NVIDIA’s GeForce GTX Titan Review, Part 2: Titan's Performance Unveiled
by Ryan Smith & Rahul Garg on February 21, 2013 9:00 AM ESTSynthetics
As always we’ll also take a quick look at synthetic performance to get a better look at Titan’s underpinnings. 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 a mix of a ROP test and a test to see if you have enough bandwidth to feed those ROPs. At the same time the smallest increase in theoretical performance for Titan over GTX 680 was in ROP performance, where a 50% increase in ROPs was met with a minor clockspeed reduction for a final increase in ROP performance of 25%.
The end result is that with gains of 28%, Titan’s lead over GTX 680 is just a hair more than its increase in theoretical ROP performance. Consequently at first glance it looks like Titan has enough memory and cache bandwidth to feed its 48 ROPs, which given where we’re at today with GDDR5 is good news as GDDR5 has very nearly run out of room.
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.
Oddly enough, despite the fact that Titan’s texture performance improvements over GTX 680 were only on the range of 46%, here Titan is measured as having 62% more texturing performance. This may be how Titan is interplaying with its improved bandwidth, or it may be a case where some of the ancillary changes NVIDIA made to the texture paths for compute are somehow also beneficial to proper texturing performance.
Finally we’ll take a quick look at tessellation performance with TessMark.
Unsurprisingly, Titan is well ahead of anything else NVIDIA produces. At 49% faster it’s just a bit over the 46% theoretical performance improvement we would expect from the increased number of Polymorph Engines the extra 6 SMXes bring. Interestingly, as fast as GTX 580’s tessellation performance was, these results would indicate that Titan offers more than a generational jump in tessellation performance, nearly tripling GTX 580’s tessellation performance. Though at this time it’s not at all clear just what such tessellation performance is good for, as we seem to be reaching increasingly ridiculous levels.
Facebook
Twitter
RSS
337 Comments
View All Comments
CeriseCogburn - Saturday, February 23, 2013 - link
Stop whining along with the rest of them, grow a set, get a job, and buy two of them.Might do you some good.
Alucard291 - Sunday, February 24, 2013 - link
Unlike you, I have a job :)chizow - Sunday, February 24, 2013 - link
Good point, I'd tend to agree with that assessment as anyone who actually works for their money would not be so eager to part with it so quickly in $1K denominations for what amounts to a glorified pinball machine.He's probably a kid who has never had to work a day in his life or a basement dweller who has no hope of ever buying one of these anyways.
CeriseCogburn - Sunday, February 24, 2013 - link
And now with the pure troll, the lying idiot conspiracist nVidia hater takes on the pure personal attack for a big fat ZERO score.Congratulations, you and your pure troll can high five each other and both be wrong anyway, for another yer or two, or the rest of your whining crybaby posting PC herd idiot mentality lives.
Alucard291 - Friday, March 8, 2013 - link
No no kid. You're the "pure troll here"So yeah go get a job and buy two of them. For yourself. Stop being angry at us for not being able to afford it
~lol~
wiyosaya - Thursday, February 21, 2013 - link
While I understand your frustrations, IMHO, this is a card aimed at those wanting the compute performance of a Tesla at 1/3 the cost. As I see it, nVidia shot themselves in the foot for compute performance with the 680, and as such, I bet that 680 sales were less than expected primarily because of its crappy compute performance in comparison to say even a 580. This may have been their strategy, though, as they might have expected $3,500 Teslas to fly off the shelf.I am also willing to bet that Teslas did not fly off the shelf, and that in order to maintain good sales, they have basically dropped the price of the first GK110s to something that is reasonable with this card. Once can now buy 3.5 Titan's for the price of the entry level GK110 Tesla, and I highly expect nVidia to make a profit rather than the killing that they might have thought possible on the GK110 Teslas.
That said, I bet that nVidia gets a sht load of orders for this card from new HPC builders and serious CAD/CAE workstation suppliers. Many CAD/CAE software packages like SolidWorks and Maple support GPGPUs in their code making this card a bargain for their builds.
My apologies, to all the gamers here but us compute nerds are drooling over this card. I only wish I could afford one to put in my i7-3820 build from July. It is more than 2x what I paid for a 580 back then, and the 580 buy was for its compute performance.
atlr - Thursday, February 21, 2013 - link
wiyosaya, I am trying to come up to speed on comparing compute performance between Nvidia and AMD options. Is the Titan drool-worthy only for software that only uses CUDA and not OpenCL? This reminds me of the days of Glide versus OpenGL APIs.trajan2448 - Friday, February 22, 2013 - link
AMDs fps numbers are overstated. They figured out a trick to make runt frames, or frames which are not actually rendered to trigger the fps monitor as a real fully rendered frame. This is real problem for AMD much worse than the latency problem. Crossfire is a disaster which is why numerous reviewers including Tech Report have written that Crossfire produces higher fps but feels less smooth than Nvidia.Check this article out. http://www.pcper.com/reviews/Graphics-Cards/NVIDIA...
chizow - Saturday, February 23, 2013 - link
That's an awesome analysis by PCPer, thanks for linking that. Might end up being the biggest driver cheat scandal in history. Runt framesgate lol.CeriseCogburn - Saturday, February 23, 2013 - link
HUGE amd cheat.It's their standard operating procedure.
The fanboys will tape their mouths, gouge out their eyes and stick fingers in their ears and chant: "I'm not listening".