Why would anyone consider having their nuts consumed like, well like a nut, a reward for successfully predicting the future, or for successfully doing anything for that matter?
funny video but might be NSFW for implied content
Why would anyone consider having their nuts consumed like, well like a nut, a reward for successfully predicting the future, or for successfully doing anything for that matter?
funny video but might be NSFW for implied content
I never mentioned OoO. OoO has nothing to do with this.
Threading is VERY different. Each (logical) x86 core has its own context. This means that each core can run a thread completely independently, and do its own branching etc. They don't even have to be the same thread, each core can run its own code.
GPGPUs are made for running large batches of the same piece of code in parallel. And they don't really branch as such, they just use masking and regroup threads from time to time to remove the 'dead' ones.
An x86 has a more generic and flexible threading and programming model than what you can use with GPGPU/DX11.
The SIMD units on the x86 cores can perform GPGPU tasks, but not the other way around.
I'm convinced that Intel is not stupid enough to not add AF to their hardware, however they plan to implement it. Intel isn't going to spend this much time and money on Larrabee only to come up with a chip that is useless with AF.
Well if nV is going to be trouble, they're just going to have to buy them out then, won't they?
I really don't see the problem.
Uhh, zbuffering is done per pixel, not per triangle.
Doesn't sound like you know how a z-buffered rasterizer works.
They didn't bother to describe AA specifically because it's trivial, as I said, just an extra iteration to refine the grid you're rasterizing.
You seem to sling some vague terms and statements around, but I think you fail to grasp the basics of rasterizing, such as supersampling/multisampling and zbuffering (or the difference between a multicore CPU and GPGPU programming model). You're not fooling me.
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,002
- 126
Oh heck, this should be good. I'm grabbing the popcorn to watch this.
I quickly figured out that arguing with you is a total waste of time, but I don't think Ben will give up so quickly.
I do have to address this woeful inaccuracy from you, however:
Wrong, completely wrong. MSAA debuted in consumer space on nVidia?s GeForce 3 (NV20), which offered variants of both 2xRGMS and 4xOGMS.
The significance of ATi?s R300 was that it offered 4xRGMS and 6xSGMS in consumer space for the first time.
Ben may not give up, doesn't mean he knows what he's talking about though.
Arguing with me is a complete waste of time if you don't know what you're talking about.
Arguing with you is a complete waste of time aswell, I see... since you sling crap like this into a discussion. I am going to have to complain about you yet again.
As far as I recall, GeForce3/4 only used that Quincunx thing, which nVidia may have called multisampling, but is not the same algorithm as the one used by ATi in the R300 and all later GPUs. THAT is the algo that is now known as multisampling in the Direct3D API. Not the one nVidia used.
Some more info... This is multisampling as we know it today in Direct3D and OpenGL:
http://en.wikipedia.org/wiki/Multisampling
"The specification dictates that the renderer evaluate one color, stencil, etc. value per pixel, and only "truly" supersample the depth value."
That is key to both the performance and the quality of AA.
This is what nVidia did with GeForce3/4:
http://www.tomshardware.com/reviews/high,294-23.html
http://www.nvidia.com/attach/151
They don't supersample the depth value any different from the rest of the scene. In fact, the depth value isn't even USED in the actual filtering of the samples.
The only thing that's different from regular supersampling at that time was that it used a clever X-shaped filter kernel rather than just a boxfilter. Which also explains why it was rather blurry, also INSIDE polygons, where MSAA only works on edges (bumpmapping isn't antialiased at all by MSAA, while it IS affected by Quincunx).
Its always hard to argue with someone who constantly makes the assumption that he knows more than everyone else.
Some good advice Scali, don't say things like 'It looks like you don't know how [insert term] works' in every post you make .. it really just makes everyone hostile towards you, and noone care to read your posts except for humor value.
As far as people thinking Larrabee will take the game market with storm, I think there is too much money and dev time behind the current game and console market.
And unless it can somehow magically run all current games at a near equal performance/price point then I do not see why more than a small handful of people would invest in it.
If it can be used together with an ATI/nV GPU in a dual GPU setup, and it somehow has amazing GPGPU/physics performance and is cheap, then maybe it can gain some solid market share.
Some good advice Scali, don't say things like 'It looks like you don't know how [insert term] works' in every post you make .. it really just makes everyone hostile towards you, and noone care to read your posts except for humor value.
As far as people thinking Larrabee will take the game market with storm, I think there is too much money and dev time behind the current game and console market.
