nVidia GT300's Fermi architecture unveiled: 512 cores, up to 6GB GDDR5

[HurleyBird]

Originally posted by: KeysplayrSo, Fermi cannot run C++ natively on chip? Ok, I don't actually know what the difference between running on a chip or not is. Some are saying it needs specific hardware for this to actually happen. Could it be that what was meant by running C++ natively on the GPU is actually code created directly with C++ and not any other third party compilers thereafter?
I've heard that Fermi can now be directly programmed on, much the same as a CPU can.
And I think C++ can be used. Visual Basic, Fortran, others.
This isn't my area to be honest. So what I'm typing here are more or less questions for those who understand this type of programming.

So, don't throw any molotov's my way, I'm just asking some questions.

Basically, NVIDIA is hyping up a new C++ compiler/assembler as if it's some kind revolutionary advance in the chip itself, stating that it runs C++ "natively". This is just marketing speak. You could make a C++ compiler and/or assembler for G80, R600, or any other programmable graphics chip. Heck, you could theoretically make a C++ compiler for a Geforce 3, although anything you ran on it would probably be very slow!

C++ has absolutely nothing to do with hardware. It's just a language that's a half step in between a computer language and a human language to make things easier to program. The compiler turns C++ language into assembly language, which is (with some exceptions) simply machine code where groups of 1's and 0's are instead represented by words and base 10 numbers. The assembler turns this into machine code, and it's more or less a direct translation. Some compilers also take out the middle man and do the assembling stage as well.

So no, NVIDIA's hype about Fermi doing C++ natively is a little bit misleading, since everything they've done is on the software side. All they have done is make a compiler/assembler that takes C++ and turns it into something Fermi can understand. There's nothing in the RV870 architecture that would stop ATI from being able to do the same.

 

[Idontcare]

Originally posted by: HurleyBird

Originally posted by: KeysplayrSo, Fermi cannot run C++ natively on chip? Ok, I don't actually know what the difference between running on a chip or not is. Some are saying it needs specific hardware for this to actually happen. Could it be that what was meant by running C++ natively on the GPU is actually code created directly with C++ and not any other third party compilers thereafter?
I've heard that Fermi can now be directly programmed on, much the same as a CPU can.
And I think C++ can be used. Visual Basic, Fortran, others.
This isn't my area to be honest. So what I'm typing here are more or less questions for those who understand this type of programming.

So, don't throw any molotov's my way, I'm just asking some questions.

Basically, NVIDIA is hyping up a new C++ compiler/assembler as if it's some kind revolutionary advance in the chip itself, stating that it runs C++ "natively". This is just marketing speak. You could make a C++ compiler and/or assembler for G80, R600, or any other programmable graphics chip. Heck, you could theoretically make a C++ compiler for a Geforce 3, although anything you ran on it would probably be very slow!

C++ has absolutely nothing to do with hardware. It's just a language that's a half step in between a computer language and a human language to make things easier to program. The compiler turns C++ language into assembly language, which is (with some exceptions) simply machine code where groups of 1's and 0's are instead represented by words and base 10 numbers. The assembler turns this into machine code, and it's more or less a direct translation. Some compilers also take out the middle man and do the assembling stage as well.

So no, NVIDIA's hype about Fermi doing C++ natively is a little bit misleading, since everything they've done is on the software side. All they have done is make a compiler/assembler that takes C++ and turns it into something Fermi can understand. There's nothing in the RV870 architecture that would stop ATI from being able to do the same.

I don't want to become "the guy" who ends up debating every other poster regarding what Fermi is and is not, but again since I've got the whitepaper open here on my computer I'm just going to quote some of it where it seems to suggest there is more to it on the hardware side than this just simply being nothing more than software:

Fermi is the first architecture to support the new Parallel Thread eXecution (PTX) 2.0 instruction set. PTX is a low level virtual machine and ISA designed to support the operations of a parallel thread processor. At program install time, PTX instructions are translated to machine instructions by the GPU driver.

