The NVIDIA GeForce GTX 1080 & GTX 1070 Founders Editions Review: Kicking Off the FinFET Generation
by Ryan Smith on July 20, 2016 8:45 AM ESTAshes of the Singularity
Sorely missing from our benchmark suite for quite some time have been RTSes, which don’t enjoy quite the popularity they once did. As a result Ashes holds a special place in our hearts, and that’s before we talk about the technical aspects. Based on developer Oxide Games’ Nitrous Engine, Ashes has been designed from the ground up for low-level APIs like DirectX 12. As a result of all of the games in our benchmark suite, this is the game making the best use of DirectX 12’s various features, from asynchronous compute to multi-threadeded work submission and high batch counts. What we see can’t be extrapolated to all DirectX 12 games, but it gives us a very interesting look at what we might expect in the future.
Once again the top spot is uncontested by the GTX 1080. However after that, things become more interesting. On the whole, Ashes is a game that favors AMD GPU over NVIDIA GPUs, and as a result the GTX 1070 does not get to lock in second place. Rather that goes to the last generation Fury X. AMD designs are very ALU-heavy, and I suspect Ashes is capable of putting those ALUs to good use, something most other games struggle with. That said, if we normalized this for price or power consumption, then the Pascal cards would be well in the lead, but it does show that on an absolute basis, GTX 1070 isn’t going to outrun the best of the last-gen cards all the time.
Meanwhile it’s interesting to note that one of the more unusual aspects of the engine behind Ashes is that it’s relatively resolution insensitive. That is, performance only drops moderately as we increase the resolution. This means that we need a GTX 1070 to sustain better than 60fps at 1080p, but that same card is still getting better than 40fps at 4K, a resolution with 4x the pixels.
Finally, looking at our NVIDIA cards on a generational basis, even without their commanding lead, the two Pascal cards show the expected generational gains. GTX 1080 improves on GTX 980 by between 65% and 70%, and GTX 1070 improves on GTX 970 by between 53% and 58%.
Facebook
Twitter
RSS
200 Comments
View All Comments
patrickjp93 - Wednesday, July 20, 2016 - link
That doesn't actually support your point...Scali - Wednesday, July 20, 2016 - link
Did I read a different article?Because the article that I read said that the 'holes' would be pretty similar on Maxwell v2 and Pascal, given that they have very similar architectures. However, Pascal is more efficient at filling the holes with its dynamic repartitioning.
mr.techguru - Wednesday, July 20, 2016 - link
Just Ordered the MSI GeForce GTX 1070 Gaming X , way better than 1060 / 480. NVidia Nail it :)tipoo - Wednesday, July 20, 2016 - link
" NVIDIA tells us that it can be done in under 100us (0.1ms), or about 170,000 clock cycles."Is my understanding right that Polaris, and I think even earlier with late GCN parts, could seamlessly interleave per-clock? So 170,000 times faster than Pascal in clock cycles (less in total time, but still above 100,000 times faster)?
Scali - Wednesday, July 20, 2016 - link
That seems highly unlikely. Switching to another task is going to take some time, because you also need to switch all the registers, buffers, caches need to be re-filled etc.The only way to avoid most of that is to duplicate the whole register file, like HyperThreading does. That's doable on an x86 CPU, but a GPU has way more registers.
Besides, as we can see, nVidia's approach is fast enough in practice. Why throw tons of silicon on making context switching faster than it needs to be? You want to avoid context switches as much as possible anyway.
Sadly AMD doesn't seem to go into any detail, but I'm pretty sure it's going to be in the same ballpark.
My guess is that what AMD calls an 'ACE' is actually very similar to the SMs and their command queues on the Pascal side.
Ryan Smith - Wednesday, July 20, 2016 - link
Task switching is separate from interleaving. Interleaving takes place on all GPUs as a basic form of latency hiding (GPUs are very high latency).The big difference is that interleaving uses different threads from the same task; task switching by its very nature loads up another task entirely.
Scali - Thursday, July 21, 2016 - link
After re-reading AMD's asynchronous shader PDF, it seems that AMD also speaks of 'interleaving' when they switch a graphics CU to a compute task after the graphics task has completed. So 'interleaving' at task level, rather than at instruction level.Which would be pretty much the same as NVidia's Dynamic Load Balancing in Pascal.
eddman - Thursday, July 21, 2016 - link
The more I read about async computing in Polaris and Pascal, the more I realize that the implementations are not much different.As Ryan pointed out, it seems that the reason that Polaris, and GCN as a whole, benefit more from async is the architecture of the GPU itself, being wider and having more ALUs.
Nonetheless, I'm sure we're still going to see comments like "Polaris does async in hardware. Pascal is hopeless with its software async hack".
Matt Doyle - Wednesday, July 20, 2016 - link
Typo in the lead sentence of HPC vs. Consumer: Divergence paragraph: "Pascal in an architecture that...""is" instead of "in"
Matt Doyle - Wednesday, July 20, 2016 - link
Feeding Pascal page, "GDDR5X uses a 16n prefetch, which is twice the size of GDDR5’s 8n prefect."Prefect = prefetch