Monday, February 3rd 2020

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

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.
Sources: @dylan522p(Twitter), The Next Platform
Add your own comment

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

#51
cucker tarlson
ratirt
And you know it isn't? Cause you were arguing that it will never be released few posts up.
I said the odds are that it will be release cause it has been mentioned by Lisa Su, other executives from AMD and lead Radeon managers. Although you didn't say if you mean RDNA or RDNA2. If AMD says it is coming then it is. AMD's 2020 GPU roadmap says, New Navi (RDNA2) in 2020 but if you say it's not then maybe tell AMD this, cause they don't know that.
no,I said no BIG navi
please read more carefully
Posted on Reply
#52
the54thvoid
This thread's a real slug fest of rumour versus unrealistic expectation. It's the worst of TPU community 'tech talk'. All based on nothing of substantial merit. Whatever Ampere brings at such a huge improvement will unlikely yield such results in gaming, the biggest market TPU members care about. As for big Navi, even more rumour.
Posted on Reply
#53
Steevo
Also, Nvidia typically stumbles with first gen node dies, I get that they are 25% more efficient with power due to the node difference between them and AMD using 7nm. But history has not been kind to Nvidia during transition to new nodes and processes. Maybe they will have this one figured out.
Posted on Reply
#54
net2007
That's a lot of fancy die talk. Comes down to what people are willing to pay. Titan Z $2500 didn't go so well. 2080ti $1200 ($200 greater than normal Titans) didn't go so well. In fact, there wasn't much reason to upgrade over 10 series. Bet they will be more cautious this time around.
Posted on Reply
#55
Minus Infinity
It's obvious from AMD's Lisa Su, Nvidia will no longer have the high-end to themselves with next gen cards. We already know AMD will offer products to compete at the 2080 and higher lever with RDNA2. 2020/2021 will give us some good competition right through the price brackets and maybe with some luck 2021/2022 will see Intel offer something at least in the low to mid-range too. I don't think Nvidia will make the same mistake with the 3xxx series prcing as they did with the 2xxx, the sales were really not great initially and look without Navi we would never have seen the super range released. They already know they now have a fight and it won't get easier with Navi+.
Posted on Reply
#56
efikkan
Everyone, please remember that both segments like "high-end" and the performance levels required for new games in e.g. 4K are moving targets. By the time Nvidia's next generation arrives, the "RTX 2080 performance level" will no longer be regarded as "high-end".
Posted on Reply
#57
Fluffmeister
Yeah also I'm not getting the Nvidia struggle on new nodes either, 16nm Pascal still remains a headache for AMD years later.... aka now.
Posted on Reply
#58
SIGSEGV
At this point, I do really wish that AMD's Radeon division going bankruptcy and disappear.
This is not good at all.
Posted on Reply
#59
renz496
gamefoo21
At least there's someone else who understands this.

Volta is really quite different than the Turing cores.

Vega V20 fights directly with Volta and provides some a lot of competition.

Navi is the compute crippled GeForce/Quadro fighter.

This is the big reason that the APUs use Vega based GFX because for office stuff OpenCL performance is king. V20 also has the same degenerated rendering hardware as Navi. So it can do GFX like Navi but crushes it in compute tasks.

A simple comparison...

Xbox One/S and the PS4 variants use Polaris GPUs and they are in the Fury/Navi line.

The Xbox One X uses a GPU based off of the 290X/Vega branch.

Radeon VII isn't much faster in gaming than a 5700XT, but it's 5.5 times faster in compute. The 5700XT has basically the same compute performance as a 2080Ti.