PTX 2.0 was specifically designed to provide full support for the C++ programming language.

Fermi and the PTX 2.0 ISA implement a unified address space that unifies the three separate address spaces (thread private local, block shared, and global) for load and store operations. In PTX 1.0, load/store instructions were specific to one of the three address spaces; programs could load or store values in a specific target address space known at compile time. It was difficult to fully implement C and C++ pointers since a pointer?s target address space may not be known at compile time, and may only be determined dynamically at run time.

With PTX 2.0, a unified address space unifies all three address spaces into a single, continuous address space. A single set of unified load/store instructions operate on this address space, augmenting the three separate sets of load/store instructions for local, shared and global. The 40-bit unified address space supports a Terabyte of addressable memory, and the load/store ISA supports 64-bit addressing for future growth.

Conclusion

The importance of data locality is recognized through Fermi?s two level cache hierarchy and its combined load/store memory path. Double precision performance is elevated to supercomputing levels, while atomic operations execute up to twenty times faster. Lastly, Fermi?s comprehensive ECC support strongly demonstrates our commitment to the high-performance computing market.

On the software side, the architecture brings forward support for C++, the world?s most ubiquitous object-orientated programming language, and Nexus, the world?s first integrated development environment designed for massively parallel GPU computing applications.

Now it seems to me that whoever put together the whitepaper does know the differences between software and hardware, and what I see are people attacking Nvidia's claims when the claims being refuted actually aren't Nvidia's claims to begin with.

 

[MODEL3]

According to Nvidia (From Fermi's whitepaper):

The implementation of a unified address space enables Fermi to support true C++ programs. In C++, all variables and functions reside in objects which are passed via pointers. PTX 2.0 makes it possible to use unified pointers to pass objects in any memory space, and Fermi?s hardware address translation unit automatically maps pointer references to the correct memory space. Fermi and the PTX 2.0 ISA also add support for C++ virtual functions, function pointers, and new and delete operators for dynamic object allocation and de-allocation. C++ exception handling operations ?try? and ?catch? are also supported.

 

[Totally]

Ale those pictures of those cards and not a single spec or benchmark?

The 4850x2 was not an official card (as far as I remember) but made by aftermarket companies right? I google the 5850x2 and didn't find anything official or official looking?

took like 30 seconds

http://lmgtfy.com/?q=5850+x2

the first link if you need further hints

 

[BFG10K]

Originally posted by: BenSkywalker

PCGA is a bunch of liars. There is no splitting hairs here, they lie, badly. They don't have any basis in reality at all, no marketing data to support their obnoxiously obscene claims and nothing resembling a shred of credibility.

I?m afraid you?re going to have to provide solid evidence of this otherwise it?s an opinion that I disagree with. Again, other industry members and analysts like EA?s CEO are claiming the exact the same things PCGA is. Is EA?s CEO lying too when he states PC gaming is growing overall?

By contrast, I linked NPD- which is able to demand thousands of dollars per annual subscription, and corporate financial reports that are bound by law to be accurate.

Your NPD figures don?t even account for digital transactions according to your own link. Not to mention that many such reports only track North American sales. PCGA has stated China is now the biggest PC gaming market so it?s pretty significant if NPD isn?t tracking it, wouldn?t you say?

They are reported in the financials I linked. My numbers are correct, the NPD number was just to use as a guideline of retail sales versus actual total sales numbers. Outside of MMOs, online sales are significantly smaller then most people seem to think. Again, they are listed, as a matter of criminal law, accurately in the financial links I provided.

Yes, your links are correct as far as the data they report. The problem is that they don?t report all of the data, and you don?t have all of the links.

Where are the financials for Valve software for example? How about Zenimax, a private company which doesn?t even disclose financials like other public companies? How about ID software?s Steam sales which aren?t done through Activision? These are but three examples you?re missing, and there are potentially hundreds (if not thousands) of titles out there that aren?t under the umbrella of a major publisher and aren?t available in B&M stores.

