NVIDIA GeForce GTX 780 Review: The New High End
by Ryan Smith on May 23, 2013 9:00 AM ESTSynthetics
As always we’ll also take a quick look at synthetic performance, though as GTX 780 is just another GK110 card, there shouldn't be any surprises here. These tests are mostly for comparing cards from within a manufacturer, as opposed to directly comparing AMD and NVIDIA cards. We’ll start with 3DMark Vantage’s Pixel Fill test.
Pixel fill is traditionally bound by ROP and memory throughput, but with enough of both the bottleneck can shift back to the shader blocks. In this case that’s exactly what happens, with the GTX 780 trailing GTX Titan by about the theoretical difference between the two cards. On the other hand it’s very odd to see the GTX 680 get so close to the GTX 780 in this test, given the fact that the latter is more powerful in virtually every way possible.
Moving on, we have our 3DMark Vantage texture fillrate test, which does for texels and texture mapping units what the previous test does for ROPs.
Unlike pixel fill, texel fill is right where we expected it to come in compared to cards both above and below the GTX 690.
Finally we’ll take a quick look at tessellation performance with TessMark.
NVIDIA’s tessellation performance is strongly coupled to their SMX count, so the high number of SMXes (12) on the GTX 780 helps it keep well ahead of the pack. In fact we’re a bit surprised it didn’t fall behind GTX Titan by more than what we’re seeing. On the other hand the lead over the GTX 580 is right where we’d expect it to be, showcasing the roughly trebled geometry performance of GTX 780 over GTX 580.
155 Comments
View All Comments
aidivn - Thursday, May 23, 2013 - link
so, how many Double Precision units are there in each SMX unit of gtx780? titan had 64 dp units in each of their SMX units which totaled to 896 dp unitsAnd can u turn them on or off from the forcewre driver menu like “CUDA – Double Precision” for gtx780?
Ryan Smith - Thursday, May 23, 2013 - link
Hardware wise this is GK110, so the 64 DP units are there. But most of them would be disabled to get the 1/24 FP64 rate.aidivn - Friday, May 24, 2013 - link
so how many are disabled and how many are enabled (numbers please)?Ryan Smith - Friday, May 24, 2013 - link
You would have only 1/8th enabled. So 8 per SMX are enabled, while the other 56 are disabled.aidivn - Saturday, May 25, 2013 - link
so, the GTX780 only has 96 DP units enabled while the GTX TITAN has 896 DP units enabled...thats a huge cut on double precisionDanNeely - Sunday, May 26, 2013 - link
That surprised me too. Previously the cards based on the G*100/110 cards were 1/8; this is a major hit vs the 580/480/280 series cards.Old_Fogie_Late_Bloomer - Thursday, May 23, 2013 - link
"GTX 780 on the other hand is a pure gaming/consumer part like the rest of the GeForce lineup, meaning NVIDIA has stripped it of Titan’s marquee compute feature: uncapped double precision (FP64) performance. As a result GTX 780 can offer 90% of GTX Titan’s gaming performance, but it can only offer a fraction of GTX Titan’s FP64 compute performance, topping out at 1/24th FP32 performance rather than 1/3rd like Titan."Seriously, this is just...it's asinine. Utterly asinine.
tipoo - Thursday, May 23, 2013 - link
Market segmentation is nothing new. The Titan really is a steal if you need DP, the next card up is 2400 dollars.Old_Fogie_Late_Bloomer - Thursday, May 23, 2013 - link
I'm well aware of the existence of market segmentation, but this is just ridiculous. Putting ECC RAM on professional cards is segmentation. Disabling otherwise functional features of hardware, most likely in the software drivers...that's just...ugh.SymphonyX7 - Thursday, May 23, 2013 - link
I just noticed that the Radeon HD 7970 Ghz Edition has been trouncing the GTX 680 in most of the benchmarks and trailing the GTX 680 in those benchmarks that traditionally favored Kepler. What the heck just happened? Didn't the review of the Radeon HD 7970 Ghz Edition say that it was basically tied with the GTX 680?