Monday, December 12th 2016
AMD "Vega" Demoed in Sonoma, California
AMD's next-generation high-end graphics card, based on the "Vega" architecture, was showcased at an event in Sonoma CA, earlier this week. While the architecture is being debuted with the Radeon Instinct MI25 deep-learning accelerator, a prototype graphics card based on the silicon was exhibited by the company, showing Vulkan API gaming.
AMD was pretty tight-lipped about the specifications of this prototype, but two details appear to have slipped out. Apparently, the chip has a floating point performance of 25 TFLOP/s (FP16), and 12.5 TFLOP/s (FP32, single-precision). On paper, this is higher than the 11 TFLOP/s (FP32) of NVIDIA TITAN X Pascal. The other important specification that emerged is that the card features 8 GB of HBM2 memory, with a memory bandwidth of 512 GB/s. This, too, is higher than the 480 GB/s of the TITAN X Pascal. It remains to be seen which market-segment AMD targets with this card.
This article was updated on Dec 15 to accommodate AMD's request to remove all info regarding the demo system, the shown game and its performance, which has been put under NDA retroactively.
Source:
Golem.de
AMD was pretty tight-lipped about the specifications of this prototype, but two details appear to have slipped out. Apparently, the chip has a floating point performance of 25 TFLOP/s (FP16), and 12.5 TFLOP/s (FP32, single-precision). On paper, this is higher than the 11 TFLOP/s (FP32) of NVIDIA TITAN X Pascal. The other important specification that emerged is that the card features 8 GB of HBM2 memory, with a memory bandwidth of 512 GB/s. This, too, is higher than the 480 GB/s of the TITAN X Pascal. It remains to be seen which market-segment AMD targets with this card.
This article was updated on Dec 15 to accommodate AMD's request to remove all info regarding the demo system, the shown game and its performance, which has been put under NDA retroactively.
120 Comments on AMD "Vega" Demoed in Sonoma, California
it still only manages to best the 1070/1080 by 11 and 8 fps respectively and uses god knows how much power doing so
this does not bode well just more pr non-sense from amd
its using 8GB HMB2 amd would not put that on anything but a top end part the 25TFP/s number also points to that
the 1070 was getting about 50
the big thing is power consumption unless amd has worked some serious magic is probably going to be well north of 225W
the only metrics you should be caring about are the ROP's/TMU's shader count and FP math are mostly second banana when it comes to running games
but hey it looks really good on paper if you have no idea what any of that means
Anyway, this preliminary result is as expected. There's no small or big Vega yet. The only thing ever spotted was a 4096 shader Vega, which is likely to be this very card. Wow, you must have fallen from the hype train before posting this.
Good times.
www.techpowerup.com/reviews/MSI/GTX_1070_Quick_Silver_OC/13.html
I expect 1070 performance in dx11 and 1080+ for vulkan/dx12 for small Vega and a Titan competitor for the big Vega.
For anyone that actually has doom, it doesn't have a benchmark tool.
This means review site numbers are not comparable directly, and AMD can cater this 'sample' to whatever they want.
Because here's my 980Ti @ 4k ultra:
i.imgur.com/diFvE7i.jpg
I don't believe any 'astonishing' numbers until it's in at least w1zz's hands.
Bonus Vulkan screenie:
i.imgur.com/xMEkWxe.jpg
Any moderator that could merge my 3 posts, please do it and sorry for not doing that correctly from the beginning.
Hexus do use Vulkan API for their testing.