• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD VEGA confirmed for 2017 H1

Status
Not open for further replies.
And some drivers that may or may not work at launch time !!!

I think AMD should name their software "The Quantum Drivers" as they will work well at some point but no one is quite sure when.:D
Fine wine tech brother. Built in free performance which will get unlocked over time. I am down with that as long as day one driver performance is at a decent level to begin with and priced accordingly :p:)

I know that means nothing to you, but I like free performance ;)
 
Shall be interesting to see what devs do when the Vega architecture becomes more ubiquitous, considering FP16 is available in the PS4 pro and partly in the Xbox scorpio.

Only real Pascal part is GP100, since it is the only one with 2xSP for 16bit Floats. Makes me wonder if they added FP16 support for the entire Volta Lineup.
 
I bloody hope so, this is the longest talked about launch ever lol.

And we said exactly the same thing last year with Polaris and then the year before with Fiji. Waiting for AMD GPUs now just seems to be a regular thing.

So far this year, it's going fairly similarly to the release of Fury. Nvidia releases their Ti a shortwhile before, AMD releases a disappointing refresh that will soon do better due to driver improvements... what are the chances that Vega will be like Fury? With terrible levels of stock, performance above an x80 but below the x80ti (except in DX12/games which prefer AMD GPUs) and pricing equivalent to the x80ti...

De Ja Vu? Could be the Fury all over again. And to be fair... this time, most of us would actually welcome it, even if the DX11 performance might make it feel like worse bang for buck than a 1080ti. I'd buy this GPU just to give AMD a boost (despite not wanting to buy a 1080ti due to its price) and also to finally make use of the FreeSync feature on my monitor.
 
And we said exactly the same thing last year with Polaris and then the year before with Fiji. Waiting for AMD GPUs now just seems to be a regular thing.

So far this year, it's going fairly similarly to the release of Fury. Nvidia releases their Ti a shortwhile before, AMD releases a disappointing refresh that will soon do better due to driver improvements... what are the chances that Vega will be like Fury? With terrible levels of stock, performance above an x80 but below the x80ti (except in DX12/games which prefer AMD GPUs) and pricing equivalent to the x80ti...

De Ja Vu? Could be the Fury all over again. And to be fair... this time, most of us would actually welcome it, even if the DX11 performance might make it feel like worse bang for buck than a 1080ti. I'd buy this GPU just to give AMD a boost (despite not wanting to buy a 1080ti due to its price) and also to finally make use of the FreeSync feature on my monitor.
No one really knows. All the info people here are going on is the doom 4K benchmark. We don't know if that was on their best vega, we don't know how much driver improvement from January to June they will have made. What we do know is they are finally brining high performance graphics card designed on 14nm, with hbm2 and a new architecture which brings a bunch of improvements.

Forget about Polaris, do not compare it with that. That is where people get confused. Think Fury X and add all the improvements mentioned on top. That is why from the start I have been expecting Titan XP performance which really at the least it needs to be. Because likely soon with Volta that level of performance will be in the mid range bracket again in the shape of a 2070.

No one really knows. But what I can say with a high degree of confidence is, big Vega will be better than 1070/1080 which is what 99% of the people here predicted/expecting it will be. So as long as AMD get pricing right, there will be a lot of happy/surprised people at least. Few upset/angry ones too. Lol :p
 
No one really knows. But what I can say with a high degree of confidence is, big Vega will be better than 1070/1080 which is what 99% of the people here predicted/expecting it will be. So as long as AMD get pricing right, there will be a lot of happy/surprised people at least. Few upset/angry ones too. Lol :p

Agree with this. Even using the flawed Fiji arch, a 500mm^2 die (seen at Vega Preview) @ ~1000MHz will have 1080 performance. Incorporate the huge changes to Vega arch, in terms of efficiency and higher clock frequencies, and it is a good bet that Vega will greatly exceed 1080 performance.

I just can't fathom that they could have lower efficiency arch than Fiji:
Fiji has 8.9B transistors and a 596mm^2 die - 15M transistors/mm^2
Polaris has 5.7B transistors and 232mm^2 die - 24.5M transistors/mm^2

The (Very) rough maths goes, that Vega will have 12.25B transistors. That's a 37% uplift in transistors alone. Then add Raja has been quite transparent of AMD trying to achieve higher clocks with Vega.
On a different line of thought, if you did a linear scaling of Polaris (Rx480) performance per transistor, it would net a 114% increase in perf before any arch improvements. I gather 1080 is 50-80% quicker than RX480 (user bench) for reference.

Obviously I'm assuming all transistors go towards performance, which is absolutely not the case. But it gives some weight to the 500mm^2 die variant of Vega >1080 claim.

What I love about this is that the 1080 has 7.7B transistors! Significantly less than Fury :/ SO much room for improvement!

Edit: The worry is if AMD can only beat the 1080 and not the 1080ti with a 500mm^2 die. Very costly!
 
Agree with this. Even using the flawed Fiji arch, a 500mm^2 die (seen at Vega Preview) @ ~1000MHz will have 1080 performance. Incorporate the huge changes to Vega arch, in terms of efficiency and higher clock frequencies, and it is a good bet that Vega will greatly exceed 1080 performance.

I just can't fathom that they could have lower efficiency arch than Fiji:
Fiji has 8.9B transistors and a 596mm^2 die - 15M transistors/mm^2
Polaris has 5.7B transistors and 232mm^2 die - 24.5M transistors/mm^2

The (Very) rough maths goes, that Vega will have 12.25B transistors. That's a 37% uplift in transistors alone. Then add Raja has been quite transparent of AMD trying to achieve higher clocks with Vega.
On a different line of thought, if you did a linear scaling of Polaris (Rx480) performance per transistor, it would net a 114% increase in perf before any arch improvements. I gather 1080 is 50-80% quicker than RX480 (user bench) for reference.

Obviously I'm assuming all transistors go towards performance, which is absolutely not the case. But it gives some weight to the 500mm^2 die variant of Vega >1080 claim.

What I love about this is that the 1080 has 7.7B transistors! Significantly less than Fury :/ SO much room for improvement!

Edit: The worry is if AMD can only beat the 1080 and not the 1080ti with a 500mm^2 die. Very costly!
Good stuff!

Shows Nvidia's efficiency in getting the most out of what is theoretically a less capable GPU.

BUT, I think you should be careful in extrapolating this kind of thing. It was a common belief that Polaris would be in the 390X range of performance based on die shrink alone when compared to Hawaii. The math supports this quite well. But it turned out weaker than that.
 
Proper facts are great.
Propppahhh!

Im an old timer now in this game but it ammuses me how people hate on Nvidia for dominating the way they are. At the end of the day their products are brilliant. Last 2 years for AMD has been a **** show and its only now their drivers are stable enough. I purchased a 7950 2 years ago and only in last 6 months have I been able to get an overclock on it cause their drivers kept constantly crashing at the hint of an overclock. Sad but true...

With that said, apart from looking for a second hand 980ti, I want to see what Vega will do and where my money will be going so i can play quake for another 20+ years with downgraded graphics and 1000fps.
 
AMD obviously still have inefficiencies in other pipelines, but their asynchronous capabilities are still a *proper* step ahead of what Nvidia have. You could argue that if AMD were more efficienct in these areas that their async benefits would be reduced, and while that could be true in certain instances, that still doesn't change the fact that AMD's setup has a much higher operational potential. And that this potential is only more likely to be realized as time goes on.
And we said exactly the same thing last year with Polaris and then the year before with Fiji. Waiting for AMD GPUs now just seems to be a regular thing.

So far this year, it's going fairly similarly to the release of Fury. Nvidia releases their Ti a shortwhile before, AMD releases a disappointing refresh that will soon do better due to driver improvements... what are the chances that Vega will be like Fury? With terrible levels of stock, performance above an x80 but below the x80ti (except in DX12/games which prefer AMD GPUs) and pricing equivalent to the x80ti...

De Ja Vu? Could be the Fury all over again. And to be fair... this time, most of us would actually welcome it, even if the DX11 performance might make it feel like worse bang for buck than a 1080ti. I'd buy this GPU just to give AMD a boost (despite not wanting to buy a 1080ti due to its price) and also to finally make use of the FreeSync feature on my monitor.

It's a shame more gamers don't think like this,
I never bought into Maxwell, I waited for Fiji simply to support AMD because if they leave the graphics market we'll all be getting screwed,
Nvidia gave us a taste of what it would be like with how they've handled the 1070/1080 release.

When I actually bought my Fury I was a little concerned about driver support too but they've done a great job over the last 18 months,
so good in fact that it gave me the confidence to go and get a Freesync monitor and not buy into Pascal.
 
It's a shame more gamers don't think like this,
I never bought into Maxwell, I waited for Fiji simply to support AMD because if they leave the graphics market we'll all be getting screwed,
Nvidia gave us a taste of what it would be like with how they've handled the 1070/1080 release.

When I actually bought my Fury I was a little concerned about driver support too but they've done a great job over the last 18 months,
so good in fact that it gave me the confidence to go and get a Freesync monitor and not buy into Pascal.
What gamers should be thinking about it how games are progressively getting worse. For example, AI hasnt really moved on in the way it should have - a great youtube channel 'WorthABuy' highlights this on a weekly basis. Developers are getting sloppier and sloppier.

They done a great job over the past 6-9 months on drivers not 18months on the driver front.
 
Propppahhh!

Im an old timer now in this game but it ammuses me how people hate on Nvidia for dominating the way they are. At the end of the day their products are brilliant. Last 2 years for AMD has been a **** show and its only now their drivers are stable enough. I purchased a 7950 2 years ago and only in last 6 months have I been able to get an overclock on it cause their drivers kept constantly crashing at the hint of an overclock. Sad but true...

With that said, apart from looking for a second hand 980ti, I want to see what Vega will do and where my money will be going so i can play quake for another 20+ years with downgraded graphics and 1000fps.

If you want more out of your 7950 flash it to a 7970 :)
 
And to be fair... this time, most of us would actually welcome it, even if the DX11 performance might make it feel like worse bang for buck than a 1080ti. I'd buy this GPU just to give AMD a boost (despite not wanting to buy a 1080ti due to its price) and also to finally make use of the FreeSync feature on my monitor.

While I can understand the sentiment, I disagree with the "charity" aspect. This used to make sense, but with the arrival and success of Ryzen lack of competition due to AMD going bankrupt is off the table. They need to deliver now.

In my eyes though, people are painting a pretty bleak picture for no reason. Take the RX480 for example: it actually has decent DX11 performance. I honestly don't feel like I'm missing out anything when using it. Sure the 1060 can get 5-10% more FPS in some games but that's not major (the 480 does the same in other games). And even though it has more power consumption, it has its own advantages (more RAM, better at new APIs, Freesync support). It's a very competitive offering.

If AMD had delivered something like that against the 1080 a year ago (close DX11/betterDX12, just with higher consumption) I would've gone for it due to Freesync and expecting DX12/Vulkan to pick up in AAA titles...
 
Status
Not open for further replies.
Back
Top Bottom