Edit: I really hope Arcturus brings it, because the sooner CUDA dies out the better. Locking so much important research to a closed system isn't good. OpenCL ftw!
OpenCL exist for a decade now. Back in 2009 they said give it two years. Two years and no one ever remember CUDA ever exist. Part of the issue is CUDA is fully controlled by nvidia. That in one aspect will speed up CUDA development and optimization on nvidia hardware. With OpenCL everyone needs to agree what spec to be accepted into the API and this where the major problem are. Because some feature might work nicely on one hardware but not others. And they will not want this to happen because that will only give certain IHV the comepetitive edge while leave out the others. So while they are disputing how the feature should be added to the spec closed API like CUDA already moving forward to implement next feature and improvement. For professional client they will favor the one that really works for them even if they really like open source. Because in pro world time is money. And second if only certain vendor going to be good with that open source API what the point? Because to get maximize percormance they will still going to be locked on that certian hardware when the point of using open source solution is not to locked themselves on one vendor only.
Posted on Reply
#60
TheGuruStud
SIGSEGV
At this point, I do really wish that AMD's Radeon division going bankruptcy and disappear.
This is not good at all.
Wut? Lay off the Huang juice.
Posted on Reply
#61
renz496
Fluffmeister
Yeah also I'm not getting the Nvidia struggle on new nodes either, 16nm Pascal still remains a headache for AMD years later.... aka now.
This. Nvidia might have issues in the past when moving to new node but after 40nm disaster they put a lot more effort in this issue. 28nm i don't any issue. Same with 16nm. Now they invest even more money to enable themselves to get a node they really want like 12nmFFN where they can reliaze the dream of GV100 with staggering die size than none of us ever think was possible. 7nm has been out for quite some time now. Just because nvidia did not launch 7nm product like AMD meaning their upcoming transition to 8nm will be more problematic because they have much less experience on it. Nvidia most likely have access to the node from much earlier just like others but they just don't need it yet. The development route they take like 12nmFFN will be their "alternative way out" if future node have issues.
Posted on Reply
#62
ratirt
cucker tarlson
no,I said no BIG navi
please read more carefully
I dead read you and I've mentioned I don't know if you are referring to RDNA or RDNA2. You need to be more specific. Besides i's been in the ether for a while that the RDNA aka BIG NAVI will not be released but The "BIG NAVI" RDNA2 will. Just wanted to get that one clear.
Posted on Reply
#63
cucker tarlson
ratirt
I dead read you and I've mentioned I don't know if you are referring to RDNA or RDNA2. You need to be more specific. Besides i's been in the ether for a while that the RDNA aka BIG NAVI will not be released but The "BIG NAVI" RDNA2 will. Just wanted to get that one clear.
both
I'm hoping for rdna2 rt card in h2 but I wouldn't put it past amd to do some sort of stupid 580->590-like refresh with 5700xt+/5700xtx and move the real rdna 2 cards to 2021.
Posted on Reply
#64
kings
Minus Infinity
It's obvious from AMD's Lisa Su, Nvidia will no longer have the high-end to themselves with next gen cards. We already know AMD will offer products to compete at the 2080 and higher lever with RDNA2. 2020/2021 will give us some good competition right through the price brackets and maybe with some luck 2021/2022 will see Intel offer something at least in the low to mid-range too. I don't think Nvidia will make the same mistake with the 3xxx series prcing as they did with the 2xxx, the sales were really not great initially and look without Navi we would never have seen the super range released. They already know they now have a fight and it won't get easier with Navi+.
We will see, the last time AMD said it was going to compete seriously in the high end, we had Vega, 15 months after Pascal, barely matching the GTX 1080.

Not to mention the "overclockers dream" or "poor Volta" thing, etc ... What CEOs say is not reliable and especially at AMD we have had recent cases of reality not matching what is said. Therefore, caution and control over expectations are necessary.
Posted on Reply
#65
cucker tarlson
kings
We will see, the last time AMD said it was going to compete seriously in the high end, we had Vega, 15 months after Pascal, barely matching the GTX 1080.

Not to mention the "overclockers dream" or "poor Volta" thing, etc ... What CEOs say is not reliable and especially at AMD we have had recent cases of reality not matching what is said. Therefore, caution and control over expectations are necessary.
amd proved that even when they have they tools like new,more efficient architecture and a 7nm tsmc node,they still can't deliver in the high end.
all this amd ceo "we'll be big next year" talk that people keep telling me to believe makes me nauseous.
I'll believe it when I see big rdna2 cards this year,if it's gonna be rdna1 refresh to go against two year old turing cards it'll be a miserable failure yet again.
Posted on Reply
#66
ratirt
kings
We will see, the last time AMD said it was going to compete seriously in the high end, we had Vega, 15 months after Pascal, barely matching the GTX 1080.

