all round.

I want a full retraction of your lie Chizow. Right now.

Furthermore, the claims that were being made by the others were to do with palletized textures and API changing yet we know that isn?t the case, so I was 100% correct.

I?ve already quoted the coder of the fix and he clearly said the issues were related to dithering being removed from nVidia drivers and also Z buffering bugs. So you need to retract your lie and retract your erroneous claims about the whole issue.

Rofl, OCD much?

Even after repeated attempts you still fail to answer basic questions. I thought that perhaps repetition would assist you, but it appears not.

You made a claim that ATI's monthly driver updates were superior to Nvidia's because you knew when an update was coming and if there was a problem, you would likely get a fix in a month. So again, if a problem in a AAA title has a hot fix released to specifically fix it, but still lingers over 2 additional hot fixes specifically targetting it and is still a problem after 3 WHQL releases, do you think the monthly driver program is working as you described? Or if the similar occurs for a pervasive problem with CF in Vista that lingers on in various permutations still, after 3 driver releases, do you think the monthly driver program is working?

And do you think nVidia?s driver schedule is working given the alt-tab issues lasted since at least 2004 to 2008? Or how about the Unreal 2 stuttering issue that lasted from Nov 06 to Jul 08. Or how about numerous TWIMTBP titles having issues for months on nVidia parts but not on ATi parts?

Oh that?s right, when I mentioned these problems you claimed they didn?t exist and/or that I wasn?t entitled to point them out because I hadn?t used Vista as my primary gaming OS. Even while nVidia?s release notes contained the fixes you were still denying they ever happened.

Nope, I was just quoting the experiences of people with actual experience with a wide variety of parts from both vendors who clearly felt ATI's drivers were worst than Nvidia's right now. Given they do this for a living and have simultaneous experience with said parts at any given time, their opinion actually means something, unlike your weighted and biased snapshots in time dominated by Nvidia parts.

For what other purpose were you quoting a comparison of drivers, other than to make a comparison of drivers yourself?

Yep, and I said that in the comments, which you selectively ignored to quote.

Bullshit you did. Show me a direct quote and a link.

And if you did, you need to retract all of your arguments you were using when you claimed nVidia was no worse than ATi. After all, according to you given they do this for a living and have simultaneous experience with said parts at any given time, their opinion actually means something,, that means when Derek says ATi?s drivers were better than nVidia?s, that means something to you.

If I link to a problem on AMD's site detailing CrossFire problems, or link to an essay where Derek/Anand take a poop on ATI's drivers, am I expressing my opinion based on my experiences? No, I'm not.

But I linked to multiple pages but you dismissed them stating I couldn?t make any such claims because I hadn?t used recent hardware.

So again I?ll ask why can you link to websites but I can?t?

Answer the question Chizow.
Answer the question.

Again, where have I claimed this? I'm very happy with Nvidia's drivers. Not only with out-of-the-box game support, but timely updates and more recently, significant performance boosts.

So again I?ll ask, what are you claiming? That the drivers are equal? If they?re equal why bother telling us about the three hot-fixes when nVidia has the same problem?

Now, if you don?t think they?re equal then you must be making comparison. You can?t say ?A and B aren?t equal, but I?m not comparing them?. That?s nonsensical - you?re simply playing rhetorical games.

Because I'm not claiming any opinion based on my experiences when I've had none recently, unlike you.

Yes you are. That?s exactly what you?re doing. You?re making claims and opinions about monthly WHQL drivers despite having not used an ATi part for the last 7 years, because of what you read on the internet.

Are you saying its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline. is neither a claim or opinion?

But when I did the same based on numerous internet links, you said I couldn?t make such claims or opinions because I hadn?t used recent hardware.

Answer the question: why can you make claims and opinions based on what you read on the internet but I can?t?

Answer the question Chizow.
Answer the question.

A clear bias towards ATI, just like the rest of this forum.

But by your own admission I?ve spent far more on nVidia?s parts. That?s a strange sort of ATi bias, wouldn?t you say?

Rofl, once again dodging the question.

But what is there to dodge? You claimed I?m biased against ATi but you keep scratching your head as to why I keep buying nVidia parts. Perhaps if you considered your original claim was wrong you might understand the bigger picture.

You said your agenda is to spend as much as you want on the vendor of your choice, but what exactly are you basing that choice on?

Lots of things, the kinds of things someone would know had they actively been using both vendors? parts since 2002. But I wouldn?t expect you to know that since you haven?t used an ATi part since 7 years ago and only have what is posted on the internet to go off.

So why not buy a 4870 1GB?

