Thursday, December 13th 2012
HD 7950 May Give Higher Framerates, but GTX 660 Ti Still Smoother: Report
The TechReport, which adds latency-based testing in its VGA reviews, concluded in a recent retrospective review taking into account recent driver advancements, that Radeon HD 7950, despite yielding higher frame-rates than GeForce GTX 660 Ti, has higher latencies (time it takes to beam generated frames onto the display), resulting in micro-stutter. In response to the comments-drama that ensued, its reviewer did a side-by-side recording of a scene from "TESV: Skyrim" as rendered by the two graphics cards, and slowed them down with high-speed recording, at 120 FPS, and 240 FPS. In slow-motion, micro-stuttering on the Radeon HD 7950 is more apparent than on the GeForce GTX 660 Ti.
Find the slow-motion captures after the break.
Source:
The TechReport
122 Comments on HD 7950 May Give Higher Framerates, but GTX 660 Ti Still Smoother: Report
The testing system. techreport.com/review/23981/radeon-hd-7950-vs-geforce-gtx-660-ti-revisited/2
The 7950 was better in sleeping dogs and medal of honor: Warfighter in their tests for latency. So it is not an always situation but in most cases.
It does sound like driver optimization to me more than anything else, but it could be less a case of "amd is faster but their drivers are worse" and more a case of "amd optimizes more for speed while nv optimizes more for less stutter" or something along those lines. Maybe there's even some hidden hitch in the CPU code for the drivers that is holding everything back.
Any particular reason why the AMD card drops to zero latency for the last few hundred frames of the chart, but the nV one doesn't?
This is skewed towards Nvidia to make them look better.
And since you're quite insistent that there is an Nvidia bias, why not read this page about Sleeping Dogs, an "AMD Gaming Evolved" title?
AMD IS BETTER!!! NVIDIA SUCKS!!!
Not much difference, it's just a tie...
Oh...THAT'S NVIDIA'S FAULT!!!Video card reviews should just focus on framerates? Testing for latency is embarrassing?
from what i remember a 7950 clocked up is faster than a 7950 boost that hits the same clocks speeds do to 0 clock speed fluctuation.
There is more going on in the background that isnt being expanded upon and tested. Essentially it seems TR took the 1 AMD gpu thats had the roughest time out of all AMD products then used a model that brings in the boost variable and its potential problems that represents.
From Hardocp testing, you seem to need more FPS to have the same "smoothness feeling" while gaming VS Nvidia. They spoke about it lots of time on lots of reviews.
I was telling about micro stuttering with SLi/CF years ago but people went mad at me and refused to believe it exists. Same thing with HD5 series broken AF. Etc. etc.
Having said that, given the standing that Scott has in the community, I'd say that someone will test the 7970, 7970GE etc. It wouldn't surprise me if AMD were optimizing for outright speed at all- that is the primary metric used in the majority of reviews, and a major bullet point in PR slides.AMD had no problem reaping the rewards of reviews using the all new 7950 Boost...are you saying that TR should gloss over any potential problem because the bulk of the other models are possibly OK ? HardOCP and Tom's Hardware amongst others are also reporting various issues in single/multi card configs- at what point do you say its worth reporting?
I get micro stutter on Fry Cry 3 with my crossfire 7970 set up, very noticeable when it should be butter smooth. But my friend with a GTX 680 also has graphical issues.
Micro stutter itself is a bit subjective. I've played games with fare worse with micro-stutter on AMD cards (according to Hard OCP) but haven't noticed any effect unlike FC 3 where it is considerably more apparent.
Meh, choose your card and see how you like it. Dishonored played beautifully on my cards. Such an awesomely well planned and coded game.
I moved away from ATI after the issues I had with my 4870X2. I'd really like to see them fix these small issues. Then I'd consider buying them again.
But I thought all of those videos looked weird. Especially the 240FPS ones looked plain wrong. At 120FPS Nvidia looked smoother, but still flaky.
They should have done the gameplay video comparison with Sleeping Dogs and not Skyrim, so that there won't be any "BUT WE ALL KNOW THERE ARE PROBLEMS WITH AMD IN THE GAME!!!" because right off the bat a big "AMD Gaming Evolved" logo will appear once you run the game. And if they still insist that Sleeping Dogs is "skewed towards Nvidia" then that also implies that "AMD Gaming Evolved" was pointless, unless they then say that United Front Games intentionally made AMD perform worse in their game...
Actually yeah, this is very subjective. The AMD card craps out as the video goes on and the foliage gets closer to the camera view but to me the 660 starts out with more noticeable jerkiness between frames.
Having watched this, I don't think it's micro stutter I've got on FC3. Something more buggy methinks.
It is like cutting your hand off when it hurts.