And unless it can somehow magically run all current games at a near equal performance/price point then I do not see why more than a small handful of people would invest in it.
If it can be used together with an ATI/nV GPU in a dual GPU setup, and it somehow has amazing GPGPU/physics performance and is cheap, then maybe it can gain some solid market share.
It's not an assumption, Ben clearly lacks understanding of zbuffering. Anyone who knows how zbuffering works, knows that his talk about z-value per triangle doesn't make sense (that's more related to the Painter's algorithm, which we don't use because it can't produce correct results anyway).
The problem is, if everyone else here ALSO doesn't understand zbuffering and rasterizing, they won't know that he's wrong. They can get hostile with me all they want, I don't care.
At least I know what I'm talking about, they don't.
Besides, if you're going to be arrogant and smug like the way Ben chose to argue, you deserve to be called out when you are talking nonsense. You asked for it.
The pursuit of truth is noble and worthy of time and effort. But forget not that it is easier to catch flies with honey, than with vinegar.
The point being if you are going to go to all the effort of presenting your information in a public forum like this, then logic would dictate that you should go to some effort to package it in a friendly digestible format so as all the more people can benefit from the fruits of your labor.
(I am speaking of the people who visit, read, lurk, but do not actively engage you in debate, as well as the people who do)
Yes some thirst for knowledge so badly they will drink it whether it tastes like vinegar or honey, but to use honey is to amplify the impact of your time spent here. The difference between inspiring to learn versus dictating to recite. You're bright, clearly, so I waste your time to push my diatribe any further.
The point being if you are going to go to all the effort of presenting your information in a public forum like this, then logic would dictate that you should go to some effort to package it in a friendly digestible format so as all the more people can benefit from the fruits of your labor.
(I am speaking of the people who visit, read, lurk, but do not actively engage you in debate, as well as the people who do)
Yes some thirst for knowledge so badly they will drink it whether it tastes like vinegar or honey, but to use honey is to amplify the impact of your time spent here. The difference between inspiring to learn versus dictating to recite. You're bright, clearly, so I waste your time to push my diatribe any further.
The thing is that I tried to be friendly, but all I get is insults, people claiming I work for this-or-that company, or taking out some small detail and coming up with some far-fetched argument (like Quincunx that just came up... ridiculous). Most of what I said is either not understood or just ignored, because I never hear people about it. They just want to argue with me to try and make themselves look smart. Then I catch them on a mistake, and their petty ego's are bruised and they get all pathetic and disrespectful.
My patience and friendliness have run out, thanks to the 'vinegar' I was met with when I tried to be friendly and patient in this discussion.
I most certainly didn't start it.
You got to let it mellow a bit, think of a pond, scali is the pebble that plops in. Makes waves. Everything is agitated for a while, the established rank and file don't know where this new guy fits in.
Do they "vet" the new guy a bit first to see what the sticking power is before they worry about it? Do they make room, even it it means maybe they move down the chain of "ranked expert" and take a second seat to you after having been an authority on subject xyz for a year or longer here?
You got to prove yourself, and that means your disposition as well as your knowledge-base. People are receptive when the other person is as well, it takes time. If your goal is to become a member of the existing community then you should want to be accepted by it.
Do you work for yourself (meaning by yourself) or do you work in an office with coworkers? If in an office then you are no doubt familiar with the "pecking order" that exists and everyone understands...if you need xyz you go to Tom, and avoid Jim on that topic at all costs. Etc. Then a new-hire comes in, he may be Einstein in disguise but if the first thing he does is start telling everyone that Tom don't know shit then the new-hire is going to be on everyone's "I hate the new guy" list and it won't matter what brainchild ideas he's got, he's already turned everyone off to listening to him.
Rather you want to see the new guy take some time getting to know his environment. Situational awareness. Embrace the existing hierarchy for a time just to get to know it and for it to get to know you. Then worry about changing the world one post at a time.
You are being vetted, your staying power is questioned, your disposition is questioned. Consider it flattering, it means people think you worth the time to see if you are worth their time. But you got to let some of it mellow, forget 99.9% of the past posts (post history here will only hold you back, forgive and forget, you'll find anyone worth your time will do the same with you and whatever has happened here on the forums prior will not be held against you either).
You are trying to sprint the 100m dash in molasses, take it slower, let us digest it, we'll all change and grow for the better from it. Over time. Not today, not by tomorrow. But eventually, in months. Like I said, you're bright, you add value to this forum, I know 99% of what I write here is already self-evident to you. At the same time I take the time to write them to show you there is method to the madness around here. But yes it is true there are some who are here simply for the madness, the misery loves company types. Avoid them, you know who they are, don't waste your time even if/when they seek you out to waste yours. Be above it.