Because Red Faction isn?t the only reason I buy video cards. But sure, if one vendor runs it but the other doesn?t, I?ll point it out. Stop by Rage3D sometime and check the regular ATi bug reports I file there.

Once again, it changes everything if you're claiming incompatibility with industry standard as your contrived reason to downplay PhysX. Its not Nvidia's responsibility to ensure PhysX runs on all IHVs hardware if they make their code compatible with the industry standard, OpenCL, as long as it is in fact compatible which it clearly will be. Or are you still contesting this?

It?s not an industry standard if only one vendor has chosen to write the implementation. That?s the crux of the issue. Just because nVidia comes up with an OpenCL wrapper for PhysX on their boards, it does not mean you?ll get hardware accelerated PhysX on all other IHVs by the developers simply writing OpenCL code.

You can?t say ?it?s an industry standard? and then turn around and say ?yeah well, it only runs on nVidia, but it?s not their problem?.

Yep, I am being serious. You want to dismiss the list with one hand, yet pimp Havok for being widely adopted and therefore, superior with the other. So again, even if that list does not represent GPU accelerated games, it clearly shows PhysX is everything that Havok is, and more, which makes your attempts at pimping another proprietary solution like Havok even more laughable.

Uh, what? How does it show PhysX is ?everything that havoc is?? If you want to list all PhysX games then we should do the same for Havok. That leads us to over 150 games and over 70 developers.

It shows PhysX has already been widely adopted, a claim which you've refuted numerous times.

But again, if you?re going purely for adoption rates then it?s a drop in the bucket compared to Havok.

The PhysX content existed prior to GPU PhysX. All that was needed was a driver that enabled GPU PhysX, at which point you could enable GPU PhysX. This refutes your earlier claims about GPU PhysX not offering any functionality for legacy PhysX titles, which is clearly false.

So you?re claiming any existing PhysX title will show a benefit from nVidia hardware acceleration? Please provide evidence of this claim or retract it immediately.

Its amazing how certain people will attempt to diminish PhysX or compare it to DX10.1, but can't even define the benefits of DX10.1.

Again, tell us what nVidia doesn?t support and we continue this discussion. There must be something or Blizzard, EA, et al, wouldn?t support it.

So again, did Carmack have to write drivers for the various IHVs for OpenGL support, or was the impetus still on the IHVs? Sure he may write vendor-specific extensions for his engine but that's only after they're compliant with the standard. This is no different than with PhysX, you can lead a horse to water but you can't make it drink.

Right, exactly. The IHVs implement an OpenGL ICD which is why it became an open standard. However at this time only one vendor implements PhysX and that?s the whole point. Something can?t be an open standard if only one vendor supports it.

PhysX isn't just an API, its the front-end SDK, API and driver. If the API and driver are portable and can be made compatible with industry standards then the front-end SDK would function very much like any proprietary game engine.

But it isn?t currently compatible with non-nVidia parts.

If you look at the flow chart near the bottom and replace the PhysX Driver/HAL with OpenCL as your API, then all you would need are compatible OpenCL wrappers for the various hardware solution drivers.

Right, and again, where are these wrappers? Do you think ATi or Intel will implement them and promote a competitor?s proprietary physics solution?

So who do you think is responsible for porting the PhysX back-end to the other IHVs or initiating interest?

Well that?s the $64,000 question, isn?t it? The IHVs aren?t going to support a competitor?s tech so the burden will either fall onto nVidia or the developer. Regardless, the point is if no-one does it then PhysX will remain propriety, OpenCL or no OpenCL.

 

[Atechie]

Originally posted by: BFG10K

Originally posted by: chizow

My comments were made over a year before you decided to update your frame of reference, so having never used a 4XXX part is irrelevant.

Actually it?s very relevant given you have yet to update your frame of reference. You also continue to make comparisons that you forbad me from making before I updated my frame of reference, despite you still not updating yours.

You stated you felt ATI's drivers were superior to Nvidia's based largely on your experience with the G80's launch drivers at a time when you had not used an ATI part in ~3 years. Are you contesting this or are you going to just try and cover your tracks with misdirection as you usually do?

You stated you felt nVidia?s drivers are superior to ATi?s despite not having used an ATi part in ~7 years. Are you contesting this or are you going to just try and cover your tracks with misdirection as you usually do?

So you were wrong that it was an Nvidia-only problem. When asked repeatedly, you insisted the problem did not exist on ATI parts, yet they did.

Nope, when I found the dithering affected ATi parts (in that same thread) I admitted it was an issue. Not only that but I?ve petitioned to get it back @ Rage3D. Of course you wouldn?t know any of this given you spend your days marketing nVidia. That and your ?understanding? of the issue is as woeful as your comments about it.

