AMD Vega GPU architectural analysis

Interesting how AMD is is pitting this against Volta, and trying to leapfrog generations.

The thing I cannot wrap my head around is how they can pit it against Volta, of which next to nothing is known (unless we're talking corporate espionage lol)??
 
The thing I cannot wrap my head around is how they can pit it against Volta, of which next to nothing is known (unless we're talking corporate espionage lol)??

A part of me thinks AMD has lost it and is desperately cloying at the walls to climb out of the economic pit they're in, breaking their fingers in the process. But another part of me giggles with excitement at the thought of Vega not only competing and beating Pascal but competing with Volta. I find that very hard to believe, but still. It was initially rumoured that Vega 20 would be Vega 10 shrunk down using the 7nm process with 16GB of HBM2 running at 1TB/s and having vastly increased clock speeds and reduced TDP. They could then rebadge 14nm Vega 10 and 11 as midrange GPU's and the Polaris 10 chip as a low-end card. I doubt that, though, as optimising for both 14nm and 7nm in tandem might not be practical.
 
A part of me thinks AMD has lost it and is desperately cloying at the walls to climb out of the economic pit they're in, breaking their fingers in the process. But another part of me giggles with excitement at the thought of Vega not only competing and beating Pascal but competing with Volta. I find that very hard to believe, but still. It was initially rumoured that Vega 20 would be Vega 10 shrunk down using the 7nm process with 16GB of HBM2 running at 1TB/s and having vastly increased clock speeds and reduced TDP. They could then rebadge 14nm Vega 10 and 11 as midrange GPU's and the Polaris 10 chip as a low-end card. I doubt that, though, as optimising for both 14nm and 7nm in tandem might not be practical.

I just don't see it happening, it beating high end Pascal, let alone Volta, given their lengthy track record of big talk and not delivering (I recall the same before Polaris came out etc.). However, I don't know until it's out of course, but I cannot help but be weary and skeptical because of the aforementioned track record and the state the company has been in for many years now. And it is frustrating because I want them to deliver; it's the better outcome for everyone involved.
 
I just don't see it happening, it beating high end Pascal, let alone Volta, given their lengthy track record of big talk and not delivering (I recall the same before Polaris came out etc.). However, I don't know until it's out of course, but I cannot help but be weary and skeptical because of the aforementioned track record and the state the company has been in for many years now. And it is frustrating because I want them to deliver; it's the better outcome for everyone involved.

I don't know how much the rejig that recently occurred at AMD—the Radeon group being sectioned off and dedicating its resources to GPU's and Raja taking over—will impact things. The Fury line could have been a slight slip up. The 290X suffered from temperatures and power consumption, but it was a stellar GPU other than that. And one of the main reasons why it suffered from a higher TDP than its competitors was because AMD were stuck with 28nm for so long. Nvidia focused on the right thing while AMD focused on the wrong thing. They're correcting that mistake. Polaris was, at first, underperforming, but it is now comparable to Nvidia and in some cases superior. Temperatures are in check, power consumption is in check, pricing is in check, memory configuration is in check, availability is in check (both both AIB partners and AMD directly), drivers are in check, VR and high resolution performance is in check, DX11 and DX12 performance is in check, even overclocking is better with Polaris than Fury. If Polaris is performing so well, and Vega isn't just a beefed up Polaris, why can't AMD pull it off again?
 
AMD has been doing well financially this year so not sure what your talking about there. AMD's marketing has always been a bit strange though, sometimes really aggressive, sometimes coy. So I never take much stock in the marketing material just specs, architecture and performance numbers
 
AMD has been doing well financially this year so not sure what your talking about there. AMD's marketing has always been a bit strange though, sometimes really aggressive, sometimes coy. So I never take much stock in the marketing material just special, architecture and performance numbers

This year, maybe. But their income has been very low from dedicated GPU's in the last seven years compared to Nvidia. Even when AMD were making superior GPU's back in 2008 and beyond, Nvidia were outselling them and making a crap-ton more money than them.

So I'll rephrase: They may be trying to claw their way out of their 'dedicated GPU crisis'.
 
They could do it, for sure. They do hurt from having a smaller and less funded R&D department than Nvidia, it's been really noticeable since they downsized there, unfortunately. Which doesn't mean that they can't deliver a stellar product of course! But it sure has its impact. For me it's just a case of needing to see it before believing it kinda thing.
 
This year, maybe. But their income has been very low from dedicated GPU's in the last seven years compared to Nvidia. Even when AMD were making superior GPU's back in 2008 and beyond, Nvidia were outselling them and making a crap-ton more money than them.

So I'll rephrase: They may be trying to claw their way out of their 'dedicated GPU crisis'.

You can't compare there GPU finance reports. They both report very differently and they include different things.
Nvidia reports purely there consumer GPU sales. AMD does that and there pro and one other thing(can't remember what it was) and as we know they aren't doing great there. So it brings there report down and look bad because one of there markets can't overcome two other losses.

Nvidia didn't out sell like crazy, it was pretty close. Just Nvidia charged more and got on custom PC websites to basically only advertise Nvidia, so they were able to make more per GPU(as those sites charge more in general)

So I'll rephrase for you. They are competing. Everyone here is way overboard on this whole AMD dieing thing. We don't even have a launch date, just relax. It ain't over yet


As for this information in the article. This is all exactly what it needed and the memory thing is a big step forward for what Navi is going to be. It is going to be a pretty potent GPU. I think a lot of people will be surprised that it'll probably have much better performance per watt over Polaris which is a very good step that AMD needs
 
Last edited:
Don't worry about AMD financially they are doing fine. Unlike Nvidia they have managed to get their fingers into a couple of worthy pies and thus will have financial backing for any f ups they make.

Why do you think Nvidia are trying everything under the sun to come up with something other than a GPU? because they know the end is nigh (Moore's Law) and they desperately need another way of making revenue. Thankfully all of their attempts have failed because they think they are amazing at everything.

They'll never get a foothold in the tablet market and I would assume they have lost a ton of money and ran away with their tail between their collective legs (see also Shield) and now they are wasting money on an online streaming service that too will fail because they have ALL failed.

When the GPU war is over (soon) they will be in big trouble unless they can find a way of selling sand to Arabs.
 
Don't worry about AMD financially they are doing fine. Unlike Nvidia they have managed to get their fingers into a couple of worthy pies and thus will have financial backing for any f ups they make.

Why do you think Nvidia are trying everything under the sun to come up with something other than a GPU? because they know the end is nigh (Moore's Law) and they desperately need another way of making revenue. Thankfully all of their attempts have failed because they think they are amazing at everything.

They'll never get a foothold in the tablet market and I would assume they have lost a ton of money and ran away with their tail between their collective legs (see also Shield) and now they are wasting money on an online streaming service that too will fail because they have ALL failed.

When the GPU war is over (soon) they will be in big trouble unless they can find a way of selling sand to Arabs.

Not entirely true, Nvidia have managed to make inroads with graphics processing SOC's like tegra. The Nintendo switch uses nvidia hardware. Although I'd say amd are far more forward thinking with their adoption and promotion of open standards.
 
AMD have console and nVidia have mindshare.
I do feel that if AMD release as late as April they will have missed the boat to an extent.
Both companies are vastly different AMD simply does not have the clout that nVidia has to plough into R&D but they are doing ok for their size whereas nVidia are taking advantage of the consumer similar to Intel.

I honestly think AMD are going to surprise, no SHOCK the industry with both Ryzen and Vega but it wont be enough to pass nvidia due to the mindshare they have.
 
You can't compare there GPU finance reports. They both report very differently and they include different things.
Nvidia reports purely there consumer GPU sales. AMD does that and there pro and one other thing(can't remember what it was) and as we know they aren't doing great there. So it brings there report down and look bad because one of there markets can't overcome two other losses.

Nvidia didn't out sell like crazy, it was pretty close. Just Nvidia charged more and got on custom PC websites to basically only advertise Nvidia, so they were able to make more per GPU(as those sites charge more in general)

So I'll rephrase for you. They are competing. Everyone here is way overboard on this whole AMD dieing thing. We don't even have a launch date, just relax. It ain't over yet


As for this information in the article. This is all exactly what it needed and the memory thing is a big step forward for what Navi is going to be. It is going to be a pretty potent GPU. I think a lot of people will be surprised that it'll probably have much better performance per watt over Polaris which is a very good step that AMD needs

I don't understand. AMD's earning reports combine Pro and desktop GPU's, yet their earnings from these collectively are significantly below Nvidia's desktop GPU's alone. So yes, AMD's Pro series might hamper the numbers, but that doesn't matter; the numbers are still low which still translates to a loss. This is my point. AMD were making superior graphics cards back in 2009, but they were selling them at a loss in some cases while Nvidia were raking it in.

And how can you say it was "pretty close" when one of the main reasons why AMD reported losses over the last few years was because Nvidia owned 80% of the dedicated desktop GPU market and Intel owned around the same amount in desktop CPU's?

I'm basing my comments on this video and the associated quarterly earning PDF's:

https://www.youtube.com/watch?v=uN7i1bViOkU


edit: While it requires you to watch the entire video to find the specific details and numbers, the point I will draw attention to is at 24:15: "No matter where you look through the product stack, you would see Nvidia cards overwhelmingly outselling AMD cards."
 
Last edited:
Not entirely true, Nvidia have managed to make inroads with graphics processing SOC's like tegra. The Nintendo switch uses nvidia hardware. Although I'd say amd are far more forward thinking with their adoption and promotion of open standards.

How many Nintendo units can you see selling VS two different PS4s (the reg and the pro) and the Xbone?

lmao they won't even get within phone call distance.

I highly expect the Nintendo to flop (because every one knows what it is, underpowered etc) and the PS4 to make the most sales. The next Xbone will probably do OK given that the power scales are tipping in M$ direction.

If it wasn't for Pokemon Nintendo would have gone bust with the Wii U. It was a resounding failure.
 
Yeah, I have very low expectations for the Nintendo Switch.


We already know that graphics wise it is going to be pants. So that just leaves Mario, Zelda etc. Enough to make a few sales for the Wii U but they are going to have to offer something pretty damn crazy to make people buy this console just for those games.

IMO they would have made more money putting Zelda and Mario etc on the PS4 and Xbone (kinda like Sega did when the Dreamcast nearly sent them bankrupt).
 
I don't understand. AMD's earning reports combine Pro and desktop GPU's, yet their earnings from these collectively are significantly below Nvidia's desktop GPU's alone. So yes, AMD's Pro series might hamper the numbers, but that doesn't matter; the numbers are still low which still translates to a loss. This is my point. AMD were making superior graphics cards back in 2009, but they were selling them at a loss in some cases while Nvidia were raking it in.

And how can you say it was "pretty close" when one of the main reasons why AMD reported losses over the last few years was because Nvidia owned 80% of the dedicated desktop GPU market and Intel owned around the same amount in desktop CPU's?

I'm basing my comments on this video and the associated quarterly earning PDF's:

https://www.youtube.com/watch?v=uN7i1bViOkU


edit: While it requires you to watch the entire video to find the specific details and numbers, the point I will draw attention to is at 24:15: "No matter where you look through the product stack, you would see Nvidia cards overwhelmingly outselling AMD cards."

I know you are basing it off that video. And that video is wrong and misinformation.
 
I know you are basing it off that video. And that video is wrong and misinformation.

podcast-75-evidence.jpg


???

Great underground rapper, by the way.
 
podcast-75-evidence.jpg


???

Great underground rapper, by the way.

Who cares? What the heck is that related to? Surely not a GPU

If you are referring to the album name, well there is hardly any in the video. Operating profit that guy is using as a "basis" for a claim is NOT how you compare finance's. You need way way more than just that.
 
Last edited:
Back
Top