Not to mention the "overclockers dream" or "poor Volta" thing, etc ... What CEOs say is not reliable and especially at AMD we have had recent cases of reality not matching what is said. Therefore, caution and control over expectations are necessary.
Well, NV is releasing its new GPU first. At this point anything is possible. Even though It has been confirmed the new RDNA2 is coming out this year, we all know this may change. I think AMD is waiting for the new NV GPU (that's a maybe). If they find out the new NV is not that improved over the Turing, maybe they will do something else? It is hard to predict what each company will actually do.
Even if NV and AMD release their cards I will still wait for benchmarks. I'm not in a hurry and I need to see what each can do to make a purchase decision.
Posted on Reply
#67
cucker tarlson
ratirt
I think AMD is waiting for the new NV GPU
what about the old ones ? are they waiting to release a 2080 competitor too?
Posted on Reply
#68
ratirt
cucker tarlson
what about the old ones ? are they waiting to release a 2080 competitor too?
Think about it. If AMD was to release RDNA NAVI to compete with 2080 and up, it would have happened by now. I mean, that is what I think. I don't know why AMD is waiting. Maybe they just want to ditch the RDNA and move on? NV is releasing new graphics in May probably (present the new graphics). What's the point for AMD to release 2080 and up competitor now, when new stuff is just around the corner from NV? So it can compete with it for 3 months tops?
Posted on Reply
#69
cucker tarlson
ratirt
Think about it. If AMD was to release RDNA NAVI to compete with 2080 and up, it would have happened by now. I mean, that is what I think. I don't know why AMD is waiting. Maybe they just want to ditch the RDNA and move on? NV is releasing new graphics in May probably (present the new graphics). What's the point for AMD to release 2080 and up competitor now, when new stuff is just around the corner from NV? So it can compete with it for 3 months tops?
During the earnings call, AMD's CEO, Dr. Lisa Su, once again confirmed new Radeon RX graphics cards in 2020 but she didn't just mention next-generation RDNA based products but also unveiled that existing Navi (RDNA1) graphics cards would be refreshed.
I just feel sorry for those who buy navi refresh in 2020 when consoles will get rdna 2 at the same time.
Posted on Reply
#70
ratirt
cucker tarlson
I just feel sorry for those who buy navi refresh in 2020 when consoles will get rdna 2 at the same time.
Will they get RDNA2? Fully fledged RDNA2? From what I've read so far it says NAVI with Ray tracing. Maybe some derivative from RDNA with some hardware accelerating ray tracing. It does not have to be RDNA2 and I think the official specs haven't been presented yet. Still it is dedicated gpu we are talking about here. Besides, even though NV has introduced RT in the Turing graphics, AMD still stayed conservative about the RT idea. PS5 may have the acceleration, but it has never been said it will be RDNA2.
Also, I think there was always this correlation between consoles and graphics. How I always pictured it, the new stuff was always introduced as a GPU for PC and then it would find it's way to consoles. That is reasonable since consoles as dedicated gaming platforms, don't need that much processing power as PCs do and (despite PS4pro), all graphics options are locked.

Why do you feel sorry? Most of the people purchasing it, know what they've purchased and for what reason. AMD was not aiming for sorry though and what I hear, people are happy with the performance for what they had to pay for it. Performance per $ and value is good in the tier where the card is currently. Also mixing up consoles and PC graphics in the way you do is not right.
I can say, I'm sorry for people who purchased Turing in 2020 since new NV GPU is going to be released soon. I don't think that would have been justified since they are happy with what they've got.
Posted on Reply
#71
cucker tarlson
ratirt
Will they get RDNA2? Fully fledged RDNA2? From what I've read so far it says NAVI with Ray tracing. Maybe some derivative from RDNA with some hardware accelerating ray tracing. It does not have to be RDNA2 and I think the official specs haven't been presented yet. Still it is dedicated gpu we are talking about here. Besides, even though NV has introduced RT in the Turing graphics, AMD still stayed conservative about the RT idea. PS5 may have the acceleration, but it has never been said it will be RDNA2.
Also, I think there was always this correlation between consoles and graphics. How I always pictured it, the new stuff was always introduced as a GPU for PC and then it would find it's way to consoles. That is reasonable since consoles as dedicated gaming platforms, don't need that much processing power as PCs do and (despite PS4pro), all graphics options are locked.

Why do you feel sorry? Most of the people purchasing it, know what they've purchased and for what reason. AMD was not aiming for sorry though and what I hear, people are happy with the performance for what they had to pay for it. Performance per $ and value is good in the tier where the card is currently. Also mixing up consoles and PC graphics in the way you do is not right.
I can say, I'm sorry for people who purchased Turing in 2020 since new NV GPU is going to be released soon. I don't think that would have been justified since they are happy with what they've got.
rdna 2 is rdna with rt and vrs on 7nm+

ratirt
I can say, I'm sorry for people who purchased Turing in 2020
me too
Posted on Reply
#72
ratirt
cucker tarlson
rdna 2 is rdna with rt and vrs on 7nm+


me too
And how can you possibly know that?
Try this for a change.
This Ampere is nothing more than Turing on a 7nm node. Nothing special about it and I wouldn't expect much. Would you agree with my statement?
cucker tarlson
me too
that was a sarcasm but sure, I give you that one :)
Posted on Reply
#73
kings
ratirt
Well, NV is releasing its new GPU first. At this point anything is possible. Even though It has been confirmed the new RDNA2 is coming out this year, we all know this may change. I think AMD is waiting for the new NV GPU (that's a maybe). If they find out the new NV is not that improved over the Turing, maybe they will do something else? It is hard to predict what each company will actually do.
Even if NV and AMD release their cards I will still wait for benchmarks. I'm not in a hurry and I need to see what each can do to make a purchase decision.
But AMD cannot wait, they are the ones who have to gain market share, increase profits, try to change the mindshare, etc ... It is AMD that is playing catch up, so if they have big guns, there is no interest in waiting.

The longer they wait, the more time they give Nvidia to respond and even subject themselves to Nvidia already having something better by then. It's a dangerous game for them, I don't think AMD is doing this on purpose.

ratirt
Think about it. If AMD was to release RDNA NAVI to compete with 2080 and up, it would have happened by now. I mean, that is what I think. I don't know why AMD is waiting. Maybe they just want to ditch the RDNA and move on? NV is releasing new graphics in May probably (present the new graphics). What's the point for AMD to release 2080 and up competitor now, when new stuff is just around the corner from NV? So it can compete with it for 3 months tops?
I agree, for me It´s clear by now that 1st gen Navi will not go beyond this, otherwise they had already launched them. And it´s not because they are waiting, it´s because it is probably not feasible to scale up to the level they need. AMD did not mind launching a Radeon VII and discontinuing it 4 months later, for that reason, I don´t believe that AMD is waiting for something new from Nvidia, they simply have nothing to offer.

Maybe with RDNA 2 things will be much different, we all hope so, but that is also unknown. People tend to think that RDNA 2 is going to be a revolution, but there is no evidence of that, only rumors. It may just be RDNA 1 with minor tweaks and RT/VRS, we don´t know. The same applies to Nvidia's rumors, we don't know if it's just going to be a slight tweak on Turing or something quite different.
Posted on Reply
#74
cucker tarlson
ratirt
And how can you possibly know that?
https://www.techpowerup.com/forums/threads/ray-tracing-and-variable-rate-shading-design-goals-for-amd-rdna2.262098/

you commented like a thousand times in this thread

ratirt
This Ampere is nothing more than Turing on a 7nm node. Nothing special about it and I wouldn't expect much. Would you agree with my statement?
no,cause turing on 7nm would be amazing
pascal was maxwell on a shrunk die and it put amd so far behind that they still can't catch up with 1080Ti in 2020 unless you're counting radeon 7

kings
Maybe with RDNA 2 things will be much different, we all hope so, but that is also unknown. People tend to think that RDNA 2 is going to be a revolution, but there is no evidence of that, only rumors.
one person thinks it and keeps spamming every thread with it
fixed that for you.
Posted on Reply
#75
ratirt
kings
But AMD cannot wait, they are the ones who have to gain market share, increase profits, try to change the mindshare, etc ... It is AMD that is playing catch up, so if they have big guns, there is no interest in waiting.

The longer they wait, the more time they give Nvidia to respond and even subject themselves to Nvidia already having something better by then. It's a dangerous game for them, I don't think AMD is doing this on purpose.
Yes, You are totally right, They need to get the market share but going irrational, rush things and pump media with false info will not give'em this. Guys, AMD , Intel Nvidia, been in the market for so damn long. AMD was almost bankrupt and yet they are doing fine now. I'm sure AMD realizes this market share and it is not that they need to gain market share, they want this. AMD is aware and I can assure you they have a plan for all of this. AMD got CPUs out of the "dead" so lets wait and see what AMD will do with the GPUs. There hundreds of people thinking about the next steps and cautiously plan next moves. AMD needs market share is not groundbreaking news in my eyes.

cucker tarlson
you commented like a thousand times in this thread
No no brother. How do you know RDNA2 is RDNA on 7nm+ with Ray Tracing.
What the OP said about PS5in the thread says Based on RDNA2 which can mean anything. Either way I haven't seen it confirmed.

cucker tarlson
no,cause turing on 7nm would be amazing
pascal was maxwell on a shrunk die and it put amd so far behind that they still can't catch up with 1080Ti in 2020 unless you're counting radeon 7
I see. that explains everything.
Posted on Reply
Add your own comment