When a company has to release 3 hot fixes/betas to specifically fix a problem and goes through 3 driver revisions without actually fixing anything, its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline.

Answer the question: why can you make comments based on what you?ve read on the internet but I can?t?

Answer the question.
Answer the question Chizow.

Answer the question: why can you make comparisons of ATi vs nVidia drivers despite you not having used an ATi part for 7 years, while you lambasted me for making such a comparison because I hadn?t used an ATi part for 3 years?

Answer the question.
Answer the question Chizow.

Answer the question: why can you discuss 4xxx parts despite you never having used them, while lambasting me for not using Vista as my primary gaming OS?

Answer the question.
Answer the question Chizow.

I've used both vendors as well thanks, and I've had great experiences with Nvidia drivers. Nvidia has been so good that ATI has given me no reason to spend money on their parts since the 9700pro.

You haven?t used both vendors in 7 years but you?re still making comparisons. I?ve used both vendors? current parts and you still were telling me that I couldn?t make a comparison because ?I swapped the cards too early? or some such nonsense.

Answer the question: why can you make comparisons but I can?t?

Answer the question.
Answer the question Chizow.

I acknowledged his comments that he felt ATI's drivers were better at the time, but I'd say that opinion has clearly changed.

So do you then acknowledge ATi?s drivers were superior at the time? Or do you simply acknowledge his comments when you think he?s changed his mind and it backs your agenda?

Anyway, we aren?t talking about Derek, we?re talking about you given you haven?t touched an ATi part in 7 years.

To quote you again: I just pointed out your perspective was clearly limited to your current experiences,.

Since your current experience with ATi parts is zero, this means your perspective is zero in terms of a comparison and also in terms of commenting about current ATi drivers and parts.

Answer the question: is the above comment true or is it a lie?

Answer the question.
Answer the question Chizow.

Uh, no, I haven't made any idiotic claims like "Nvidia drivers are far superior to ATI's based on my extensive experience with both" when my last ATI part was a 9700pro.

You?re right, you simply made the claim that nVidia?s drivers are superior and left out the last bit telling us that you hadn?t touched ATi parts in 7 years.

Answer the question Chizow: why can you talk about current ATi drivers, having not used an ATi part since 7 years ago, but I can?t do the same despite having current experience with both vendors? latest parts?

Answer the question.
Answer the question Chizow.

I don't have a problem pointing out a product's flaws, its just clearly obvious your bias prevents you from doing so equally.

A bias towards whom? ATi or nVidia?

For example, in your latest act of purchasing hypocrisy, why buy a GTX 216 instead of a 1GB 4870 given price was similar, ATI drivers are superior in your opinion, and the 4870 generally outperformed the GTX 260 in reviews? You don't overclock, warranty obviously does you no good wherever you live, and you clearly have a preference for ATI parts. If there was ever a time to break your Nvidia addiction, surely this would be it. Yet you bought an Nvidia part.

Why buy a GTX260? The fact that you even have to ask that question demonstrates you know even less about the relative strengths of the vendors than I thought you did. Please conduct some research to educate yourself and then we can continue this conversation.

But ATI parts run Red Faction better right?

They sure do

He's assuming there's something written in Brook+ worth porting, when there isn't, so there is no problem.

Uh, no. He?s making the obvious point that proprietary standards don?t mix, standards like PhysX.

That's something they need to work out, the point is that PhysX and CUDA are fully compatible with the industry standard, something you contested which is clearly FUD.

No they aren?t; CUDA and PhysX are proprietary nVidia standards. Adding OpenCL to the mix doesn?t change anything unless you can demonstrate that you?ll achieve hardware accelerated PhysX on all IHVs by developers running OpenCL code.

Its already been linked to you numerous times.

You?re referring to the list from nZone? You?re not being serious, right? LMAO.

All those titles have licensed versions of the PhysX SDK, even if they do not offer GPU accelerated support.

If they don?t support nVidia GPU acceleration then how is PhysX an advantage for nVidia in those titles?

If they?re console games then how is PhysX an advantage for nVidia?

Contrary to what you stated earlier, there have been legacy games that have had GPU PhysX forced on them, UT3, GRAW/2, and Warmonger and there are 2 new titles releasing in weeks that were developed with PhysX from the ground-up.

Nothing was ?forced? into them. The first three had updates to apply cherry-picked PhysX maps which added nothing to the core gameplay, and the last game was designed with hardware PhysX in mind.

And before you drivel on about unreleased games, Cryostasis has already been released in Russia, I'm sure the delay is only due to localization. Mirror's Edge releases later this week and given its already out on PS3 and 360 I think its safe to say its going to release and not turn into PhysX vaporware.