The fact is, the data you provided isn?t a complete picture of PC gaming by any stretch of the imagination. Here?s a summary of where PCGA?s data comes from (from the PDF):

The Horizon?s report encompasses PC Gaming revenue from every region around the world, and from every part of the PC gaming industry compiled including retail, online gaming, digital distribution, and online advertising.

That?s far more comprehensive than the four links your provided.

 

[Cogman]

Originally posted by: Idontcare

Originally posted by: HurleyBird

Originally posted by: KeysplayrSo, Fermi cannot run C++ natively on chip? Ok, I don't actually know what the difference between running on a chip or not is. Some are saying it needs specific hardware for this to actually happen. Could it be that what was meant by running C++ natively on the GPU is actually code created directly with C++ and not any other third party compilers thereafter?
I've heard that Fermi can now be directly programmed on, much the same as a CPU can.
And I think C++ can be used. Visual Basic, Fortran, others.
This isn't my area to be honest. So what I'm typing here are more or less questions for those who understand this type of programming.

So, don't throw any molotov's my way, I'm just asking some questions.

Basically, NVIDIA is hyping up a new C++ compiler/assembler as if it's some kind revolutionary advance in the chip itself, stating that it runs C++ "natively". This is just marketing speak. You could make a C++ compiler and/or assembler for G80, R600, or any other programmable graphics chip. Heck, you could theoretically make a C++ compiler for a Geforce 3, although anything you ran on it would probably be very slow!

C++ has absolutely nothing to do with hardware. It's just a language that's a half step in between a computer language and a human language to make things easier to program. The compiler turns C++ language into assembly language, which is (with some exceptions) simply machine code where groups of 1's and 0's are instead represented by words and base 10 numbers. The assembler turns this into machine code, and it's more or less a direct translation. Some compilers also take out the middle man and do the assembling stage as well.

So no, NVIDIA's hype about Fermi doing C++ natively is a little bit misleading, since everything they've done is on the software side. All they have done is make a compiler/assembler that takes C++ and turns it into something Fermi can understand. There's nothing in the RV870 architecture that would stop ATI from being able to do the same.

I don't want to become "the guy" who ends up debating every other poster regarding what Fermi is and is not, but again since I've got the whitepaper open here on my computer I'm just going to quote some of it where it seems to suggest there is more to it on the hardware side than this just simply being nothing more than software:

Fermi is the first architecture to support the new Parallel Thread eXecution (PTX) 2.0 instruction set. PTX is a low level virtual machine and ISA designed to support the operations of a parallel thread processor. At program install time, PTX instructions are translated to machine instructions by the GPU driver.

PTX 2.0 was specifically designed to provide full support for the C++ programming language.

Fermi and the PTX 2.0 ISA implement a unified address space that unifies the three separate address spaces (thread private local, block shared, and global) for load and store operations. In PTX 1.0, load/store instructions were specific to one of the three address spaces; programs could load or store values in a specific target address space known at compile time. It was difficult to fully implement C and C++ pointers since a pointer?s target address space may not be known at compile time, and may only be determined dynamically at run time.

With PTX 2.0, a unified address space unifies all three address spaces into a single, continuous address space. A single set of unified load/store instructions operate on this address space, augmenting the three separate sets of load/store instructions for local, shared and global. The 40-bit unified address space supports a Terabyte of addressable memory, and the load/store ISA supports 64-bit addressing for future growth.

Conclusion

The importance of data locality is recognized through Fermi?s two level cache hierarchy and its combined load/store memory path. Double precision performance is elevated to supercomputing levels, while atomic operations execute up to twenty times faster. Lastly, Fermi?s comprehensive ECC support strongly demonstrates our commitment to the high-performance computing market.

On the software side, the architecture brings forward support for C++, the world?s most ubiquitous object-orientated programming language, and Nexus, the world?s first integrated development environment designed for massively parallel GPU computing applications.

