I think you need to add 10 to the GT600. It has 10 billion transistors!! Don't get too carried away with the copy and paste. lol
- Status
I think you need to add 10 to the GT600. It has 10 billion transistors!! Don't get too carried away with the copy and paste. lol
http://vr-zone.com/articles/-rumour-nvidia-gt300-architecture-details-revealed/7763.html?doc=7763
According to this..the codename would be GF100. GPU would have over 3B transistors.
According to this..the codename would be GF100. GPU would have over 3B transistors.
dont know why the vrzone author thinks its 256 shaders. that diagram shows combined 32 scalar fp32 sp alus + interpolators in one MIMD unit with a total of 16 MIMD units under the blurred area.
= 512 shaders. Except madd = mul are redesigned and the interpolators are in the text units now?
= 512 shaders. Except madd = mul are redesigned and the interpolators are in the text units now?
Don't know what the four dots are above the sp's. memory should be 384bit ~ 230gb/sec woot! 512 shaders / 230 gb/sec / 40nm / 3.2bln transistors - sounds fast to me!
I wonder if these rumor sites are tripping all over each other with the barrage of codenames because they haven't figured out (or been told) that there are some four unique IC designs coming, each with their own codename of course?
It's not a rumor. Rys @B3D posted his screenshot blurred out last night: http://forum.beyond3d.com/show...1341596&postcount=2552
It's been known for some time that it would be 512 shaders and DDR5. With a 384-bit memory interface it should perform very nicely indeed. 30% faster than a 295 is a distinct possibility.
http://www.brightsideofnews.co...c-up-to-6gb-gddr5.aspx
Onboard cache, native C++ support, these things will be amazing compute engines (and will probably put larrabee back another year as Intel try to come up with something that doesn't look useless in comparison). It's a folders dream.
On the other hand will all that gpu compute stuff compromise graphics performance vs the 5870, which is a pure graphics chip with few gpu compute dreams?
Onboard cache, native C++ support, these things will be amazing compute engines (and will probably put larrabee back another year as Intel try to come up with something that doesn't look useless in comparison). It's a folders dream.
On the other hand will all that gpu compute stuff compromise graphics performance vs the 5870, which is a pure graphics chip with few gpu compute dreams?
I noticed in Nvidia's whitepaper on Fermi that they repeatedly made mention/comparison to G80 and GT200 but not once is there any reference to a "GT300" nor a "GF100".
It is simply Fermi, nothing more, nothing less, in the whitepaper.
Has Nvidia publicly referred to Fermi as GF100 or GT300 at any point in time? I would think if the terms were actually used and synonymous then they'd be present somewhere at least once in the whitepaper.
It is simply Fermi, nothing more, nothing less, in the whitepaper.
Has Nvidia publicly referred to Fermi as GF100 or GT300 at any point in time? I would think if the terms were actually used and synonymous then they'd be present somewhere at least once in the whitepaper.
Thats a good guess. I've seen people say guesses from anywhere from 15-20% faster than an HD5870 to 60% faster than a GTX295.
Barfo
Lifer
- Jan 4, 2005
- 27,554
- 212
- 106
Hahah :thumbsup:Hopefully this doesn't turn out to be the Prescott of GPUs, don't like that code name at all
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
- Status
TRENDING THREADS
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 10K
-
Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)- Started by Vattila
- Replies: 13K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 7K
-
Facebook
Twitter