Again, we?ve already covered those two games.

For the future we have 3 major studios, EA, 2K, and THQ that recently announced PhysX support to add to the likes of Epic and GameBryo that already use PhysX in their titles.

Likewise there are at least three major studios that have announced DX10.1 support.

That's the point, it won't become obsolete because OpenCL is now the standard, PhysX will build on that standard going forward just as the various Quake engines built on OpenGL.

It doesn?t matter ? unless OpenCL provides hardware accelerated PhysX on non-nVidia parts then it?s not a standard. This is unlike OpenGL where its code provides hardware acceleration on any part with an OpenGL ICD, and at least four current consumer IHVs have OpenGL ICDs.

Did people stop using the Quake engine because it was proprietary and they had to pay for it? Of course not.

What on Earth are you babbling about? A game engine is not the same as an API. Making such a comparison is akin to saying Far Cry is the same as DirectX.

The Quake engine runs under OpenGL. So does Quake 3. So does Quake Wars. The engine is irrelevant - all you need is an OpenGL ICD and at least four consumer IHVs currently provide one, which means all of them can run those games.

PhysX is a totally different API to OpenCL. That PhysX can be wrapped to OpenCL is no different to claiming Direct3D and Glide are the same thing because there are Glide wrappers that run under Direct3D.

Furthermore, since PhysX only runs on nVidia cards then it doesn?t really matter if an OpenCL wrapper is made because someone has to port the PhysX back-end to the other IHVs to provide hardware acceleration through that API. That anyone could potentially do it is the same as saying anyone could potentially win lotto as long as they buy a ticket.

How can a troll like this be a moderator on a forum?
This post contains the largest amount of fallacies I have ever seen in a post i years

 

[AmberClad]

Originally posted by: Atechie
How can a troll like this be a moderator on a forum?
This post contains the largest amount of fallacies I have ever seen in a post i years

Unless the moderator bolds/signs their post (like I am doing so here), then the person is posting as a regular member and not as a mod. As such, they're entitled to express their opinions like any other member and their mod status does not factor in.

Please refrain from the personal attacks (calling members trolls) and off-topic comments in the future.

- AmberClad (Video Moderator)

 

[chizow]

Originally posted by: BFG10K
Don?t twist the issue. You were asked why you were making driver comparisons between ATi and nVidia despite not having used an ATi part for 7 years while blasting me for not having used one for 3 years. We aren?t talking about me, we?re talking about you.

Answer the question Chizow.
Answer the question.

Again, you don't seem to understand the difference. I haven't made any claims based on my experiences with the parts, unlike you. If I link and reference to recent problems, those would be based on other people's experiences. There is a clear difference, one you don't seem to be able to distinguish between. Now, going back to comments made before, you claimed ATI's drivers were better than Nvidia's based on your experiences with parts from both vendors when you had not used an ATI part in 3 years with at least an overall 3:1 ratio in favor of Nvidia parts.

So are you saying both drivers are equal? Is that what you?re trying to say now? :roll:

From this very thread:

When a company has to release 3 hot fixes/betas to specifically fix a problem and goes through 3 driver revisions without actually fixing anything, its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline.

I've used both vendors as well thanks, and I've had great experiences with Nvidia drivers.

You made a comparison and your comparison is based on what you had read on then internet about ATi because you hadn?t touched an ATi part in 7 years.

Like I said before, I'm sure both vendors have issues, but I'm not going to make an idiotic claim that one has better drivers than the other based on my experiences when I haven't used an ATI part in 3 years, like you did.

But when I made such a comparison by demonstrating a plethora of issues with nVidia?s Vista drivers and how they were cured when people moved to ATi, you engaged in your hand-waving and tried to claim I wasn?t entitled to make such a comparison because I hadn?t used Vista as my primary gaming OS.

Rofl? Yes I'm sure you linked to numerous cases of people being "cured" by moving to ATI. I remember quite well, you went on about the Vista Ready lawsuit, which turned out to be a joke like your argument. At which point I argued it was obvious why Nvidia users would have more problems under Vista as their parts were the only parts worth buying at the time. Numerous Vista Hot Fixes had eased the problem greatly in the July/August timeframe and were recommended as vital updates by both Nvidia and game developers as fixes for the TDM errors. And I can say I experienced this since I did use Vista at the time. But since you seem to think these problems were Nvidia only, maybe you can "cure" these problems too?

22 pages of ATI ATIKMDAG Error

So please stop insulting the collective intelligence of this forum by claiming you have not made any driver comparisons. This is an outright lie.