Do they "vet" the new guy a bit first to see what the sticking power is before they worry about it? Do they make room, even it it means maybe they move down the chain of "ranked expert" and take a second seat to you after having been an authority on subject xyz for a year or longer here?
You got to prove yourself, and that means your disposition as well as your knowledge-base. People are receptive when the other person is as well, it takes time. If your goal is to become a member of the existing community then you should want to be accepted by it.
Do you work for yourself (meaning by yourself) or do you work in an office with coworkers? If in an office then you are no doubt familiar with the "pecking order" that exists and everyone understands...if you need xyz you go to Tom, and avoid Jim on that topic at all costs. Etc. Then a new-hire comes in, he may be Einstein in disguise but if the first thing he does is start telling everyone that Tom don't know shit then the new-hire is going to be on everyone's "I hate the new guy" list and it won't matter what brainchild ideas he's got, he's already turned everyone off to listening to him.
Rather you want to see the new guy take some time getting to know his environment. Situational awareness. Embrace the existing hierarchy for a time just to get to know it and for it to get to know you. Then worry about changing the world one post at a time.
You are being vetted, your staying power is questioned, your disposition is questioned. Consider it flattering, it means people think you worth the time to see if you are worth their time. But you got to let some of it mellow, forget 99.9% of the past posts (post history here will only hold you back, forgive and forget, you'll find anyone worth your time will do the same with you and whatever has happened here on the forums prior will not be held against you either).
You are trying to sprint the 100m dash in molasses, take it slower, let us digest it, we'll all change and grow for the better from it. Over time. Not today, not by tomorrow. But eventually, in months. Like I said, you're bright, you add value to this forum, I know 99% of what I write here is already self-evident to you. At the same time I take the time to write them to show you there is method to the madness around here. But yes it is true there are some who are here simply for the madness, the misery loves company types. Avoid them, you know who they are, don't waste your time even if/when they seek you out to waste yours. Be above it.
I've been reading this thread for what feels like a week now and have been rather entertained by it. I don't fully understand everything that is being said, especially when it begins to get highly technical (like the recent few exchanges between Scali and BenSkywalker), but I feel like I grasp some of the basic concepts better for having read them. So, I guess I'd just like to say, as someone that really just lurks and doesn't feel like he has much in the way of tech-smarts to add to conversations enough to justify posting, that I'm thankful these arguments exist.
I would say though, as a generally impartial (as much as anyone can be I suppose) observer, that some people are a little bit high-strung in here.
I think Forumpanda was rather close to the mark on this one. The problem comes from what appears to be the fact that Scali might have a better working knowledge of these things than most others here.
In reading this thread, you (Scali) certainly do seem like you know your stuff (not that I know enough to say anything in that regard, but one can deduce certain things from the manner of speech -- writing, as it were -- and informed responses to your posts). However, you often come off as rather arrogant when you are forced to defend your statements.
You're right; most people wouldn't have known, at all, who was right and who was wrong on this topic. Hell, I still don't. There is so much foundation involved in understanding this for a layman like myself that I could never feasibly grasp what is going on. However, you're posting on a forum. You seem to be doing so only in a corrective fashion at this point however, and that fails as a method if/when you are wrong. Presupposing you know everything invalidates the necessity of open discourse -- not that I am ascribing that bit of hyperbole to you, but perhaps a subset with relation to the topic at hand is valid?
The problem with that is that perception has a great deal to do with who 'wins' a debate / argument. As well, when communication breaks down to personal jabs and the slandering of credentials, the topic is never resolved. When things do break down before there are clear, agreeable answers, then no one but, perhaps, those involved know exactly what has happened, and certainly no progress has been or can be made. That is precisely a lose-lose situation.
All I'm trying to say is a little diplomacy goes a long way. You don't need to imply that someone is fully ignorant about a topic in order to question their understand of one aspect. Neither do you need to be as confrontational when you do so, as you are in this quote:
You could easily have eluded to a misconception on his part about an issue or two instead of wholesale throwing out his understanding of the topic. That, to me, would have been rather insulting, were I him.
Anyway, you should probably just ignore me and keep on keepin' on. I'm sure I shouldn't be commenting on others like this anyway (sorry if there is a forum rule I'm breaking), but after reading 30 pages of this stuff I really wanted to chime in with something!
