The NVIDIA GeForce RTX 2070 Founders Edition Review: Mid-Range Turing, High-End Price
by Nate Oh on October 16, 2018 9:00 AM ESTCompute & Synthetics
Shifting gears, we'll look at the compute and synthetic aspects of the RTX 2070. Though it has its own GPU in the form of TU106, the hardware resources at hand are similar in progression to what we've seen in TU102 and TU104.
Starting off with GEMM tests, the RTX 2070's tensor cores are pulled into action with half-precision matrix multiplication, though using binaries originally compiled for Volta. Because Turing is backwards compatible and in the same compute capability family as Volta (sm_75 compared to Volta's sm_70), the benchmark continues to work out-of-the-box, though without any Turing optimizations.
At reference specifications, peak theoretical tensor throughput is around 107.6 TFLOPS for the RTX 2080 Ti, 80.5 TFLOPS for the RTX 2080, and 59.7 TFLOPS for the RTX 2070. Unlike the 89% efficiency with the Titan V's 97.5 TFLOPS, the RTX cards are essentially at half that level, with around 47%, 48%, and 45% efficiency for the RTX 2080 Ti, 2080, and 2070 respectively. A Turing-optimized binary should bring that up, though it is possible that the GeForce RTX cards may not be designed for efficient tensor FP16 operations as opposed to the INT dot-product acceleration. After all, the GeForce RTX cards are for consumers and ostensibly intended for inferencing rather than training, which is the reasoning for the new INT support in Turing tensor cores.
In terms of SGEMM efficiency though, the RTX 2070 is hitting a ridiculous 97% of its touted 7.5 TFLOPS, though to be fair the reference specifications here are done manually rather with a reference vBIOS. The other two GeForce RTX cards are at similar 90+% levels of efficiency, though a GEMM test like this is specifically designed for maximum utilization.
The breakdown of the GB4 subscores seems to reveal a similar uplift like we spotted with the Titan V, which had scored in excess of 509,000 points. We'll have to investigate further but Turing and Volta are clearly accelerating some of these workloads beyond what was capable in Pascal and Maxwell.
Given that TU106 has 75% of the hardware resources of TU104, the tessellation performance is in line with expectrations. For reference, we noted earlier that the Titan V scored 703 while the Titan Xp scored 604.
121 Comments
View All Comments
beisat - Tuesday, October 16, 2018 - link
Thanks for the review, nice as always.Was hoping to upgrade my 970 before Turing was announced, but I feel like I'm getting ripped of with these cards. The review did nothing to change that feeling, but that was to be expected.
Luke212 - Tuesday, October 16, 2018 - link
Please investigate why Turing is slower than Volta for HGEMM. If it was using the tensor cores they should be not that slow.SMOGZINN - Tuesday, October 16, 2018 - link
On the "The Test" page you show that the "NVIDIA GeForce GTX 1070 Ti Founders Edition" is one of the cards being compared, but it does not show up in the benches.Targon - Tuesday, October 16, 2018 - link
From the information, seeing Vega 64 going up to a temp of 86C would put it into thermal throttle range, which would cripple performance. From my own experience, manually adjusting the fan settings in Global Wattman to go up to 4500rpm and with a temperature target of 75C will avoid the throttle issues in the first place and also improving performance significantly, even without tweaking clock speeds or voltages.So, if Vega 64 is getting throttled and still hitting the numbers reported, that implies that with the fan profile adjusted as I suggested, we would be seeing Vega 64 doing a bit better in terms of framerates.
The_Assimilator - Tuesday, October 16, 2018 - link
Let's be honest: Vega isn't here for competition purposes, it's just included as a courtesy.atl - Tuesday, October 16, 2018 - link
Would be good to have some SLI & Cryptocurrency benchmarks includedTEAMSWITCHER - Tuesday, October 16, 2018 - link
These RTX cards are going to be a fantastic value......next summer when they drop the prices.
eva02langley - Tuesday, October 16, 2018 - link
Even there, I don`t know if Navi can really be a 250$ GPU with 1080 GTX performances.sandman74 - Tuesday, October 16, 2018 - link
980 owner here gaming at 1440p. Really wanted to upgrade but when I cost everything up, PC gaming has suddenly become a very expensive hobby.Decided to completely abandon the PC as a future gaming platform mostly thanks to the pricing of the new gpu cards.
2.5yrs since the 1080 for barely better performance. RTX isn’t viable on this card. My own view is the new line up sucks.
Practically all my mates are on consoles these days which is a shame but it’s a sign of the times. Tried the BF5 beta on my xbox one S and was blown away at how decent it was. Had real fun playing with friends which is what matters.
So I can only imagine it’s even better on the Xbox One X which you can buy for the price of just this GPU.
Prices have gone insane, so I’m stepping out. Total respect for those that can justify the prices and carry on PC gaming. I can’t.
The_Assimilator - Tuesday, October 16, 2018 - link
tl;dr rather get a heavily discounted 1080 Ti (which will probably be factory overclocked and have a beefier cooler).