So again I?ll ask why you can make comparisons about something you have read on the internet but I can?t?

Answer the question Chizow.
Answer the question.

I've never claimed that I've never made comparisons, I just haven't made any idiotic claims that they were based on my experiences when they weren't, like you did. You still don't seem to understand this. You clearly and repeatedly claimed ATI drivers were better, in your experience, when you had not used an ATI part in over 3 years. Or are you disputing this? If you quote something from the internet, is that your experience, or someone elses? This is not a hard question, I think even the collective intelligence of this forum would agree......

From the same thread:

http://forums.anandtech.com/me...AR_FORUMVIEWTMP=Linear

Originally posted by: BFG10K

Yep, it looks like ATi are dropping the ball with legacy compatibility on their unified parts too.

This is bad news for consumers all round.

I want a full retraction of your lie Chizow. Right now.

Furthermore, the claims that were being made by the others were to do with palletized textures and API changing yet we know that isn?t the case, so I was 100% correct.

I?ve already quoted the coder of the fix and he clearly said the issues were related to dithering being removed from nVidia drivers and also Z buffering bugs. So you need to retract your lie and retract your erroneous claims about the whole issue.

Lie? Erroneous claims? No, I never cared about the details, I stopped caring about the details of the issue once it proved you were wrong that Thief 2 was an Nvidia-only problem when it clearly was not. You didn't admit anything until it was proven to you by someone who owned an ATI part, which again, proved my claim your frame of reference precluded you from making any relevant comparison based on your experience.

Even after repeated attempts you still fail to answer basic questions. I thought that perhaps repetition would assist you, but it appears not.

Repeating an irrelevant question repeatedly and throwing a tantrum isn't going to get you an answer any more than stating it once. It just makes you look childish.

And do you think nVidia?s driver schedule is working given the alt-tab issues lasted since at least 2004 to 2008? Or how about the Unreal 2 stuttering issue that lasted from Nov 06 to Jul 08. Or how about numerous TWIMTBP titles having issues for months on nVidia parts but not on ATi parts?

Oh that?s right, when I mentioned these problems you claimed they didn?t exist and/or that I wasn?t entitled to point them out because I hadn?t used Vista as my primary gaming OS. Even while nVidia?s release notes contained the fixes you were still denying they ever happened.

Their driver release schedule has worked great for me and as I said then, especially with newer titles. I also stated I didn't particularly know or care about the issues you were concerned about and that most buyers of new cards were interested in new games. I never outright dismissed them, I just linked to various problems with ATI parts showing they were far from perfect as well. At which point you went on again about how ATI's monthly driver schedule was better because you got quality and faster fixes and knew when you'd get updates etc. So again, do you still feel that's true? Or are you just getting whatever is ready at 4:30pm PST on Friday before cut-off day?

For what other purpose were you quoting a comparison of drivers, other than to make a comparison of drivers yourself?

And again, I've never said I wasn't making a comparison or pointing out a problem with ATI's drivers, I just wasn't going to claim them as my experience.

Bullshit you did. Show me a direct quote and a link.

And if you did, you need to retract all of your arguments you were using when you claimed nVidia was no worse than ATi. After all, according to you given they do this for a living and have simultaneous experience with said parts at any given time, their opinion actually means something,, that means when Derek says ATi?s drivers were better than nVidia?s, that means something to you.

Nvidia's Unified Drivers are back!
Probably easier if you just thoroughly re-read that. I thanked him for his input after he directly stated he felt Nvidia had more problems and also stated I felt input from him and other reviewers was particularly relevant because they had direct access to comprehensive array of hardware from both vendors simultaneously. He also acknowledged hot fixes and 3rd party apps and drivers also had an impact on the driver situation and directly refuted claims by you about ATI drivers not breaking stuff. He also detailed why he thought Monthly drivers were ineffective.

There's nothing for me to retract as I've maintained throughout both vendors have problems. Based on Derek's opinion Nvidia may have been worst during the time in question, but his opinion has certainly changed now.

But I linked to multiple pages but you dismissed them stating I couldn?t make any such claims because I hadn?t used recent hardware.

So again I?ll ask why can you link to websites but I can?t?

Answer the question Chizow.
Answer the question.

Again, you don't seem to understand this simple concept. If you link to external sources, you can no longer claim your opinion is based on your experiences with the hardware.

So again I?ll ask, what are you claiming? That the drivers are equal? If they?re equal why bother telling us about the three hot-fixes when nVidia has the same problem?

Now, if you don?t think they?re equal then you must be making comparison. You can?t say ?A and B aren?t equal, but I?m not comparing them?. That?s nonsensical - you?re simply playing rhetorical games.