Now it seems to me that whoever put together the whitepaper does know the differences between software and hardware, and what I see are people attacking Nvidia's claims when the claims being refuted actually aren't Nvidia's claims to begin with.

Thanks for posting that, it really does clear up what nVidia's claim is (Not the claim of this reviewer).

I haven't read the whitepaper, I was only going off of the comment made by the website in the OP. Taken out of context, it sound ludicrous. However, put in the context of "Now fermi can handle pointers" It really means a lot. For C++ it is an absolutely needed feature, not so much for C.

Thank IDC.

 

[BenSkywalker]

I?m afraid you?re going to have to provide solid evidence of this otherwise it?s an opinion that I disagree with.

I have provided the financial reports to the six largest PC publishers in the world, you can not provide a single factual number at all(neither can your sources as they flat out lie, they are a marketing group).

. Again, other industry members and analysts like EA?s CEO are claiming the exact the same things PCGA is. Is EA?s CEO lying too when he states PC gaming is growing overall?

MMO revenue is increasing markedly, everything else on the PC is in free fall as far as revenue is concerned. The numbers I provided display this rather clearly. All of the companies that seperate them out are showing huge growth in the MMO space, massice contraction everywhere else.

Where are the financials for Valve software for example? How about Zenimax, a private company which doesn?t even disclose financials like other public companies? How about ID software?s Steam sales which aren?t done through Activision? These are but three examples you?re missing, and there are potentially hundreds (if not thousands) of titles out there that aren?t under the umbrella of a major publisher and aren?t available in B&M stores.

You think Valve and Zenimax are larger then EA and Activision by a staggering amount? Talking about id is redundant, they are a subsidiary of Zenimax and not a private company. Another note about id, they aren't publishing through Activision anymore, the last game that they will publish before being solely Zenimax is published by EA.

What you are claiming at this point, is that a couple of indie publishers and fully indie developers are more then four times the size of every major PC publisher in the world if you honestly believe the numbers you are quoting.

That?s far more comprehensive than the four links your provided.

No, it isn't in the slightest. They are "guessing" for marketing purposes, nothing more. Go ahead and check for their sources on their, heh, 'study'. You won't find it, although you can easily check NPD or check with the SEC about the numbers I've linked. They are a bunch of advertisers who lie throught their teeth, real numbers show them for what they are. Unless you think Valve and Zenimax both did $5Billion dollars each in revenue(which if you did, not much to talk about on this subject) while EA failed to hit $0.8Billion then you must realize their numbers are shockingly wrong. This isn't like they are tweaking things a little bit, they aren't remotely in the league of being accurate.

 

[Keysplayr]

Originally posted by: Cogman


Thanks for posting that, it really does clear up what nVidia's claim is (Not the claim of this reviewer).

I haven't read the whitepaper, I was only going off of the comment made by the website in the OP. Taken out of context, it sound ludicrous. However, put in the context of "Now fermi can handle pointers" It really means a lot. For C++ it is an absolutely needed feature, not so much for C.

Thank IDC.

Now to put it in english for the rest of us non-programmers, does this mean that a developer can sit down at a Fermi equipped rig, and start up visual studio and begin programming right on the GPU using C++ ? Instead of having to write using the CPU and compile for Fermi afterward? Can someone clear this up a little for us?

Secondly, how does this make things easier/faster (if at all) for the dev as opposed to coding for current GPUs? Nvidia or AMD?

Thanks

 

[rpglord]

BenSkywalker you need to get your head examined if you think those 6 companies make 90 % of the games sold for pc platform

 

[novasatori]

I am excited for this chip, I was originally excited for 5800 series, and kinda meh on the NV part expecting them to be similar, but boy was I wrong.

Just saw a translated link (assuming it was correct) that mentioned $399 for the GTX, if true, I am REALLY excited.

