Seems that the only good card here is Vega 56. Vega 64 and Liquid are a dissaster.
But the real star will be Vega Nano. When is launch, very few cards of the same kind of Nano will stand a chance against it.
On a brief aside, the number of compute engines has been an unexpectedly interesting point of discussion over the years. Back in 2013 we learned that the then-current iteration of GCN had a maximum compute engine count of 4, which AMD has stuck to ever since, including the new Vega 10. Which in turn has fostered discussions about scalability in AMD’s designs, and compute/texture-to-ROP ratios.
Talking to AMD’s engineers about the matter, they haven’t taken any steps with Vega to change this. They have made it clear that 4 compute engines is not a fundamental limitation – they know how to build a design with more engines – however to do so would require additional work. In other words, the usual engineering trade-offs apply, with AMD’s engineers focusing on addressing things like HBCC and rasterization as opposed to doing the replumbing necessary for additional compute engines in Vega 10.
Vega 56 faster than VEga 64? A lot wrong going on in that game.....From Anandtech's review:
So basically, RTG's engineers are claiming they could have increased the number of compute engines (and thus the number of triangles/clock) if they wanted, but they just didn't care. This reinforces my belief that they are basically giving up on gaming. If they were taking gaming seriously, they would have increased the number of compute engines to 8 and the number of ROPs to 128. They had plenty of room to do this while still fitting within the reticle limit, and it would have prevented disgraceful results like this one:
Instead, we get a card that basically retains almost all of the potential bottlenecks of Fiji. In a handful of cases, the drivers can try to use tricks to reduce the amount of geometry rendered, but when these optimizations are not done or when the geometry actually does have to be drawn, then Vega is no better than Fiji.
I don't expect anything worthwhile in gaming for Navi, either. They aren't going to bother to make two GPUs act transparently as one - they'll expect the developers to do it with explicit multi-adapter in DX12. The developers won't; instead, they will continue to write easier-to-code DX11 titles that Nvidia's driver team will optimize to the point that they are as fast as DX12. Navi will still be limited to four compute engines and 64 ROPs because RTG simply doesn't care to do any better.
From Anandtech's review:
So basically, RTG's engineers are claiming they could have increased the number of compute engines (and thus the number of triangles/clock) if they wanted, but they just didn't care. This reinforces my belief that they are basically giving up on gaming. If they were taking gaming seriously, they would have increased the number of compute engines to 8 and the number of ROPs to 128. They had plenty of room to do this while still fitting within the reticle limit, and it would have prevented disgraceful results like this one:
Instead, we get a card that basically retains almost all of the potential bottlenecks of Fiji. In a handful of cases, the drivers can try to use tricks to reduce the amount of geometry rendered, but when these optimizations are not done or when the geometry actually does have to be drawn, then Vega is no better than Fiji.
I don't expect anything worthwhile in gaming for Navi, either. They aren't going to bother to make two GPUs act transparently as one - they'll expect the developers to do it with explicit multi-adapter in DX12. The developers won't; instead, they will continue to write easier-to-code DX11 titles that Nvidia's driver team will optimize to the point that they are as fast as DX12. Navi will still be limited to four compute engines and 64 ROPs because RTG simply doesn't care to do any better.
The delay, hyper focus on Zen, extreme focus on compute.. if that didn't tell you AMD's intentions, I don't know what else would. They have given up on gaming, but also failed at the GPU datacenter.
The pro-consumer market isn't large enough to sustain it, but whatever.
AMD isn't being dishonest in what you posted above. Vega 64 is the fastest GPU in pure compute under $500. And it will get over 60FPS in pretty much every game out. Now saying they are fastest based on compute alone is kind of cherry picking, but getting over 60fps in those games is not dishonest. Sure it may only get 90fps to a 1080Ti's 130fps, but those extra frames mean nothing to the vast majority of people with 60Hz displays.
And since the graphic was touting 4k performance (where Vega 64 is best) those extra frames above 60fps are meaningless from a visual standpoint. I have a 4k Freesync monitor, a really nice LG, and once the game is in the Freesync band any changes in fps aren't noticeable.
From Anandtech's review:
So basically, RTG's engineers are claiming they could have increased the number of compute engines (and thus the number of triangles/clock) if they wanted, but they just didn't care. This reinforces my belief that they are basically giving up on gaming. If they were taking gaming seriously, they would have increased the number of compute engines to 8 and the number of ROPs to 128. They had plenty of room to do this while still fitting within the reticle limit, and it would have prevented disgraceful results like this one:
Instead, we get a card that basically retains almost all of the potential bottlenecks of Fiji. In a handful of cases, the drivers can try to use tricks to reduce the amount of geometry rendered, but when these optimizations are not done or when the geometry actually does have to be drawn, then Vega is no better than Fiji.
I don't expect anything worthwhile in gaming for Navi, either. They aren't going to bother to make two GPUs act transparently as one - they'll expect the developers to do it with explicit multi-adapter in DX12. The developers won't; instead, they will continue to write easier-to-code DX11 titles that Nvidia's driver team will optimize to the point that they are as fast as DX12. Navi will still be limited to four compute engines and 64 ROPs because RTG simply doesn't care to do any better.
Guys how was the Wait for Vega?
Guys how was the Wait for Vega?
Its hard to tell if AMD (ATI) got worse over the years or if Nvidia just got way better. I honestly can't tell what happened, but here we are. I am starting to have the same feelings for AMD's GPU's as I did for their CPU's before Ryzen came along. I think AMD will need a massive jump in GPU performance to catch up, because they aren't catching up to 1080ti, which is already impossible enough. They will have to go straight from Vega performance to trying to cope with Volta. I doubt a Volta Ti card is going to play very nice with AMD next round.
Thanks to betting everything on HBM2, they have way less wiggle room when it comes to price.One of the most underwhelming GPU launches in recent memory, while Fermi had power and heat issues it had no trouble dethroning the 5870, this one struggles to beat even a 1080.The asking price is also quite atrocious to be honest, frankly after doing some initial benches they should have have skipped vega entirely and go straight to Navi.
Thanks to betting everything on HBM2, they have way less wiggle room when it comes to price.
It is sad that going full HBM did not pay off at all for them.
- They dont have a speed-advantage or power-advantage in the prosumer/datacenter market. From my POV, going full HBM *could* have made sense if they got some advantage out of it. But they did not.
- The gaming market does not need HBM at all. It only makes your GPU more expensive...and cuts your margins.
- No kidding, those TPU performance summary charts are going to start looking a little ridiculous with the 1080, 1080 ti, 2060, 2070, 2080 and 2080ti sitting uncontested at the top of the pile...
You're assuming that either of those is what's holding Vega back, which we have no proof of.From Anandtech's review:
So basically, RTG's engineers are claiming they could have increased the number of compute engines (and thus the number of triangles/clock) if they wanted, but they just didn't care. This reinforces my belief that they are basically giving up on gaming. If they were taking gaming seriously, they would have increased the number of compute engines to 8 and the number of ROPs to 128. They had plenty of room to do this while still fitting within the reticle limit, and it would have prevented disgraceful results like this one:
Instead, we get a card that basically retains almost all of the potential bottlenecks of Fiji. In a handful of cases, the drivers can try to use tricks to reduce the amount of geometry rendered, but when these optimizations are not done or when the geometry actually does have to be drawn, then Vega is no better than Fiji.
I don't expect anything worthwhile in gaming for Navi, either. They aren't going to bother to make two GPUs act transparently as one - they'll expect the developers to do it with explicit multi-adapter in DX12. The developers won't; instead, they will continue to write easier-to-code DX11 titles that Nvidia's driver team will optimize to the point that they are as fast as DX12. Navi will still be limited to four compute engines and 64 ROPs because RTG simply doesn't care to do any better.
How old games count?I agree with you.
Where are the 100+ games running 4k @ 60+ FPS? I can find 2/3...