I've never claimed anything more than both vendors have problems and that Nvidia's drivers have been very good in my experience. You made idiotic claims "based on your experiences" that clearly fell out of that realm of experience. You further extended this by claiming ATI's monthly driver program was superior to Nvidia's TWIMTBP and cumulative update model.

I'm asking a simple question, do you still feel the Monthly program is better, or that its giving you "rapid progress"? You cry about Nvidia's leaked, unofficial Betas and lack of official WHQL but really, what exactly makes ATI's drivers better, the "official" status even if they're broken and don't fix anything at all?

Yes you are. That?s exactly what you?re doing. You?re making claims and opinions about monthly WHQL drivers despite having not used an ATi part for the last 7 years, because of what you read on the internet.

Are you saying its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline. is neither a claim or opinion?

But when I did the same based on numerous internet links, you said I couldn?t make such claims or opinions because I hadn?t used recent hardware.

Answer the question: why can you make claims and opinions based on what you read on the internet but I can?t?

Answer the question Chizow.
Answer the question.

Yep, my opinion is that ATI's monthly marketing placebo driver model is garbage, but that's not based on my experiences, nor am I claiming it is. That's based on other's people's experiences, observation and just common sense. When a hot fix is released for a specific problem and it doesn't actually fix anything, and the problem still isn't fixed over numerous driver revisions, or fixes aren't carried over from WHQL to WHQL, or a "newer" driver is actually older than a previous driver, or the common fix is to revert to a driver that is older than the latest driver, its obvious to me the monthly driver program isn't accomplishing the goals it is commonly praised for.

But by your own admission I?ve spent far more on nVidia?s parts. That?s a strange sort of ATi bias, wouldn?t you say?

Yep it just shows you have conflicted opinions, an agenda or hold Nvidia parts to higher standards.

But what is there to dodge? You claimed I?m biased against ATi but you keep scratching your head as to why I keep buying nVidia parts. Perhaps if you considered your original claim was wrong you might understand the bigger picture.

I do understand the bigger picture, you prefer Nvidia parts and they've bled you dry over the years, so you expect more from them. As such you're willing to devote the time and effort to expose their problems, but since they ignore you and your 8 year old games, you've come to hate them and that animosity shows in your posts. Typical love-hate relationship.

But that doesn't answer the question, why would you buy a GTX Core 216 over a 1GB 4870 when it clearly suits your needs better and you feel their drivers are better? Let me guess? PhysX support. LMAO. :laugh:

Lots of things, the kinds of things someone would know had they actively been using both vendors? parts since 2002. But I wouldn?t expect you to know that since you haven?t used an ATi part since 7 years ago and only have what is posted on the internet to go off.

Mmmhmm, I just broke down all the pros/cons based on your preferences, which is why you're ducking the question. Its because you're full of shit and are willing to say one thing and do the complete opposite when it comes buying time.

Because Red Faction isn?t the only reason I buy video cards. But sure, if one vendor runs it but the other doesn?t, I?ll point it out. Stop by Rage3D sometime and check the regular ATi bug reports I file there.

I gave 5-6 other reasons that are clearly more important, but running your favorite 8 year old games should've been icing on the cake considering they played such a large factor in your determination ATI's drivers were better than Nvidia's.

It?s not an industry standard if only one vendor has chosen to write the implementation. That?s the crux of the issue. Just because nVidia comes up with an OpenCL wrapper for PhysX on their boards, it does not mean you?ll get hardware accelerated PhysX on all other IHVs by the developers simply writing OpenCL code.

You can?t say ?it?s an industry standard? and then turn around and say ?yeah well, it only runs on nVidia, but it?s not their problem?.

Rofl what? It is an industry standard. All the logos and representation on the Khronos Group make it so, not your arbitrary criteria. They all know the spec, they know the requirements, its the individual IHVs responsibility to implement it based on the standard, not Nvidia's.

Uh, what? How does it show PhysX is ?everything that havoc is?? If you want to list all PhysX games then we should do the same for Havok. That leads us to over 150 games and over 70 developers.

I'm not basing it on adoption, I'm basing it on capability, which you finally seem to be coming to grips with. If you're going to claim Havok is somehow superior because of its flexibility you need to acknowledge PhysX can do everything Havok can, and *MORE*. Those PhysX lists prove just that, PhysX in games on multiple platforms with various hardware implementations.

But again, if you?re going purely for adoption rates then it?s a drop in the bucket compared to Havok.

I'm not basing anything on adoption rate, I'm basing it on capability. You've claimed Havok is the better option, when it clearly is not as PhysX can do everything Havok can, and much much more.

