NVIDIA's Next-Generation "Ampere" GPUs Could Have 18 TeraFLOPs of Compute Performance

[AleksandarK]

NVIDIA will soon launch its next-generation lineup of graphics cards based on a new and improved "Ampere" architecture. With the first Tesla server cards that are a part of the Ampere lineup going inside Indiana University Big Red 200 supercomputer, we now have some potential specifications and information about its compute performance. Thanks to the Twitter user dylan552p(@dylan522p), who did some math about the potential compute performance of the Ampere GPUs based on NextPlatform's report, we discovered that Ampere is potentially going to feature up to 18 TeraFLOPs of FP64 compute performance.

With Big Red 200 supercomputer being based on Cray's Shasta supercomputer building block, it is being deployed in two phases. The first phase is the deployment of 672 dual-socket nodes powered by AMD's EPYC 7742 "Rome" processors. These CPUs provide 3.15 PetaFLOPs of combined FP64 performance. With a total of 8 PetaFLOPs planned to be achieved by the Big Red 200, that leaves just a bit under 5 PetaFLOPs to be had using GPU+CPU enabled system. Considering the configuration of a node that contains one next-generation AMD "Milan" 64 core CPU, and four of NVIDIA's "Ampere" GPUs alongside it. If we take for a fact that Milan boosts FP64 performance by 25% compared to Rome, then the math shows that the 256 GPUs that will be delivered in the second phase of Big Red 200 deployment will feature up to 18 TeraFLOPs of FP64 compute performance. Even if "Milan" doubles the FP64 compute power of "Rome", there will be around 17.6 TeraFLOPs of FP64 performance for the GPU.


View at TechPowerUp Main Site

 

[ratirt]

With NV I'm not worried about the performance but mostly about the price. I only hope AMD delivers as well or otherwise we will have these cards for crapload of money. That would be a disaster.

 

[Space Lynx]

Final Fantasy X Summoning!!! Ultima Bahamut with Yojimbo RTX cores!!!!

I have a feeling these rumors are true, mainly cause they are afraid of Big Navi.

Glorious time for gamers! Rejoice brothers! Glory is here!

 

[ratirt]

Final Fantasy X Summoning!!! Ultima Bahamut with Yojimbo RTX cores!!!!

I have a feeling these rumors are true, mainly cause they are afraid of Big Navi.

Glorious time for gamers! Rejoice brothers! Glory is here!

Hope you are not mistaken. I like your charisma bro. I really hope the big navi will give some sort of stimulus to NV's RTX cards.

 

[cucker tarlson]

I have a feeling these rumors are true, mainly cause they are afraid of Big Navi.

Just like they have been afraid of everything amd has done for five years that ultimately allowed them to sell their xx106 dies at $500

 

[Chomiq]

One thing's for sure, they'll cost more. At least in high end, since that's where AMD struggles to compete with Nvidia.

 

[Anymal]

One thing's for sure, they'll cost more. At least in high end, since that's where AMD struggles to compete with Nvidia.

Turing will be discounted though. 2070super for 350eur, njami!

 

[T1beriu]

AleksandarK , please ask your editor to stop you from taking napkin math and turning it into news.

 

[dj-electric]

This new architecture might be incredibly fast, but i can't even begin to imagine the pricing on products...

 

[cucker tarlson]

One thing's for sure, they'll cost more. At least in high end, since that's where AMD struggles to compete with Nvidia.

Yup.expect reasonably priced 9-10 tflop cards and big premiums on faster ones.
I dont think were getting 2080ti performance cheaper than $699

 

[ppn]

They need at least 8192 Cuda at 2.1Ghz 1:2 FP64 to do 17.2 Tflops and if we look at NAvi 7nm is only 60% denser that 12nm. So it is similarly sized as Volta but 2.6 faster.

 

[Hyderz]

i dont think nvidia is going to change the price on their top tier cards.
Lets say big navi comes out and the performance of it is better than the rtx 2080ti.
Nvidia will drop the price of the 2080ti to match the big navi price
then when the ampere card launches later on, nvidia will jack up the top tier card back to $999 saying its faster to justify the cost

 

[cucker tarlson]

i dont think nvidia is going to change the price on their top tier cards.
Lets say big navi comes out and the performance of it is better than the rtx 2080ti.
Nvidia will drop the price of the 2080ti to match the big navi price
then when the ampere card launches later on, nvidia will jack up the top tier card back to $999 saying its faster to justify the cost

There will be no big navi.
They cant even get the small navi to work properly.
Ampere will compete with turing and next gen consoles (and their desktop equivalents).