 

[Cogman]

Originally posted by: Keysplayr
Now to put it in english for the rest of us non-programmers, does this mean that a developer can sit down at a Fermi equipped rig, and start up visual studio and begin programming right on the GPU using C++ ? Instead of having to write using the CPU and compile for Fermi afterward? Can someone clear this up a little for us?

Secondly, how does this make things easier/faster (if at all) for the dev as opposed to coding for current GPUs? Nvidia or AMD?

Thanks

Here's what it means, code still has to be compiled for the fermi platform. What has been implemented into the fermi platform that wasn't in previous platforms is a uniform way to access memory.

For c++, what makes c++, C++, is basically what it does with memory. OOP requires a lot of memory referencing and dereferencing (IE looking up a memory address and then getting the value for it, or getting the memory address for a given value).

Basically, without this feature you are unable to have high level languages. So now fermi does support the use of high level languages because it has all the instructions needed to perform the memory tasks that are so crutial to a high level language (By high level, I mean a language that far from machine language, or that is closer to human language)

So yeah, the ease comes in the fact that now you can use proper OOP on Fermi, which is just a faster and often clearer way to develop code. (most major programs are written with an OOP paradigm in mind).

One thing that I'm not entirely sure on is if this means that a pointer (memory reference) passed from the CPU will be valid on the GPU (and visa versa). I doubt it, but it would be cool if it were.

 

[Keysplayr]

Originally posted by: rpglord
BenSkywalker you need to get your head examined if you think those 6 companies make 90 % of the games sold for pc platform

Then you need to post something that convinces us why you think so and that you're right, and forget about BenSkywalkers head.
Just saying he needs his head examined doesn't do a thing to support your argument. Show us.

 

[Keysplayr]

Originally posted by: Cogman

Originally posted by: Keysplayr
Now to put it in english for the rest of us non-programmers, does this mean that a developer can sit down at a Fermi equipped rig, and start up visual studio and begin programming right on the GPU using C++ ? Instead of having to write using the CPU and compile for Fermi afterward? Can someone clear this up a little for us?

Secondly, how does this make things easier/faster (if at all) for the dev as opposed to coding for current GPUs? Nvidia or AMD?

Thanks

Here's what it means, code still has to be compiled for the fermi platform. What has been implemented into the fermi platform that wasn't in previous platforms is a uniform way to access memory.

For c++, what makes c++, C++, is basically what it does with memory. OOP requires a lot of memory referencing and dereferencing (IE looking up a memory address and then getting the value for it, or getting the memory address for a given value).

Basically, without this feature you are unable to have high level languages. So now fermi does support the use of high level languages because it has all the instructions needed to perform the memory tasks that are so crutial to a high level language (By high level, I mean a language that far from machine language, or that is closer to human language)

So yeah, the ease comes in the fact that now you can use proper OOP on Fermi, which is just a faster and often clearer way to develop code. (most major programs are written with an OOP paradigm in mind).

One thing that I'm not entirely sure on is if this means that a pointer (memory reference) passed from the CPU will be valid on the GPU (and visa versa). I doubt it, but it would be cool if it were.

Thanks!
Aren't the pointers part of an ISA standard? Meaning if the hardware is to be compliant, bit it CPU or GPU, it needs to meet a standard? Sorry for these seemingly dumb questions, just trying to better understand it.

 

[Cogman]

Originally posted by: Keysplayr
Thanks!
Aren't the pointers part of an ISA standard? Meaning if the hardware is to be compliant, bit it CPU or GPU, it needs to meet a standard? Sorry for these seemingly dumb questions, just trying to better understand it.

nVidia's ISA is different from intel's ISA (x86). They can do whatever they want with pointers so long as they are consistent.

Previously, they had 3 different types of pointers, with fermi there is now one sort of dynamic pointer that will translate into one of the 3 other pointers at run time. The more I read it, the less it sounds like the pointers are for system memory, and the more it sounds like the pointers are for the local video card memory.

 

[Keysplayr]

Ah, ok.
If the pointers are to local video card memory, it kind of makes more sense. Why have ECC memory and not use it. I would think so anyway. ??