So you?re claiming any existing PhysX title will show a benefit from nVidia hardware acceleration? Please provide evidence of this claim or retract it immediately.

Nope, I've never made that claim. You claimed legacy PhysX titles did not benefit from GPU acceleration like games did with 32-bit color and/or HW T&L, when that was clearly not true. A handful of legacy titles benefitted from GPU acceleration, just as a handful benefitted in examples you brought up.

Again, tell us what nVidia doesn?t support and we continue this discussion. There must be something or Blizzard, EA, et al, wouldn?t support it.

What's there to discuss about DX10.1? Nothing. lol.

Right, exactly. The IHVs implement an OpenGL ICD which is why it became an open standard. However at this time only one vendor implements PhysX and that?s the whole point. Something can?t be an open standard if only one vendor supports it.

OK, so again, is Carmack writing OpenGL drivers for all the different IHVs. No, he's not. It doesn't matter if only one vendor implements PhysX, just as it didn't matter that only one vendor implemented OpenGL via wrapper for GLQuake. For someone who was just arguing the other side of the coin this shouldn't be a difficult parallel to draw.

But it isn?t currently compatible with non-nVidia parts.

Guess they better get working on some drivers/wrappers then. Or they can keep blowing sunshine.

Right, and again, where are these wrappers? Do you think ATi or Intel will implement them and promote a competitor?s proprietary physics solution?

Who knows, but they sure as hell aren't going to be able to hide behind the "proprietary standards" excuse much longer. In the end adoption won't be up to ATI or Intel, it'll be up to developers and making PhysX compatible with the industry standard OpenCL is a great step in that direction.

Well that?s the $64,000 question, isn?t it? The IHVs aren?t going to support a competitor?s tech so the burden will either fall onto nVidia or the developer. Regardless, the point is if no-one does it then PhysX will remain propriety, OpenCL or no OpenCL.

Again, that's certainly the decision of the individual IHVs, they're just going to have to start coming up with better excuses for not implementing PhysX than "its proprietary", or better yet, maybe they'll come up with a hardware solution of their own.

 

[BFG10K]

Originally posted by: chizow

Again, you don't seem to understand the difference. I haven't made any claims based on my experiences with the parts, unlike you. If I link and reference to recent problems, those would be based on other people's experiences.

But I linked to other peoples? experiences as well but you claimed they didn?t count. Here are some choice quotes from you about it:

http://forums.anandtech.com/me...AR_FORUMVIEWTMP=Linear

Originally posted by: chizow

Again, a statement you can't back up with hard data or even comprehensive personal experience.

Well lets see: you have no personal experience with NV cards in Vista and you have no personal experience with ATI cards in Vista. So that leaves you with second-hand experience and conjecture as the only basis for your generalizations

Online feedback = not your experiences.

My point stands. No relevant experience, just conjecture and second-hand feedback.

So again I?ll ask why can you quote online feedback, but dismiss it when I use it?

There is a clear difference, one you don't seem to be able to distinguish between.

Funny, you couldn?t seem to distinguish in that other thread when I was telling you online evidence was backing my claims, as well as personal experiences. You just kept repeating I couldn?t use online feedback because I didn?t have relevant experience, as indicated in the above quotes from you.

Yet you appear to have no qualms to make sweeping generalizations about the state of monthly drivers from ATi despite having not touched an ATi part for seven years. Your rampant double-standards are hardly surprising.

Like I said before, I'm sure both vendors have issues, but I'm not going to make an idiotic claim that one has better drivers than the other based on my experiences when I haven't used an ATI part in 3 years, like you did.

So if you?re saying you?re not making any claim about who is better, when you say the following:

When a company has to release 3 hot fixes/betas to specifically fix a problem and goes through 3 driver revisions without actually fixing anything, its clearly obvious monthly drivers are just marketing fluff. You're not getting quality, you're getting whatever slop they manage to throw together to meet a monthly deadline.

Are you stating nVidia and ATi are equal?

And again, you?re making sweeping generalizations about monthly driver support despite having no relevant experience in the last 7 years with them. To quote you above: my point stands. No relevant experience, just conjecture and second-hand feedback.

Every comment you now make is an absolute farce given it?s possible to contradict it with some other random comment you made in the other thread.

Rofl? Yes I'm sure you linked to numerous cases of people being "cured" by moving to ATI. I remember quite well, you went on about the Vista Ready lawsuit, which turned out to be a joke like your argument. At which point I argued it was obvious why Nvidia users would have more problems under Vista as their parts were the only parts worth buying at the time.

But then this was argued by Derek when he said outright that your claims were inaccurate. He also stated the GF7 parts actually had more issues than the GF8 parts, thereby nullifying your claims that it was a new architecture and DX10 that somehow excused nVidia from the issue.