 

[ppn]

EUV will be limited to 429 mm2. AMD card will draw twice the power. So that 429mm2 draws 429 watts, and in case of nvidia 215 watts for the same performance. So there you have it.

 

[Vya Domus]

I have my doubts, if they maintain the 1:2 ratio it would take a chip close to the current 16nm reticle limit running in excess of 2.1 Ghz to achieve anywhere close to 18 TFLOPS. Not only that it would be incredibly expensive and probably have abysmal yields but the power will be insane. I just don't can't see them clocking a Tesla part that high, it's not feasible.

1:1 ratio is an option but it would come with massive die space cost which again, I don't see them doing it.

Something else is going on.

 

[_Flare]

I think a doubling of the TPC in every GPC is likely, going from Turings 6 in the "big GPC" to 12. (Turing small GPC are 4 TPC)
So every "big GPC" will grow to 1536 Cores, so with 6 GPC we reach 9216 Cores.
At 1.9 GHz we reach 17.51 TFlops FP64.

Nvidia did the big(6 TPC) and small (4 TPC) GPC experiment the first time with Turing, i bet that was for a reason.
Turing was also the first time that a non-top-tier chip like the RTX 2080 got 6 GPC when the top-tier has also 6 GPC.
Going over 6 GPC would likely drive the chip complexity without any significant gains, but putting more horsepower in every GPC makes sense.
Maybe Ampere brings even more variance in TPC count per GPC.
12 TPC = 1536 Cores
10 TPC = 1280
_8 TPC = 1024
_6 TPC = _768

at 6 GPC that would bring 9216 or 7680 or 6144 or 4608 Cores (TU102)

 

[ratirt]

There will be no big navi.
They cant even get the small navi to work properly.
Ampere will compete with turing and next gen consoles (and their desktop equivalents).

I'm sure about High-end NAVI (RDNA) not coming but RDNA2 will hit the market this year. Lisa Su has already confirmed that the RDNA2 high-end graphics release in 2020 and CES will have a keynote about it and this will also have Ray tracing support. If this RDNA2 will compete with NV graphics is unknown but also we don't know much about new NV release. I read that the high-end RDNA2 will flood markets with outstanding 4k performance. If that is going to happen only time will tell.

 

[AnarchoPrimitiv]

AleksandarK , please ask your editor to stop you from taking napkin math and turning it into news.

Agreed

 

[cucker tarlson]

I read that the high-end RDNA2 will flood markets with outstanding 4k performance. If that is going to happen only time will tell.

Outstanding,flood - where did you get that?
Nvidia calls anything tu104 based their high end and AMD seems happy to follow so there is no useful information in what you cited.

 

[ratirt]

Outstanding,flood - where did you get that?
Nvidia calls anything tu104 based their high end and AMD seems happy to follow so there is no useful information in what you cited.

AMD might detail new RDNA2 graphics card on March 5

On Financial Analysts Day AMD would share the first details about the new graphics cards that AMD has in the works. According to Mithun Chandrasekhar, the man behind Radeon product management, 'Big N...

www.guru3d.com

flood the market meaning disrupting the GPU market.
As you read, "change the 4k gaming as ryzen did for CPUs", in my words outstanding. What other conclusion would you have in mind?
It's just how I get it. Will it be true? How could you or I tell?

 

[cucker tarlson]

flood the market meaning disrupting the GPU market.
As you read, "change the 4k gaming as ryzen did for CPUs", in my words outstanding. What other conclusion would you have in mind?
It's just how I get it. Will it be true? How could you or I tell?

What is this gibberish?
Is there a card coming or not? How does it perform?

 

[ratirt]

What is this gibberish?
Is there a card coming or not? How does it perform?

Is there a card coming? Yes
How will it perform? Hasn't been revealed yet. CES 2020 hold your horses.

 

[cucker tarlson]

Is there a card coming? Yes
How will it perform? Hasn't been revealed yet. CES 2020 hold your horses.

When is the card coming ? Q1 or Q4?

 

[ratirt]

When is the card coming ? Q1 or Q4?

latest info says it will be out 2020. Which quarter? nobody knows.

 

[Super XP]

Turing will be discounted though. 2070super for 350eur, njami!

Still overpriced lol

A cute little Cardinal birdie told me that Nvidia will not release Ampere until AMD releases RDNA2. And another cute little birdie, this time the birdie was a Blue Jay told me AMD won't release RDNA2 until Nvidia releases Ampere.

Which cute little birdie will you believe?

When is the card coming ? Q1 or Q4?

Q1 Launch Ampere Server GPUs.
Q2 Release Ampere Server GPUs & Launch Ampere Desktop GPUs.
Q3-Q4 Release Ampere Desktop GPUs.
Or listen to the birdies.