 

[rpglord]

Originally posted by: Keysplayr

Originally posted by: rpglord
BenSkywalker you need to get your head examined if you think those 6 companies make 90 % of the games sold for pc platform

Then you need to post something that convinces us why you think so and that you're right, and forget about BenSkywalkers head.
Just saying he needs his head examined doesn't do a thing to support your argument. Show us.

Ok,
World of warcraft made and published by firm not on his list ( Blizzard )
Fallout 3 published by Bethesda
Armed Assault 2 ( Got Game Entertaiment and 505 games )
Risen (Deep Silver )
Neverwinter nights 2 and it's numerous expansions ( Atari )
Sacred 2 ( Deep Silver )
Cryostasis ( Aspyr )
Bioshock ( its little older but bioshock 2 is coming soon with same publisher ) 2k games
Resident Evil 5 ( capcom )
Street Fighter IV ( capcom )
Fear 2 ( warner bros )
A Vampyre Story ( Crimson Cow )
GTA IV ( RockStar )
Should I just continue to list recent games which are published by a company not on his list ? Because its getting ridiculous. Those 6 companies are big,but to say they own 90 % of the market ? Thats just crazy...maybe 30-40 % nowhere close to 90 %

 

[Zstream]

Originally posted by: rpglord

Originally posted by: Keysplayr

Originally posted by: rpglord
BenSkywalker you need to get your head examined if you think those 6 companies make 90 % of the games sold for pc platform

Then you need to post something that convinces us why you think so and that you're right, and forget about BenSkywalkers head.
Just saying he needs his head examined doesn't do a thing to support your argument. Show us.

Ok,
World of warcraft made and published by firm not on his list ( Blizzard )
Fallout 3 published by Bethesda
Armed Assault 2 ( Got Game Entertaiment and 505 games )
Risen (Deep Silver )
Neverwinter nights 2 and it's numerous expansions ( Atari )
Sacred 2 ( Deep Silver )
Cryostasis ( Aspyr )
Bioshock ( its little older but bioshock 2 is coming soon with same publisher ) 2k games
Resident Evil 5 ( capcom )
Street Fighter IV ( capcom )
Fear 2 ( warner bros )
A Vampyre Story ( Crimson Cow )
GTA IV ( RockStar )
Should I just continue to list recent games which are published by a company not on his list ? Because its getting ridiculous. Those 6 companies are big,but to say they own 90 % of the market ? Thats just crazy...maybe 30-40 % nowhere close to 90 %

Don't bother with it, someone could say Nvidia/Ati walks on water and they would still ask you to disprove it with data...

Just because you use a blinker, does not give you the right to act stupid.

 

[Zstream]

Originally posted by: Cogman

Originally posted by: Keysplayr
Now to put it in english for the rest of us non-programmers, does this mean that a developer can sit down at a Fermi equipped rig, and start up visual studio and begin programming right on the GPU using C++ ? Instead of having to write using the CPU and compile for Fermi afterward? Can someone clear this up a little for us?

Secondly, how does this make things easier/faster (if at all) for the dev as opposed to coding for current GPUs? Nvidia or AMD?

Thanks

Here's what it means, code still has to be compiled for the fermi platform. What has been implemented into the fermi platform that wasn't in previous platforms is a uniform way to access memory.

For c++, what makes c++, C++, is basically what it does with memory. OOP requires a lot of memory referencing and dereferencing (IE looking up a memory address and then getting the value for it, or getting the memory address for a given value).

Basically, without this feature you are unable to have high level languages. So now fermi does support the use of high level languages because it has all the instructions needed to perform the memory tasks that are so crutial to a high level language (By high level, I mean a language that far from machine language, or that is closer to human language)

So yeah, the ease comes in the fact that now you can use proper OOP on Fermi, which is just a faster and often clearer way to develop code. (most major programs are written with an OOP paradigm in mind).