Since Derek (and other reviewers) actually used both parts but you hadn?t their comments hold more weight than yours since according to your words, your perspective on the issue is limited by your experience, which is zero.

Again Derek?s comments backed my claims that ATi?s drivers were superior to nVidia?s in the earlier Vista days. Are you going to retract your claims to the contrary in the other thread now, or are you going to keep playing your little games?

Lie? Erroneous claims? No, I never cared about the details, I stopped caring about the details of the issue once it proved you were wrong that Thief 2 was an Nvidia-only problem when it clearly was not.

But the sky problem was an nVidia only problem. Are you going to admit that now?
The Red Faction problem was an nVidia only problem. Are you going to admit that now?

The other part was that dithering was an nVidia driver issue regardless of whether ATi had it and, I still have yet to get an admission from you.

Answer the question: do you acknowledge existence of the two nVidia driver issues, as noted by the coder of the fix?

At which point you went on again about how ATI's monthly driver schedule was better because you got quality and faster fixes and knew when you'd get updates etc. So again, do you still feel that's true?

Yes, absolutely, but then I?ve actually experienced recent benefits of monthly WHQL drivers (namely fast official application fixes), unlike you.

Probably easier if you just thoroughly re-read that. I thanked him for his input after he directly stated he felt Nvidia had more problems and also stated I felt input from him and other reviewers was particularly relevant because they had direct access to comprehensive array of hardware from both vendors simultaneously.

I?m not going to re-read anything. You made the claim so now provide the evidence. Please provide direct link & quotes or retract your claims. Thanks.

You felt his input was relevant yet you never retracted your arguments when I was arguing exactly what Derek and other reviewers confirmed later ? that nVidia indeed had worse drivers than ATi on Vista. Are you going to admit I turned out to be right or are you still going to play rhetorical games?

I don?t give a shit if you think you?re using personal experiences or not. Stop playing games and answer the question: did Derek (and other reviewers) end up backing my claims?

He also acknowledged hot fixes and 3rd party apps and drivers also had an impact on the driver situation and directly refuted claims by you about ATI drivers not breaking stuff. He also detailed why he thought Monthly drivers were ineffective.

He also pointed out that if third party issues were breaking nVidia?s drivers more then it means their drivers aren?t as robust. Do you admit he?s right?

There's nothing for me to retract as I've maintained throughout both vendors have problems. Based on Derek's opinion Nvidia may have been worst during the time in question, but his opinion has certainly changed now.

Don?t change the issue; will you admit you were wrong given Derek (and other reviewers) ended up backing my claims that ATi had better drivers than nVidia during the early Vista days?

Yep, my opinion is that ATI's monthly marketing placebo driver model is garbage, but that's not based on my experiences, nor am I claiming it is. That's based on other's people's experiences, observation and just common sense.

But I had the opinion that nVidia?s drivers weren?t as robust as ATi?s on Vista and that's based on other's people's experiences, observation and just common sense, and Derek (and other reviewers) ended up agreeing with me. Are you going to retract your erroneous claims now?

Yep it just shows you have conflicted opinions, an agenda or hold Nvidia parts to higher standards.

Nope ? the same standards, namely expecting games to work.

I do understand the bigger picture, you prefer Nvidia parts and they've bled you dry over the years, so you expect more from them. As such you're willing to devote the time and effort to expose their problems, but since they ignore you and your 8 year old games, you've come to hate them and that animosity shows in your posts. Typical love-hate relationship.

I expect exactly the same of my ATi card. Again go to Rage3D and check out any of my bug reports. Since ATi have an online bug report form there?s no need for me to get as vocal since they not only read my reports but I get fixes very quickly (in many cases in the next official driver).

But that doesn't answer the question, why would you buy a GTX Core 216 over a 1GB 4870 when it clearly suits your needs better and you feel their drivers are better?

AF, AA, and application profiles. Of course these concepts would be alien to someone that has no clue as to what the real differences of the vendors are. Such a person would also repeatedly have to ask as to why I would pick an nVidia card.

Its because you're full of shit and are willing to say one thing and do the complete opposite when it comes buying time.

Actually you?re full of shit with your blatant double standards. You were also full of shit about your claims about Vista driver quality given Derek (and other reviewers) ended up backing my claims, people that had actually used the hardware unlike you.

Again I?m still waiting for a retraction from you about the issue, along with a retraction that online feedback isn?t valid.

 

[Keysplayr]

Guys, we really do know what each of you has said. Your arguments are full throttle to destination nowhere.
Say when for all our sakes!