• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

FRONTPAGE Our Take On NVIDIA's FCAT GPU Testing

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Overclockers.com

Member
Joined
Nov 1, 1998
If you follow the GPU reviewing world, you have probably seen several articles pop up regarding a new GPU testing method using a tool called FCAT. Our editing team has discussed this extensively and we believe we've come to the right decision, but are of course willing to listen to our readers. Please have a look at this article, weigh the options and chime in with your two cents.

... Return to article to continue reading.
 
I agree that while the difference may not be noticeable or even relevant to most gamers, it still is an issue that needs to be addressed by the manufacturers and testers. If this wasn't brought up, CrossFireX would run for years the same way, possibly even getting worse, while claiming the FPS is great.

The attention it is getting is due to the fact that people have been talking about microstutter for years with SLI/CFX setups but did not have a proven metric to refer to, just anecdotal evidence. A few years later, we now have (albeit some expensive) hardware that actually allows us to measure and report this accurately, the very same issue that has plagued the few who could afford the high end setups. I for one am glad that we have come this far in this issue. While this may be a small or negligible issue for most, small or negligible issues will either grow or multiply in the future if not addressed. It could also one of those issues that you don't notice until you see it like TN vs IPS (or trying on audio headphones of different quality).

With that said, I can totally understand and relate the overclockers.com editing staff to not research this issue themselves. It takes a lot of testing, analysis, and time to research and document this issue, and on an all volunteer force it's just not feasible. I do appreciate you guys bringing up this topic though on the front page.
 
Can't argue with that at all cullam3n; it's great this was brought to everyone's attention. Thankfully, now all that remains is for AMD to fix their drivers (which they are working on for a mid-summer release I'm told by them directly).
 
Another thing that many of the "reviews" fail to mention that almost all issues with microstutter can be remedied via software implementation such as RadeonPro.

Although it is still an issue that needs to be addressed on a driver level there are still plenty of ways to get around this whole misconception that micro-stuttering is a problem that can't be alleviated.

Great article :thup:
 
i find the micro Stuttering more visible that the fps difference between 80fps and 110 fps. so i personally endorse it.
 
Last edited:
this is new and interesting to me.

reading all those references next. looks like AMD doesn't like this cause it clearly puts nvidia in front only. But I suppose I need to see for myself..

are there any videos comparing this effect mentioned?
 
Looks llike a good way to test.:cool::popcorn:

well, the result tells one thing. Crossfire is broken til they fix it.
it is a fact for now, that crossfire 'doesn't help' animation, in fact, makes it a little worse by inserting grunts in the animation sequence. My lab mates have talked about this since a long time ago, and it is one of the reason why my machine is a GTX680 3xSLI + 7970 and not the other way around.

I was able to observe the stuttering myself, and was able to see a clear difference in graphics properties. the stutter cannot be 'observed' so to say, but it results in the overall animation being a lot more 'fake'. Not sure how to say it, but when you have 2 machines side by side, one is crossfire, one is in SLI, and I was asked 'which feels better?', I was able to call it out 100% of the time.

I have 6 associates in the lab, and of the 7 of us, 2 see no difference, so I must say this is a very personal 'perception' thing. :)

for AMD cards, for now, 1 GPU is about all you need, whatever FPS you get extra with it with crossfire, actually harms animation. (great for benching though! just not for the eye)

Here's a video explaining.. I found this recently. :)

 
Aye, that video was linked from our article.

I agree it's a very good thing that this came out. It discovered an issue and AMD is already on an Alpha driver that fixes it (with good results).

If this piece came across as indicating any of us thought it was a bad thing that people did this, then that was not the intended impression. We're basically saying that AMD is fixing it (almost done fixing it actually) and once that's done, tried and true testing methods are back to being 100% reliable. We're not trying to minimize the importance of the discovery, just basically saying why we feel it's not worth investing in this type of system for our site as we go down the road.
 
Aye, that video was linked from our article.

I agree it's a very good thing that this came out. It discovered an issue and AMD is already on an Alpha driver that fixes it (with good results).

If this piece came across as indicating any of us thought it was a bad thing that people did this, then that was not the intended impression. We're basically saying that AMD is fixing it (almost done fixing it actually) and once that's done, tried and true testing methods are back to being 100% reliable. We're not trying to minimize the importance of the discovery, just basically saying why we feel it's not worth investing in this type of system for our site as we go down the road.

:thup:

yup, as soon as AMD fixed CFX then this problem is outta the window!

lol, didn't know the video is linked in the article already. thanks for pointing out!
 
wait.

How is this not a big deal?

the above shows that AMD CrossFire is a 'scam' with increased FrameRates being no more than an artifact.

For all we know, until CrossFire is 'proven fixed' (which we have no first hand evidence) AMD crossfire technology actually 'doesn't exist'. While AMD say they have Alpha drivers to fix it, it is all 'words' this moment. