Vega/Navi Rumors (Updated)

[tential]

I am actually staggered that nobody spotted a rumor about Vega GPUs, that appeared on another forum, few days ago...

Enlighten me. I may look at other forums once a month or so for tech. Mostly when I read something really really bad on here.

I'm not sure what the rush was with Ryzen. It's not like Intel has something to crush them if it was delayed another 3 months.

Vega has to compete against a chip that's already been out (titan xp is the 1080ti chip). That should take priority. I shouldn't be waiting for news for Vega. I should be like 4 weeks tops away from Vega now, but Vega should have been within 3 weeks of the 1080ti release.

 

[Glo.]

The problem is that I think it will create unnecessary hype for Vega architecture.

Let me spoil one thing. According to the post, AMD believes that they will achieve with Vega 50% of Marketshare.

With High-End GPUs, only? No way.

 

[Mopetar]

I'm assuming AMD would count APUs in that 50% assuming they use Vega-based graphics.

Unless they have a monstrous performance advantage over Nvidia, I don't see them getting to 50% in the discrete market.

Market share stats that aren't qualified in any way are meaningless. That statement leaves a lot to interpretation.

 

[Glo.]

I'm assuming AMD would count APUs in that 50% assuming they use Vega-based graphics.

Unless they have a monstrous performance advantage over Nvidia, I don't see them getting to 50% in the discrete market.

Market share stats that aren't qualified in any way are meaningless. That statement leaves a lot to interpretation.

There is a problem with this idea. Even if Raven Ridge is based on Zen+Vega architecture, this year, there is supposedly only Mobile.

I think overall we all agree that this number accounts for dGPUs and APUs with Vega architecture.

 

[w3rd]

Vega uarch is more powerful than Pascal. And will be cheaper than $699.

The 1080ti is the best NVidia can do with pascal, the only thing left is for NVidia to drop the price on the 1080ti to $499 when Vega hits. Right?

 

[Mopetar]

There is a problem with this idea. Even if Raven Ridge is based on Zen+Vega architecture, this year, there is supposedly only Mobile.

I think overall we all agree that this number accounts for dGPUs and APUs with Vega architecture.

Doesn't matter if it's mobile only. A lot of low cost notebooks that can run games like Dota 2, LoL, CS, etc. at good settings will net AMD a lot of sales, probably more so than having desktop APUs would.

If the get it out during the summer they'll get a lot of back to school sales and college kids that want a notebook for class as well as gaming. Now that AMD doesn't have anemic CPU performance, they'll be a clear favorite over Intel for that part of the market.

 

[happy medium]

Vega uarch is more powerful than Pascal. And will be cheaper than $699.

Says who?

The 1080ti is the best NVidia can do with pascal, the only thing left is for NVidia to drop the price on the 1080ti to $499 when Vega hits. Right?

no, there is a full faster Pascal.

answers and question in red.

 

[Elfear]

answers and question in red.

+1

I'd love for Vega to trounce Pascal if only to help competition in the high-end GPU market, but I haven't seen anything to indicate that will be the case.

 

[Ajay]

The problem is that I think it will create unnecessary hype for Vega architecture.

Let me spoil one thing. According to the post, AMD believes that they will achieve with Vega 50% of Marketshare.

With High-End GPUs, only? No way.

What do you have to worry about? Do you work for AMD???
Pls post - teases are pretty lame and I couldn't find any new info via a google search.

 

[Magee_MC]

What do you have to worry about? Do you work for AMD???
Pls post - teases are pretty lame and I couldn't find any new info via a google search.

Here you go. I found the reference.

http://semiaccurate.com/forums/showthread.php?t=9377&page=75

 

[w3rd]

answers and question in red.

Says logic. Vega is meant to compete with Volta.

And full Pascal is EOL. Or do you expect Pascal to compete with Volta..? Your logic is not making any sense. So, the GTX1080ti will end up being the 8800gt of this era. We'll probably see GTX1080ti's selling for under $400 bucks in one years time.

Your answers are in red.

 

[Ajay]

Here you go. I found the reference.

http://semiaccurate.com/forums/showthread.php?t=9377&page=75

Thanks!

 

[happy medium]

Says logic. Vega is meant to compete with Volta.

And full Pascal is EOL. Or do you expect Pascal to compete with Volta..? Your logic is not making any sense. So, the GTX1080ti will end up being the 8800gt of this era. We'll probably see GTX1080ti's selling for under $400 bucks in one years time.

Your answers are in red.

Did your logic ever here of Navi?
Never mind, don't bother.

 

[Ajay]

The problem is that I think it will create unnecessary hype for Vega architecture.

Let me spoil one thing. According to the post, AMD believes that they will achieve with Vega 50% of Marketshare.

With High-End GPUs, only? No way.

Geez, 50% market share in one GPU generation. Nvidia isn't sitting still, so I doubt it. The answer to his question was also a marketing answer, not a technical one. I sure hope the performance is high, as it would create more options and better pricing for consumers (and help keep AMD in Business).

 

[Valantar]

Vega uarch is more powerful than Pascal. And will be cheaper than $699.

The 1080ti is the best NVidia can do with pascal, the only thing left is for NVidia to drop the price on the 1080ti to $499 when Vega hits. Right?

Says logic. Vega is meant to compete with Volta.

And full Pascal is EOL. Or do you expect Pascal to compete with Volta..? Your logic is not making any sense. So, the GTX1080ti will end up being the 8800gt of this era. We'll probably see GTX1080ti's selling for under $400 bucks in one years time.

No. You have no basis for speculating like that, let alone with the level of certainty you're giving off. The only demo we've seen of Vega has shown it slightly beating a 1080 in a relatively friendly game. We don't know how well (or badly) optimized the drivers were at that point, the clocks of the ES chip in question, or really anything else. We don't know the specs of any Vega GPU. Do I believe AMD has the knowledge, skills and engineers to compete with Pascal, and even Volta? Sure. But it would take a lot for them to leapfrog Pascal completely. After all, they'd need to (at the very least) catch up on efficiency for that to be feasible. Or do you see AMD launching a 300W+ GPU with Vega? That would be suicide. Rather than getting your hopes up for something sensational, how about we all keep a sane critical distance and hope AMD stays competitive? I'd much rather get a happy surprise than be disappointed.

There is a problem with this idea. Even if Raven Ridge is based on Zen+Vega architecture, this year, there is supposedly only Mobile.

I think overall we all agree that this number accounts for dGPUs and APUs with Vega architecture.

Raven Ridge is only mobile? Says who? What's with the APU support on all the AM4 motherboards, then? They're definitely not relaunching some Excavator chip for the desktop. Not happening. And aren't people pretty sure RR is Zen + Polaris? Zen + Vega would be great, but it seems optimistic at best.

 

[tamz_msc]

Raven Ridge is only mobile? Says who? What's with the APU support on all the AM4 motherboards, then? They're definitely not relaunching some Excavator chip for the desktop. Not happening. And aren't people pretty sure RR is Zen + Polaris? Zen + Vega would be great, but it seems optimistic at best.

RR desktop is coming next year. This year it's mobile that gets it first. Bristol Ridge is launching for AM4 desktop this year. See AT interview with Lisa Su.

 

[Valantar]

RR desktop is coming next year. This year it's mobile that gets it first. Bristol Ridge is launching for AM4 desktop this year. See AT interview with Lisa Su.

That statement is (probably intentionally) vague. The question is specific, but she pivots to talking about "an APU" rather than mentioning BR specifically. While I do believe RR on mobile is their priority (it'd probably sell 100x or more compared to what it would on desktop, so that's a no-brainer), I'm reluctant to believe they'd release BR to retail a) a year after its launch, and b) six months after launching a vastly superior architecture. I'm not saying BR is bad (I'm very fond of the Kaveri in my HTPC), but that's like Intel releasing a new Core2Quad part alongside Sandy Bridge. It would not be a smart move, even if the upgrade path for those PCs would be amazing.

I mean, what would be the limiting factor in terms of releasing desktop RR alongside mobile RR? Fab capacity? Doubt it. Yields? also doubtful - GloFo isn't a small company, and their 14nm process is no longer new. And, after all, the die would be the same, just binned differently. I'd rather see a desktop RR launch with limited volume (due to the majority going to mobile) than retail BR launching alongside mobile RR. I guess that'd be the first time a company launches a mobile CPU(or APU) that could run circles around their desktop counterpart.

On the other hand, you might be right. They might have some kind of plan in place. Or they might just be thinking "oh, f*ck it."

But now we've veered pretty far off the Vega/Navi topic, no?

 

[Glo.]

Raven Ridge is only mobile? Says who? What's with the APU support on all the AM4 motherboards, then? They're definitely not relaunching some Excavator chip for the desktop. Not happening. And aren't people pretty sure RR is Zen + Polaris? Zen + Vega would be great, but it seems optimistic at best.

Raven Ridge is mobile for this year. Next year, according to leaked roadmaps we will get desktop RR chips.

APUs will get Zen+ Vega. This is not optimistic, it is a fact.

Whole history about the launch of the APUs is at least muddy, and uncertain to large degree.

 

[w3rd]

No. You have no basis for speculating like that, let alone with the level of certainty you're giving off. The only demo we've seen of Vega has shown it slightly beating a 1080 in a relatively friendly game. We don't know how well (or badly) optimized the drivers were at that point, the clocks of the ES chip in question, or really anything else. We don't know the specs of any Vega GPU. Do I believe AMD has the knowledge, skills and engineers to compete with Pascal, and even Volta? Sure. But it would take a lot for them to leapfrog Pascal completely. After all, they'd need to (at the very least) catch up on efficiency for that to be feasible. Or do you see AMD launching a 300W+ GPU with Vega? That would be suicide. Rather than getting your hopes up for something sensational, how about we all keep a sane critical distance and hope AMD stays competitive? I'd much rather get a happy surprise than be disappointed.


Raven Ridge is only mobile? Says who? What's with the APU support on all the AM4 motherboards, then? They're definitely not relaunching some Excavator chip for the desktop. Not happening. And aren't people pretty sure RR is Zen + Polaris? Zen + Vega would be great, but it seems optimistic at best.

My friend, you seem extremely uninformed.

Coincidentally, I have all the basis in the world to say the things I did. Counter to my argument, YOU would have to prove to everyone else, that AMD's Vega technology (being released in 2 months time), would be worse than Nvidia's Pascal, that was released last year.

Essentially, You are saying the AMD can not possibly ever beat Nvidia, because Nvidia yo...



There is zero logic in thinking NVidia's 1 year old architecture, is as advance as AMD's Vega. What proof do you have..? I know the slides I have seen from Vega presentation, Pascal doesn't have anything like those features, right..? Taken further, there is no proof that Nvidia's Volta next year, will be as advanced as Vega either.

Sorry about your nearsightedness.

 

[zinfamous]

My friend, you seem extremely uninformed.

Coincidentally, I have all the basis in the world to say the things I did. Counter to my argument, YOU would have to prove to everyone else, that AMD's Vega technology (being released in 2 months time), would be worse than Nvidia's Pascal, that was released last year.

Essentially, You are saying the AMD can not possibly ever beat Nvidia, because Nvidia yo...



There is zero logic in thinking NVidia's 1 year old architecture, is as advance as AMD's Vega. What proof do you have..? I know the slides I have seen from Vega presentation, Pascal doesn't have anything like those features, right..? Taken further, there is no proof that Nvidia's Volta next year, will be as advanced as Vega either.

Sorry about your nearsightedness.

The reality that you are ignoring is that what matters in the end is actual performance. Sure, one could fairly argue that Vega is more "Advanced" than Pascal, simply because it does things differently, simply because it has a greater theoretical TFlop advantage, but in the end--none of that matters if those super cool internal silicon whatsitgizmos are grossly underutilized by the intended software, and thus represents a real performance loss on the table. What good is 12.3TFlop potential, if the end result from application to application is only something in the neighborhood of 10.5 TFlop? This is the ongoing story with AMD, and you do yourself a credibility disservice by ignoring this historical truth.

Yes, I would like to see AMD come out of nowhere and show with a Vega 10 @, say, $500-600 that actually outperforms the latest and greatest 1080Ti from nVidia, but that is pure speculation to assume this happens, and is essentially based on ignoring the fundamental history of AMD's design with relation to actual developers. It is the same story over and over again, and I am not sure why you keep buying it.

I will say this: with the surprising pricing of 1080Ti, I do think nVidia has a pretty good idea of what Vega will bring, and I do see this as acknowledgement that Vega is probably a bit more legitimate than Fury and Fury X--so I see this as good news. I wouldn't hold out hope, though. All we actually know about Vega is that it has a 0-5% advantage over a 1 year-old card that, itself, now shows a ~20-30% disadvantage compared to its successor which is actually "available" now. Sure, you can say that those early Vega demos were unoptimized with crap drivers and crap systems...but does that represent a recoverable 20-25% gain in performance? I mean...I'd like for that to be the case, but my the rational side simply laughs at the proposition.

In the end, you have to look at price/performance. I'm going with the best-case scenario of Vega 10 coming it at a 10-15% disadvantage to 1080Ti @ 20-25% less cost. This leaves the expected potential for yet another "AMD ages much better" scenario where, at some point, that Vega will come very close to a 1080Ti at whatever future time, if not matching it, with a lower up-front cost. I feel that leaves enough room for a pleasant surprise, and far less disappointment.

 

[Headfoot]

My friend, you seem extremely uninformed.

Coincidentally, I have all the basis in the world to say the things I did. Counter to my argument, YOU would have to prove to everyone else, that AMD's Vega technology (being released in 2 months time), would be worse than Nvidia's Pascal, that was released last year.

Essentially, You are saying the AMD can not possibly ever beat Nvidia, because Nvidia yo...



There is zero logic in thinking NVidia's 1 year old architecture, is as advance as AMD's Vega. What proof do you have..? I know the slides I have seen from Vega presentation, Pascal doesn't have anything like those features, right..? Taken further, there is no proof that Nvidia's Volta next year, will be as advanced as Vega either.

Sorry about your nearsightedness.

Show a link of Vega performance of any kind. You have no facts to base your opinion on. You are speculating and passing it off as fact. You're not fooling anyone here.

 

[Valantar]

My friend, you seem extremely uninformed.

Coincidentally, I have all the basis in the world to say the things I did. Counter to my argument, YOU would have to prove to everyone else, that AMD's Vega technology (being released in 2 months time), would be worse than Nvidia's Pascal, that was released last year.

Essentially, You are saying the AMD can not possibly ever beat Nvidia, because Nvidia yo...



There is zero logic in thinking NVidia's 1 year old architecture, is as advance as AMD's Vega. What proof do you have..? I know the slides I have seen from Vega presentation, Pascal doesn't have anything like those features, right..? Taken further, there is no proof that Nvidia's Volta next year, will be as advanced as Vega either.

Sorry about your nearsightedness.

You're joking, right? Just because Vega is newer, it has to be vastly superior?

Again: the only data we have to go on is that it's slightly faster than the 1080 in ES form in a single game (Battlefront) a few months before launch. This is probably not at final clock speeds, and might (seriously) lack driver optimizations, but expecting huge gains from this seems overly optimistic. 30% better than this would be nigh on miraculous. I don't at all doubt that Vega will be good. Can it be great? Sure. But as of now, we truly have no way of knowing.

Another thing entirely is the simple fact that we know very, very little about the impact on gaming performance from the technological advancements AMD has made. HBCC seems tailor made for datacenters, workstations and large dataset computing. While streaming textures and other large data files straight from storage rather than through the CPU might show some performance gains, it also might not. Few games exceed 4GB of VRAM usage, let alone 8 or 12. On the other hand, anything limiting CPU and driver overhead is great. How dependent this will be of game-specific implementations is another question. And then there's efficiency. All we have to go by is that the ES cards have a 6-and-8-pin power setup. If that means the final high-end card is at or below 250W, that's great. Although at that power level, it'd better be far closer to the 1080Ti than the 180W 1080.

Now, please don't lose your cool. I'm not dissing AMD. I'm not - in any way whatsoever - saying Nvidia is better at making GPUs than them (hell, I'm very happy with my Fury X). I'm simply saying it'll be a great feat alone if AMD can come close to catching up with Nvidia's efficiency lead while also providing the same level of performance. That's what I'm hoping for, although expecting might be too strong a word. Hoping is more fitting. Now, expecting it to exceed this massively, like you do? To me, that sounds naïve at best. Don't raise your expectations too high, and please stop evangelizing based on speculation and hopes. That can do nothing but hurt AMD in the long run (they've seen more than their fair share of overhyped products, which has seriously damaged public perception of them).

 

[Glo.]

And I think, you know now why I was so hesitant to bring the "rumor" on this forum.

My friend, you seem extremely uninformed.

Coincidentally, I have all the basis in the world to say the things I did. Counter to my argument, YOU would have to prove to everyone else, that AMD's Vega technology (being released in 2 months time), would be worse than Nvidia's Pascal, that was released last year.

Essentially, You are saying the AMD can not possibly ever beat Nvidia, because Nvidia yo...



There is zero logic in thinking NVidia's 1 year old architecture, is as advance as AMD's Vega. What proof do you have..? I know the slides I have seen from Vega presentation, Pascal doesn't have anything like those features, right..? Taken further, there is no proof that Nvidia's Volta next year, will be as advanced as Vega either.

Sorry about your nearsightedness.

And you seem to be extremely confident in what you know about the Vego GPUs.

Nobody is buying what you are selling.

P.S. What does mean to you "that Nvidia 1 year old architecture is not as advanced as Vega is?". How does it translate into actual SOFTWARE performance?

 

[w3rd]

The reality that you are ignoring is that what matters in the end is actual performance. Sure, one could fairly argue that Vega is more "Advanced" than Pascal, simply because it does things differently, simply because it has a greater theoretical TFlop advantage, but in the end--none of that matters if those super cool internal silicon whatsitgizmos are grossly underutilized by the intended software, and thus represents a real performance loss on the table. What good is 12.3TFlop potential, if the end result from application to application is only something in the neighborhood of 10.5 TFlop? This is the ongoing story with AMD, and you do yourself a credibility disservice by ignoring this historical truth.

Yes, I would like to see AMD come out of nowhere and show with a Vega 10 @, say, $500-600 that actually outperforms the latest and greatest 1080Ti from nVidia, but that is pure speculation to assume this happens, and is essentially based on ignoring the fundamental history of AMD's design with relation to actual developers. It is the same story over and over again, and I am not sure why you keep buying it.

I will say this: with the surprising pricing of 1080Ti, I do think nVidia has a pretty good idea of what Vega will bring, and I do see this as acknowledgement that Vega is probably a bit more legitimate than Fury and Fury X--so I see this as good news. I wouldn't hold out hope, though. All we actually know about Vega is that it has a 0-5% advantage over a 1 year-old card that, itself, now shows a ~20-30% disadvantage compared to its successor which is actually "available" now. Sure, you can say that those early Vega demos were unoptimized with crap drivers and crap systems...but does that represent a recoverable 20-25% gain in performance? I mean...I'd like for that to be the case, but my the rational side simply laughs at the proposition.

In the end, you have to look at price/performance. I'm going with the best-case scenario of Vega 10 coming it at a 10-15% disadvantage to 1080Ti @ 20-25% less cost. This leaves the expected potential for yet another "AMD ages much better" scenario where, at some point, that Vega will come very close to a 1080Ti at whatever future time, if not matching it, with a lower up-front cost. I feel that leaves enough room for a pleasant surprise, and far less disappointment.

What matters at the end..?
OK, buy a 1080ti & Vega this summer. See which one is more powerful in 2 years time, when you are playing 2018 releases.

And again, Pascal is EOL bro. Nvidia's tech is EOL, that is why they are redesigning new chip with new technology for 2018. Right?


But you really give yourself away here:
" ..but that is pure speculation to assume this happens, and is essentially based on ignoring the fundamental history of AMD's design with relation to actual developers. It is the same story over and over again, and I am not sure why you keep buying it. "

What fundamental part of history..?
That AMD makes CPUs, APUs and GPUs..? While Nvidia only has experience in making a GPU..? Or the fundamental aspect that NVidia doesn't have, nor invested in HSA uArch..?


History..?
Nvidia has always had a GPU advantage, because Jen-Hsun Huang ALWAYS penned better fab deals, than AMD's CEOs. (Vega & Volta will be on par.) So it will come down to tech, not fab. Nvidia's cuda tech has dried up.

 

[Glo.]

But you really give yourself away here:
" ..but that is pure speculation to assume this happens, and is essentially based on ignoring the fundamental history of AMD's design with relation to actual developers. It is the same story over and over again, and I am not sure why you keep buying it. "

What fundamental part of history..?
That AMD makes CPUs, APUs and GPUs..? While Nvidia only has experience in making a GPU..? Or the fundamental aspect that NVidia doesn't have, nor invested in HSA uArch..?


History..?
Nvidia has always had a GPU advantage, because Jen-Hsun Huang ALWAYS penned better fab deals, than AMD's CEOs. (Vega & Volta will be on par.) So it will come down to tech, not fab. Nvidia's cuda tech has dried up.

If we would be looking for history we would be expecting Ryzen CPUs to be crap, as Bulldozer was.

When you get technology - nothing is 100% sure. Until we will see how hardware performance of Vega translates into Software performance your claims about it being more advanced are rendered baseless.