I would say though, as a generally impartial (as much as anyone can be I suppose) observer, that some people are a little bit high-strung in here.
I think Forumpanda was rather close to the mark on this one. The problem comes from what appears to be the fact that Scali might have a better working knowledge of these things than most others here.
In reading this thread, you (Scali) certainly do seem like you know your stuff (not that I know enough to say anything in that regard, but one can deduce certain things from the manner of speech -- writing, as it were -- and informed responses to your posts). However, you often come off as rather arrogant when you are forced to defend your statements.
You're right; most people wouldn't have known, at all, who was right and who was wrong on this topic. Hell, I still don't. There is so much foundation involved in understanding this for a layman like myself that I could never feasibly grasp what is going on. However, you're posting on a forum. You seem to be doing so only in a corrective fashion at this point however, and that fails as a method if/when you are wrong. Presupposing you know everything invalidates the necessity of open discourse -- not that I am ascribing that bit of hyperbole to you, but perhaps a subset with relation to the topic at hand is valid?
The problem with that is that perception has a great deal to do with who 'wins' a debate / argument. As well, when communication breaks down to personal jabs and the slandering of credentials, the topic is never resolved. When things do break down before there are clear, agreeable answers, then no one but, perhaps, those involved know exactly what has happened, and certainly no progress has been or can be made. That is precisely a lose-lose situation.
All I'm trying to say is a little diplomacy goes a long way. You don't need to imply that someone is fully ignorant about a topic in order to question their understand of one aspect. Neither do you need to be as confrontational when you do so, as you are in this quote:
You could easily have eluded to a misconception on his part about an issue or two instead of wholesale throwing out his understanding of the topic. That, to me, would have been rather insulting, were I him.
Anyway, you should probably just ignore me and keep on keepin' on. I'm sure I shouldn't be commenting on others like this anyway (sorry if there is a forum rule I'm breaking), but after reading 30 pages of this stuff I really wanted to chime in with something!
Nemesis 1
Lifer
- Dec 30, 2006
- 11,366
- 2
- 0
I really didn't no what to make of scali. But seems to be waymore techy than fanbois. Alot like Duvie.
But keep going I been trying to nail down larrabee forever . I couldn't . Still can't . But I learned something I should have known . But It doesn't help me nail down larrabee . What did I learn . Larrabee= Software= What ever you want it to be . That all I have figured out for FACT other than hardware . What Intel is doing on the software side of the gpu . Isn't real clear to me scali can you help me with the coding thing on Vector unit . Its said larrabee van handle differant code lengths. Can you talk about the longest code the backend on larrabbee can handle. Vs, The longest code used by vector unit. I have always assumed that vector processing would move completely to VLIW backends and in hand helds its proving to be the case . But really I don't understand the differant code lengths larrabee can use . Give me some insight there please.
SunnyD
Belgian Waffler
Did you too get the feeling over the last page or two that this thread has nothing to do anymore with the original topic?
Pass the popcorn please. I brought the :beer: so we can enjoy the ego-fight.
That has much to do with the manner in which I am 'asked' to defend my statements, or in some cases am ascribed things I never said or did.
Well I think the problem here is that some people act like they know everything. I don't know the level of knowledge of the other people in this thread. So if they present themselves as being experts on the topic, I don't feel the need to explain everything, but merely correct them when I see them saying something wrong.
As you can probably tell from my posts, I don't just say "you're wrong" but I actually do go into detail about that particular subject and provide links to relevant information when available.
You're right that there is a LOT of foundation involved. That's the problem, it takes YEARS before one understands how a modern renderer works, with all the little details, effects, optimizations and hacks involved. I've seen that since the GeForce 2 or so, most hardware reviewers don't seem to really know what the cards do anymore either. It's become a sort of 'black art'.
In fact, I noticed that the 'newer generation' of programmers who started out with 3d hardware and API's like Direct3D and OpenGL don't know exactly what the hardware and API's do either.
It's impossible to explain such topics to a 'layman' in a simple forum discussion, you have to have some basic working knowledge.
It's like me going to a fashion forum. I know nothing about fashion. Things like fuschia, plaid, teal... I dunno. I counldn't heads nor tails out of most of the conversation, I'd need to educate myself on some of the basics first. I can't expect people to explain that to every random newbie in every random thread, can I? It would derail all conversation. Maybe there would be a FAQ or something.
Another thing is that nobody really asks questions. I'm willing to explain things, but you have to tell me what you want me to explain. I don't know what people know and what they don't know, or what they want to know more about.