One thing that I'm not entirely sure on is if this means that a pointer (memory reference) passed from the CPU will be valid on the GPU (and visa versa). I doubt it, but it would be cool if it were.

I did not get that from the report. It seems that the code can be dynamically written to the address space per request. All it is doing is improving the amount of data and the variables allowed to use the space.

What guarantees the space will be available and have access by hardware and app?
Does the GPU support ram from being overwritten? I was unaware that any GPU has this support. ATI just now started to put hardware CRC into the memory.

The overall story is that the card can not run an OS by itself, it will need an OS. Until you see a video card with a certain chunk of the PCB having flash storage I would not expect to see it running an app or OS.

 

[Nemesis 1]

Originally posted by: MODEL3
According to Nvidia (From Fermi's whitepaper):

The implementation of a unified address space enables Fermi to support true C++ programs. In C++, all variables and functions reside in objects which are passed via pointers. PTX 2.0 makes it possible to use unified pointers to pass objects in any memory space, and Fermi?s hardware address translation unit automatically maps pointer references to the correct memory space. Fermi and the PTX 2.0 ISA also add support for C++ virtual functions, function pointers, and new and delete operators for dynamic object allocation and de-allocation. C++ exception handling operations ?try? and ?catch? are also supported.

It would be nice if someone would be kind enough to explain what the bolded part above actually means.

 

[Keysplayr]

Originally posted by: rpglord

Originally posted by: Keysplayr

Originally posted by: rpglord
BenSkywalker you need to get your head examined if you think those 6 companies make 90 % of the games sold for pc platform

Then you need to post something that convinces us why you think so and that you're right, and forget about BenSkywalkers head.
Just saying he needs his head examined doesn't do a thing to support your argument. Show us.

Ok,
World of warcraft made and published by firm not on his list ( Blizzard )
Fallout 3 published by Bethesda
Armed Assault 2 ( Got Game Entertaiment and 505 games )
Risen (Deep Silver )
Neverwinter nights 2 and it's numerous expansions ( Atari )
Sacred 2 ( Deep Silver )
Cryostasis ( Aspyr )
Bioshock ( its little older but bioshock 2 is coming soon with same publisher ) 2k games
Resident Evil 5 ( capcom )
Street Fighter IV ( capcom )
Fear 2 ( warner bros )
A Vampyre Story ( Crimson Cow )
GTA IV ( RockStar )
Should I just continue to list recent games which are published by a company not on his list ? Because its getting ridiculous. Those 6 companies are big,but to say they own 90 % of the market ? Thats just crazy...maybe 30-40 % nowhere close to 90 %

Excellent!! Now show us their SEC filings! You do realize that all you did here was list some games and their developers? That's it? You're pulling % numbers out of thin air. Guessing. Ben actually reviewed miles of pages for his info. You'll have to do much, MUCH better.
Oh and another thing, I believe they were excluding MMO's, which is on the rise.

 

[Modelworks]

Originally posted by: Cogman

One thing that I'm not entirely sure on is if this means that a pointer (memory reference) passed from the CPU will be valid on the GPU (and visa versa). I doubt it, but it would be cool if it were.

If they could actually get that to work it would be a major accomplishment, I'm doubtful as well. Arm did this with the last incarnation of the powerVR chipsets to boost video performance and Nvidia has been working very closely with ARM in the past year so who knows what they have shared in tech.

 

[Keysplayr]

Originally posted by: Zstream

Originally posted by: rpglord

Originally posted by: Keysplayr

Originally posted by: rpglord
BenSkywalker you need to get your head examined if you think those 6 companies make 90 % of the games sold for pc platform

Then you need to post something that convinces us why you think so and that you're right, and forget about BenSkywalkers head.
Just saying he needs his head examined doesn't do a thing to support your argument. Show us.

