Compute

Shifting gears, let’s take a look at compute performance on GTX 1060.

As we already had the chance to categorize the Pascal architecture’s compute performance in our GTX 1080 review, there shouldn’t be any surprises here. But it will be interesting to see whether the GTX 1060’s higher ratio of memory bandwidth per FLOP materially impacts overall compute performance.

Starting us off for our look at compute is LuxMark3.1, the latest version of the official benchmark of LuxRender. LuxRender’s GPU-accelerated rendering mode is an OpenCL based ray tracer that forms a part of the larger LuxRender suite. Ray tracing has become a stronghold for GPUs in recent years as ray tracing maps well to GPU pipelines, allowing artists to render scenes much more quickly than with CPUs alone.

Compute: LuxMark 3.1 - Hotel

While GTX 1060 could hang with GTX 980 in gaming benchmarks, we don’t start off the same way with compute benchmarks, with the last-generation flagship holding about 17% ahead. Unfortunately for NVIDIA, this is about where GTX 1060 needed to be to best RX 480; instead it ends up trailing the AMD competition. Otherwise the performance gain versus the GTX 960 stands at 65%.

For our second set of compute benchmarks we have CompuBench 1.5, the successor to CLBenchmark. CompuBench offers a wide array of different practical compute workloads, and we’ve decided to focus on face detection, optical flow modeling, and particle simulations.

Compute: CompuBench 1.5 - Face Detection

Compute: CompuBench 1.5 - Optical Flow

Compute: CompuBench 1.5 - Particle Simulation 64K

Like with GTX 1080, relative performance is all over the place. GTX 1060 wins with face detection, loses at optical flow, and wins again at particle simulation. Even the gains versus GTX 960 are a bit more uneven, though at the end of the day GTX 1060 ends up being significantly faster than its predecessor with all 3 sub-benchmarks.

Moving on, our 3rd compute benchmark is the next generation release of FAHBench, the official Folding @ Home benchmark. Folding @ Home is the popular Stanford-backed research and distributed computing initiative that has work distributed to millions of volunteer computers over the internet, each of which is responsible for a tiny slice of a protein folding simulation. FAHBench can test both single precision and double precision floating point performance, with single precision being the most useful metric for most consumer cards due to their low double precision performance. Each precision has two modes, explicit and implicit, the difference being whether water atoms are included in the simulation, which adds quite a bit of work and overhead. This is another OpenCL test, utilizing the OpenCL path for FAHCore 21.

Compute: Folding @ Home Single Precision

Compute: Folding @ Home Double Precision

Finally, in Folding@Home, we see the usual split between single precision and double precision performance. GTX 1060 is solidly in the lead when using FP32, but NVIDIA’s poor FP64 rate means that if double precision is needed, RX 480 will pull ahead.

Hitman Synthetics
Comments Locked

189 Comments

View All Comments

  • Younanomous - Monday, August 8, 2016 - link

    So would it be a good investment to go for the 1070 for 1080p if I plan t keep the card for 3+ years? I'm always told that the X70's are overill for 1080p.
  • JamesAnthony - Monday, August 8, 2016 - link

    Thanks for another Excellent review.
    The review shows the GTX 1060 is an excellent value/performance card.
    While there is a bunch of arguing back and forth I think it's fairly clear to say that:

    a. There is a huge installed base of people with monitors running 1920x1080

    b. The GTX1060 offers excellent performance from a price / performance standpoint

    c. For most intents and purposes the GTX 1060 and RX 480 8gb have similar performance under most applications and give or take a bit are similar in price & the prices will drift down a bit as supply finally fully catches up with demand. (I was able to land 2 of the Overclocked Zotac ones at $275 each on launch day).

    d. If you have a lot of brand loyalty (in my case Nvidia drivers just work better for me), you'll pick based on your preference, and then people who truly don't care (I think there are less of those) can pick based on specific things they want, but neither is a bad choice unless you have to have FP64 then it's AMD only.

    The RX 460 is going to be quite interesting from the extreme cheap budget / power capped viewpoint if it comes in around the $120 range & has a 75w power draw that doesn't need a power connector.

    I'd love to see a GTX 1050 that was PCIe power only and could go standard single slot or low profile single / dual slot to compete. (with it being able to go single slot, low profile as the ultimate idea).
  • loop - Monday, August 8, 2016 - link

    buy rx 480 or gtx 1060?
  • elessar25 - Monday, August 8, 2016 - link

    I read on Eurogamer that AMD's GPUs outperform NVIDIA on "close to metal" APIs like DX12. Although this review proofs without a doubt that the GTX 1060 is more powerful overall than the RX 480, I'm curious if this will hold true going forward? For someone who wants to future-proof their system, which is the better GPU? Disclaimer: I'm not a fanboy and I currently run a GTX 670.
  • Greeba77 - Tuesday, August 9, 2016 - link

    This is pretty much what I was getting at in my above comment, curious to know...
  • Tech-Curious - Wednesday, August 10, 2016 - link

    We don't know, but it seems reasonable to guess, based on recent history, that:

    1.) AMD's relative performance will likely increase, whereas nVidia's will probably decrease with time, as a result of driver optimizations for the former and the lack thereof for the latter. Keep in mind that this isn't necessarily an indictment of nVidia or a compliment to AMD; it could be interpreted as nVidia's optimizing their drivers better at the outset, leaving them less headroom to improve. But for whatever reason, past-gen nVidia cards seem to fall back relative to their competition later on.

    2.) AMD will tend to gain more from DX 12. Whether "gaining more" means that any particular AMD card will outperform its nVidia counterpart is a whole 'nother issue.
  • Hrel - Tuesday, August 9, 2016 - link

    Wow, based on these numbers AMD needs to drop their prices at least $40 to even be viable, let's not even talk about competitive.

    Fuck, so now we officially have a monopoly in the CPU and GPU space.
  • Younanomous - Wednesday, August 10, 2016 - link

    It's not a monopoly if you have another choice, even if that other choice isn't as good in your eyes.
  • Casecutter - Wednesday, August 10, 2016 - link

    Would've been nice to see where a $249 unit (with lower end dual fans) slots into this line up. This Asus doesn't really impress for $315 (working a $15 rebate right today).
  • thunderwave_2 - Wednesday, August 10, 2016 - link

    For the upcoming RX 480 review, could you post power, temps and noise results for the Compatibility mode? (they perform almost identically, so performance results might be redundant)

Log in

Don't have an account? Sign up now