I mainly see people who get angry and rude because they suddenly realize they may not be as smart as they thought they were.
SirPauly
Diamond Member
- Apr 28, 2009
- 5,187
- 1
- 0
Also, the GeForce 4 had a new marketing name called, "Accuview!"
Edit: Been so long and tough to find a link but will use this site's info:
http://www.anandtech.com/video/showdoc.aspx?i=1583&p=3
Edit: Been so long and tough to find a link but will use this site's info:
http://www.anandtech.com/video/showdoc.aspx?i=1583&p=3
Dam Nemesis give the man some time to sleep and eat breakfast :laugh: Or are you testing to see if he is a 24/7 autobot?
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,002
- 126
Are we talking about Ben, or you?
LMFAO, are you for real?
Seriously, I don?t think an open forum is the best place for you. Maybe you should start a blog with you as the sole registered user. That way you can post misinformation and sling insults without fear of retribution. Don?t think you?re going to get a free ride here though.
Who gives a shit what you recall? Are we speaking in facts, or in flawed in recollections like yours?
Your recollection is wrong. Also the fact that you ignored the two examples of MSAA I gave you and instead discussed Quincunx tells me, to quote you, ?that you don?t really understand? the significance of those modes. Also to quote yourself: ?you don?t fool me?.
Did you even read your own links? I was going to provide my own links, but you?ve done all the work for me.
To quote Tom:
Also Anand: http://www.anandtech.com/showdoc.aspx?i=1426&p=7:
If you still don?t get this, I suggest you consult a dictionary for the defintions of terms ?besides?, ?also? and ?combining? so that you understand that Quincunx is made possible by the fact that the hardware offers multi-sampling at the base level.
2xRGMS forms the basis of Quincunx; without it, the GF3?s implementation of Quincunx wouldn?t be possible. This is trivial to check with any AA tester app, and therefore trivial to prove you wrong
Also when you claimed the GF4 only offered super-sampling (?nVidia only had the GeForce4, which only supported SSAA?), you were wrong there as well given by your own admission they supported Quincunx, which is not SSAA.
You need to retract both of your false claims Scali, and stop posting misinformation.
Oh, and don?t forget to report me again for proving you wrong. LMAO.
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,002
- 126
http://developer.nvidia.com/object/geforce3_faq.html
But hey, I guess nVidia, Tom, Anand (et el) are all wrong when they say the card supports MSAA. I'm glad we have Scali here to "correct" them all. :roll:
Ooops, I guess I?ll be reported again for addressing your misinformation.
Hmm, two samples?as in 2xRGMS, which is exactly what I said earlier.
But hey, I guess nVidia, Tom, Anand (et el) are all wrong when they say the card supports MSAA. I'm glad we have Scali here to "correct" them all. :roll:
Ooops, I guess I?ll be reported again for addressing your misinformation.
Keysplayr
Elite Member
- Jan 16, 2003
- 21,211
- 50
- 91
Maybe you guys should stick to "current" events? Thread IS about Nvidia's bleak outcome for CUDA and PhysX.
And Scali, Idontcare, Senpuu, and forumpanda had some really sweet advice for you. I don't know if you're the type of guys who takes advice, but after all, you are never too old, and never too knowledgeable, to learn something new each and every single day. Now that you've let the dreggs drag you through the muck and entice you into older matters that really don't pertain to this thread, you've lost your edge. Granted, you've been bombarded by countless dudes in this thread, and quite frankly, I'm surprised to retained your composure as long as you did before sinking to that level. One of my favorite quotes from a Darth Vader poster, "You can only call a guy "Annie" so many times before he snaps!!"
And Scali, Idontcare, Senpuu, and forumpanda had some really sweet advice for you. I don't know if you're the type of guys who takes advice, but after all, you are never too old, and never too knowledgeable, to learn something new each and every single day. Now that you've let the dreggs drag you through the muck and entice you into older matters that really don't pertain to this thread, you've lost your edge. Granted, you've been bombarded by countless dudes in this thread, and quite frankly, I'm surprised to retained your composure as long as you did before sinking to that level. One of my favorite quotes from a Darth Vader poster, "You can only call a guy "Annie" so many times before he snaps!!"
Keysplayr
Elite Member
- Jan 16, 2003
- 21,211
- 50
- 91
Yes, relax is good. I'm just trying to lighten the dismal mood in here.
Hey, I can moon everybody if it would help!!
Say Cheese
Hey, I can moon everybody if it would help!!
Say Cheese
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter