AMD Vega 10, Vega 20, and Vega 11 GPUs Detailed

[Audiophizile]

While it is proven AMD makes up a very small amount of ground in dx12(non-vulkan) and does really well with the 1 vulkan title it seems silly to buy a card that is STILL SLOWER in almost every or every game. I am trying to hold off for Vega (originally wanted to rebuild before bf1 release) but my hopes are understandably not crazy high. If replacing a GPU every 2 or so years, right now nvidia is a better bet as amd has really only proven it can win in vulkan. How many titles that you will actually play be vulkan before your next GPU upgrade? Probably WAY less than non-vulkan titles. I want Vega to be good. I want competition. I would prefer to not wait a full year from the 1080 release to get competition. If Vega doesn't outright beat a 1070 at $300-350 or beat a 1080 at $450-525 I will pretty much have no hope left in AMD. Volta will release in less time than between pascal and Vega so amd will need to work miracles on pricing to be a valid option.

 

[Divide Overflow]

Somewhere between all the hate and the hype exists the real product, for which I will patiently wait for W1zzard to review.

 

[Captain_Tom]

...except the benchmark you shared proves that the 1070 does better than the Fury, and that's before OC.

Proves what? The 980 Ti is below even the $300 Fury, and the 1070 ties the Fury X!!!


And overclock?! ROFL! My Fury is at 1135 MHz - I get 10% more performance. Pascal overclocks usually yield a 7% boost - WOW!

 

[Audiophizile]

Proves what? The 980 Ti is below even the $300 Fury, and the 1070 ties the Fury X!!!


And overclock?! ROFL! My Fury is at 1135 MHz - I get 10% more performance. Pascal overclocks usually yield a 7% boost - WOW!

The 1070 beats a furyx most games stock v stock. As it should being a gen later and a step below. That's how generations work usually... Next gen step down = last gen step up +/- a few % and much cheaper than the last gen step up launch price. But let's talk about current gen v current gen high end. 1070, 1080 and titan x. That ends this discussion.

Edit: just noticed a big mistake. The 1070 is actually 2 steps below a fury x as a fury x was meant to compete with the ti model. So things are looking pretty good there.

 

[looncraz]

And overclock?! ROFL! My Fury is at 1135 MHz - I get 10% more performance. Pascal overclocks usually yield a 7% boost - WOW!

Yeah, there's usually not much point in overclocking video cards.

The only real exception is with Polaris... GPU overclocks aren't worth much alone, but memory overclocks combined with higher GPU clocks can be quite good. Still, even with that, you're only talking about 15% more performance over reference... and I consider a 30% improvement to be the smallest worthwhile improvement to consider spending money on... unless you're just at the cusp of stable frame rates or just barely outside of your monitor's FreeSync range, when even 5~10% more performance can make the difference between stutters and silky-smooth experiences.

For me, though, I usually underclock my video cards. My RX 480 is currently running 640Mhz on the core and 1Ghz on the RAM. I can play most of the simple games and do all of the normal tasks that I want - while consuming only ~35W under load (above idle usage of ~20W).

I then have three other performance presets (using Afterburner).

900/1500 - Somewhat more demanding games (Crysis [Warhead] mostly)
1000/2000 - Demanding games (BF4, Hitman [Absolution], Civ V)
1288/2150 - Seriously Demanding Games (BF1 the only one so far)

Mind you, I have to keep frame rates at or below 146Hz and above 30Hz for FreeSync, so I tune things appropriately. All of my profile have significant under-volting applied and have had a few hours, at least, of games played with them.

The 1070 is actually 2 steps below a fury x as a fury x was meant to compete with the ti model. So things are looking pretty good there.

This is expected, Pascal and Polaris both represent a double-generation jump (well, Polaris anyway... Pascal is basically just a shrunken Maxwell with clock speed improvements enough for two generations of performance increase).

The only effort this generation that should really be applauded, IMHO, is AMD's. GCN4 is ~15% faster per clock, more efficient, brings quite a bit of meaningful new tech, and still clocks 15%+ higher... the fact that it was only used on a mid-range GPU is irrelevant from a technology perspective. Pascal is just the same old stuff with a few tweaks and clock speed improvements... boring stuff, really, albeit the feat is still impressive... it's basically just teams looking for critical paths and optimizing them away.

 

[Audiophizile]

This is expected, Pascal and Polaris both represent a double-generation jump (well, Polaris anyway... Pascal is basically just a shrunken Maxwell with clock speed improvements enough for two generations of performance increase).

The only effort this generation that should really be applauded, IMHO, is AMD's. GCN4 is ~15% faster per clock, more efficient, brings quite a bit of meaningful new tech, and still clocks 15%+ higher... the fact that it was only used on a mid-range GPU is irrelevant from a technology perspective. Pascal is just the same old stuff with a few tweaks and clock speed improvements... boring stuff, really, albeit the feat is still impressive... it's basically just teams looking for critical paths and optimizing them away.

While it is a good step for AMD and possibly better from a technology standpoint the efficiency is still not as good as nvidias. A 480 uses as much or more power than a 1070 while being substantially slower. That part is not so good.

 

[Captain_Tom]

The 1070 beats a furyx most games stock v stock. As it should being a gen later and a step below. That's how generations work usually... Next gen step down = last gen step up +/- a few % and much cheaper than the last gen step up launch price. But let's talk about current gen v current gen high end. 1070, 1080 and titan x. That ends this discussion.

Edit: just noticed a big mistake. The 1070 is actually 2 steps below a fury x as a fury x was meant to compete with the ti model. So things are looking pretty good there.

Edit: Thank you

This is expected, Pascal and Polaris both represent a double-generation jump (well, Polaris anyway... Pascal is basically just a shrunken Maxwell with clock speed improvements enough for two generations of performance increase).

The only effort this generation that should really be applauded, IMHO, is AMD's. GCN4 is ~15% faster per clock, more efficient, brings quite a bit of meaningful new tech, and still clocks 15%+ higher... the fact that it was only used on a mid-range GPU is irrelevant from a technology perspective. Pascal is just the same old stuff with a few tweaks and clock speed improvements... boring stuff, really, albeit the feat is still impressive... it's basically just teams looking for critical paths and optimizing them away.

It's very clear that AMD is just buying dirt cheap silicon to spit out chips so they can capture marketshare. If they were using the more expensive TSMC 16nm or if GloFlo 14nm was mature.... It would be a bloodbath, and Nvidia knows it.

 

[looncraz]

While it is a good step for AMD and possibly better from a technology standpoint the efficiency is still not as good as nvidias. A 480 uses as much or more power than a 1070 while being substantially slower. That part is not so good.

If AMD were to strip their GPU they could be "efficient," too!

Jokes aside, AMD does need to make some more serious efforts towards improving architectural efficiency. It mostly comes down to under-utilization, IMHO. If AMD had the same utilization rates as nVidia, they would be pretty close to just as efficient... instead AMD crams 35%+ more computing power to break even.

Per TFLOP, AMD isn't less efficient... but per FPS, they certainly are.

 

[Audiophizile]

If AMD were to strip their GPU they could be "efficient," too!

Jokes aside, AMD does need to make some more serious efforts towards improving architectural efficiency. It mostly comes down to under-utilization, IMHO. If AMD had the same utilization rates as nVidia, they would be pretty close to just as efficient... instead AMD crams 35%+ more computing power to break even.

Per TFLOP, AMD isn't less efficient... but per FPS, they certainly are.

The 480 has the same TFLOP as the 1070? I guess with vulkan we are starting to see more utilization with AMD which does start to level or even tip in amds favor at their price points but we're a long way off vulkan becoming the standard api. Again though, I think amds pricing for Vega is going to have to be extremely low because volta is most likely going to be a pretty large step over pascal. Vega coming out over a half year after pascal with Volta following most likely more closely than that... AMD needs to beat the 1070 and/or 1080 performance AND cost $50-75 less per card compared to the 1070/1080 or Volta will annihilate it.

 

[Captain_Tom]

The 480 has the same TFLOP as the 1070?

No the 1070 has 20% more TFLOP's. However they have the same 256 GB/s of bandwidth, and as such it should even out to the 1070 being 10% stronger if the 480 is fully utilized. However the wild card here i Async compute. I would wonder if that could allow AMD to gain even more than Nvidia per TFLOP (But I would highly doubt it for now).


Consider though that by the time Vega launches almost all AAA games will be DX12/Vulkan, and Vega will have at least 12 TFLOP's while the Titan has 11. Vega will also have more bandwidth too...

 

[thesmokingman]

Consider though that by the time Vega launches almost all AAA games will be DX12/Vulkan, and Vega will have at least 12 TFLOP's while the Titan has 11. Vega will also have more bandwidth too...

That's a bit, no... way freaking optimistic man. I want pure DX12 games as much as the next guy whose low level API biased, but devs are not going to jump on that gravy train because its not a gravy train since it cuts out many of their potential customers. I do think it will happen... eventually, years from now.

 

[Captain_Tom]

That's a bit, no... way freaking optimistic man. I want pure DX12 games as much as the next guy whose low level API biased, but devs are not going to jump on that gravy train because its not a gravy train since it cuts out many of their potential customers. I do think it will happen... eventually, years from now.

I mean what games won't be? Dishonered 2 and a freaking Skyrim re-release. Those games don't need it lol.

 

[looncraz]

The 480 has the same TFLOP as the 1070? I guess with vulkan we are starting to see more utilization with AMD which does start to level or even tip in amds favor at their price points but we're a long way off vulkan becoming the standard api. Again though, I think amds pricing for Vega is going to have to be extremely low because volta is most likely going to be a pretty large step over pascal. Vega coming out over a half year after pascal with Volta following most likely more closely than that... AMD needs to beat the 1070 and/or 1080 performance AND cost $50-75 less per card compared to the 1070/1080 or Volta will annihilate it.

RX 480 has 5.8TFLOPS @ 1.266Ghz - GTX 1070 has 6.5TFLOPS at 1.683GHz. GTX 1060 has 4.4TFLOPS at 1.709GHz... and is about the same performance as RX 480 at 1.266GHz if both are pegged at the stated clocks (GTX 1060 usually runs at > 1.8GHz stock, RX 480 at ~1.2GHz stock, giving the GTX 1060 about a 10% lead in performance).

Vulkan/DX 12 are allowing AMD GPUs to fill in the gaps left by their scheduler windows... gaps which really shouldn't exist as much as they do in the first place. AMD needs a new ABI, scheduler, and driver in order to get rid of more of those gaps using DX11... but they could, conceivably, then see a ~25% boost in performance... without providing more hardware processing power...

Vega will need to come out with a low price, no doubt. This is one of the major reason I think AMD will use a 2048-bit HBM2 bus... two stacks are cheaper than four.. and even if the costs are not so much better, you still have a smaller die/interposer or both to help.

Volta will trash Vega 10, no doubt. The only way this wouldn't happen is if Vega is using the long-rumored new ABI... then we could see Vega 10 being
as much as twice as fast as the Fury X. I have serious doubts about this, though... the same team that created Polaris created Vega just six months later, so it's most certainly a direct a descendant of GCN 4... and there's some indication in my own circle that AMD has abandoned the new ABI altogether as the software is catching up and they anticipate DX11 performance to become irrelevant - as well as their utilization issues... all with no work on their part.

 

[Audiophizile]

That's a bit, no... way freaking optimistic man. I want pure DX12 games as much as the next guy whose low level API biased, but devs are not going to jump on that gravy train because its not a gravy train since it cuts out many of their potential customers. I do think it will happen... eventually, years from now.

Yeah. Not even bf1 is pure dx12, dx11 with "some dx12 features". And dx12 alone doesn't greatly favor AMD. Vulkan being a standard is not anytime soon sadly. The fact Volta will follow Vega closer than Vega followed pascal and Volta will have a sync is AMDs biggest problem. Their timing is very poor.

 

[phanbuey]

 

[looncraz]

Yeah. Not even bf1 is pure dx12, dx11 with "some dx12 features". And dx12 alone doesn't greatly favor AMD. Vulkan being a standard is not anytime soon sadly. The fact Volta will follow Vega closer than Vega followed pascal and Volta will have a sync is AMDs biggest problem. Their timing is very poor.

nVidia just doesn't have as much to gain from DX12/Vulkan as AMD. nVidia spent a great deal of resources optimizing their drivers for DX11 and OpenGL - so their performance is already quite great in those APIs relative to their hardware capabilities. Async compute is good, but its greatest benefit will be seen in architectures with large empty time slices in its execution - AMD's hardware has large periods of time when it's just waiting for work... nVidia's does not. I don't see Volta changing that dynamic (though there are always ways to improve asynchronous workloads).

Timing, however, is a major issue for AMD. Unless Vega 10 is much better than anticipated, Volta will wipe the floor with Vega... and AMD's own timeline doesn't have Navi until 2018... giving nVidia the top tiers for nearly two years.

 

[Audiophizile]

nVidia just doesn't have as much to gain from DX12/Vulkan as AMD. nVidia spent a great deal of resources optimizing their drivers for DX11 and OpenGL - so their performance is already quite great in those APIs relative to their hardware capabilities. Async compute is good, but its greatest benefit will be seen in architectures with large empty time slices in its execution - AMD's hardware has large periods of time when it's just waiting for work... nVidia's does not. I don't see Volta changing that dynamic (though there are always ways to improve asynchronous workloads).

Timing, however, is a major issue for AMD. Unless Vega 10 is much better than anticipated, Volta will wipe the floor with Vega... and AMD's own timeline doesn't have Navi until 2018... giving nVidia the top tiers for nearly two years.

Agreed. I have hope but even if it doesn't work out bad for them financially their timing is hurting us(consumers) no matter what way you look at it. Unless, as you stated, Vega is much better than anticipated. That rarely is ever the case with any company however.

 

[HTC]

nVidia just doesn't have as much to gain from DX12/Vulkan as AMD. nVidia spent a great deal of resources optimizing their drivers for DX11 and OpenGL - so their performance is already quite great in those APIs relative to their hardware capabilities. Async compute is good, but its greatest benefit will be seen in architectures with large empty time slices in its execution - AMD's hardware has large periods of time when it's just waiting for work... nVidia's does not. I don't see Volta changing that dynamic (though there are always ways to improve asynchronous workloads).

Timing, however, is a major issue for AMD. Unless Vega 10 is much better than anticipated, Volta will wipe the floor with Vega... and AMD's own timeline doesn't have Navi until 2018... giving nVidia the top tiers for nearly two years.

Not necessarily: remember why nVidia dropped allot of it's compute power in the 1st place? Volta will have async but it will come @ a cost: speed AND efficiency.

I've got serious doubts nVidia will manage to keep it's speed advantage over AMD (their cards run @ a much higher speed then AMD's) and, if so, that also means efficiency will suffer: time will tell. If they do manage to keep the speed WHILE incorporating async, then i agree with you.

 

[Prima.Vera]

I mean even if AMD lunches something on Spring 2016, doesn't mean that nVidia wont do the same. I have a feeling we are going to see the 11xx series much earlier next year, probably to be lunched on a time frame simmilar with AMDs. Or maybe will see an 1080Ti delayed until next year?

 

[thesmokingman]

When Vegans attack is when we will see the rumored Ti.

 

[laszlo]

over 100 posts arguing & debating what?

is like a never ending soap tv show when every week a new rumor appear...

"The future has a way of arriving unannounced." George Will

 

[ppn]

It's just 25% overclock over 14nm shrink, all it is. will be. 2018, 7nm is just another shrink. 15% GCN optimizations don't count on it. FuryX wasn't particularly efficient compared to 390X. add 25% to it and you get the picture., assuming memory will not overclock (memory bandwidth is the same) it could be as low as 15%. Here where 15% GCN+memory compression comes into play to pull it back to 25%.

 

[the54thvoid]

I have to ask what games it is "trouncing" in:

https://www.techpowerup.com/forums/...a-11-gpus-detailed.226012/page-5#post-3526293


On average the Fury X is beating it in every resolution but 1080p (And even then it loses by 4%). Furthermore the best indicators of this fall and next year's game performance are DOOM and Deus EX where the Fury X does in fact TROUNCE the 980 Ti by 20%.

I meant that my card trounces fps. Not that it trounces Fury X. I'm getting (@1440p) framerates that are high enough to support usually maxing out all settings or just dropping a little. Also, using TPU graphs and checking benchmarks, my card is 15-25% faster (in actual fps) than the TPU benches (I'm at 1501 under water). So my card is holding up very well indeed. So much so, I don't see any point upgrading (1080 is only 20% faster and Titan X is £1100 . On my cards performance, a Fury X is a downgrade/sidegrade. And again, for me, I use DX11. I'm moving to W10 when i get a new cpu build (Zen or Kabylake).

I did see this tidbit from WCCFtech that got me thinking WTF?

And the recently leaked Vega 20 flagship GPU. A cutting edge graphics chip debuting on 7nm FinFet with 12 teraflops of single precision compute horsepower and double that in half precision compute. Making it the fastest ever graphics chip we’ve learned about to date, edging out Nvidia’s monstrous GP100 GPU and Tesla P100 accelerator by a fair margin.

It's hype to the max when a website is looking at a card to be released in 3 years time and then comparing it against a card out now (not AMD's fault). This is part of an ongoing problem when AMD start to release info, the rest of the web starts to pump it up, unfortunately.

Anyhoo, I've said enough here. Suffice to say let's not diss the 980ti as it still is a very powerful card when it's let loose, especially so if you're still on DX11 and lots of us still are. And in the UK, the main suppliers don't have many (if any) Fury X on sale and they are still £500 (not $300). Conversely, the 980ti's are under £350.

Funny ol' world.

ps- can people stop double and triple posting? Use multi-quote.

EDIT: I had to come back

Given this was a year ago and I checked the only 2 games left in TPU suite (Witcher 3 and BF4) both give gains to the 980ti a year on (as does Fury X). To understand at least my point and that of enthusiasts with overclocked 980ti's. My card runs pretty much where the OC Lightning is. You need to understand that performance to see why people defend their beloved 980ti's so much. A stock 980ti is meaningless when looked at alone. That was why I never bought a Fury X - it released at an awful OC scenario. Even with voltage it pales in comparison to a 980ti.

So for some performance perspective, this is why you shouldn't write of a 980ti. In that example below it's >30% above stock. That's where my card games.

 

[$ReaPeR$]

Is the 1070 EOL? Fury Nitro matches it, and the Fury X trades blows with the 1080 in the latest games. Say whatever you want - an EOL AMD product matches Nvida's new High-end cards.


All future poofing means for me is that the product will continue to provide the same experience, or even a better experience for 2-3 years. People who bought an i7 5 years ago continue to get 60-100 FPS in the latest games while the old i3's need to be replaced (SB i3's can't really do 60 FPS anymore). Of course nothing is completely future proof, but I expect my products to maintain their relative value for 3 years, and frankly they should never lose performance to their launch competition.

mate, my point was that it is pretty much pointless to compare an EOL product to a brand new one for the majority of people. AMD at this point needs numbers, market share, and it won't get it from an EOL product, no matter how good it is.

 

[LemmingOverlord]

Did AMD say they're positioning between GP104 and GP102? If so, that's bad news. Given the time frame of 2nd half 2017, not first quarter, it's almost definitely a Summer release for 2017. That gives Nvidia a huge leeway for current pricing and Volta development.
Also, of tremendous importance is that this is an investors conference so they need to say all the absolute best things.
I have a bad feeling about Vega. Even if it's better than GTX1080, AMD are saying it won't beat Titan X?
Sad face.

I've been listening to Papermaster's presentation at the DBTC, but I failed to hear any reference to Vega... The presentation was short (10 mins or so) with the rest being Q&A. I haven't heard it all yet, but... if true it's a massive letdown.