Ok,
World of warcraft made and published by firm not on his list ( Blizzard )
Fallout 3 published by Bethesda
Armed Assault 2 ( Got Game Entertaiment and 505 games )
Risen (Deep Silver )
Neverwinter nights 2 and it's numerous expansions ( Atari )
Sacred 2 ( Deep Silver )
Cryostasis ( Aspyr )
Bioshock ( its little older but bioshock 2 is coming soon with same publisher ) 2k games
Resident Evil 5 ( capcom )
Street Fighter IV ( capcom )
Fear 2 ( warner bros )
A Vampyre Story ( Crimson Cow )
GTA IV ( RockStar )
Should I just continue to list recent games which are published by a company not on his list ? Because its getting ridiculous. Those 6 companies are big,but to say they own 90 % of the market ? Thats just crazy...maybe 30-40 % nowhere close to 90 %

Don't bother with it, someone could say Nvidia/Ati walks on water and they would still ask you to disprove it with data...

Just because you use a blinker, does not give you the right to act stupid.

Can it, happiness.

 

[rpglord]

Originally posted by: Keysplayr

Originally posted by: rpglord

Originally posted by: Keysplayr

Originally posted by: rpglord
BenSkywalker you need to get your head examined if you think those 6 companies make 90 % of the games sold for pc platform

Then you need to post something that convinces us why you think so and that you're right, and forget about BenSkywalkers head.
Just saying he needs his head examined doesn't do a thing to support your argument. Show us.

Ok,
World of warcraft made and published by firm not on his list ( Blizzard )
Fallout 3 published by Bethesda
Armed Assault 2 ( Got Game Entertaiment and 505 games )
Risen (Deep Silver )
Neverwinter nights 2 and it's numerous expansions ( Atari )
Sacred 2 ( Deep Silver )
Cryostasis ( Aspyr )
Bioshock ( its little older but bioshock 2 is coming soon with same publisher ) 2k games
Resident Evil 5 ( capcom )
Street Fighter IV ( capcom )
Fear 2 ( warner bros )
A Vampyre Story ( Crimson Cow )
GTA IV ( RockStar )
Should I just continue to list recent games which are published by a company not on his list ? Because its getting ridiculous. Those 6 companies are big,but to say they own 90 % of the market ? Thats just crazy...maybe 30-40 % nowhere close to 90 %

Excellent!! Now show us their SEC filings! You do realize that all you did here was list some games and their developers? That's it? You're pulling % numbers out of thin air. Guessing. Ben actually reviewed miles of pages for his info. You'll have to do much, MUCH better.
Oh and another thing, I believe they were excluding MMO's, which is on the rise.

I am not going to do that,because actually I was only commenting on the fact that those 6 companies do not publish 90 % of the games for pc.Not even close. Just listing recent,popular games proved that
I don't know if pc gaming is "dying" or not,and I dont really want to get into that

 

[T2k]

I really want to see some hard evidence of those handful of companies owning 90% of PC market..... what? Sales...? Number of titles or franchises...?

 

[BenSkywalker]

World of warcraft made and published by firm not on his list ( Blizzard )

Blizzard is owned by Activision, I listed them(and even listed WoW revenue seperately).

Neverwinter nights 2 and it's numerous expansions ( Atari )

$30Million for Infrogames which owns Atari.

Bioshock ( its little older but bioshock 2 is coming soon with same publisher ) 2k games

I listed 2K games.

Resident Evil 5 ( capcom )

Capcom reports things a bit differently. According to their numbers they have 0.93% of the US market(console and PC combined).

GTA IV ( RockStar )

RockStar is owned by Take2, which I listed in my original list.

I am not going to do that,because actually I was only commenting on the fact that those 6 companies do not publish 90 % of the games for pc.

We aren't talking about the quantity of titles, we are talking about sales, and 90% is likely a bit conservative on my part. The only two titles on your list that approached 1Million units sold on the PC were GTA and WoW, and both of them were included in my first post listing companies. The other publishers you listed either don't break down their data as much as needed(WB) or are small time indies(Aspyr, Deep Silver and Crimson Cow who is miniscule).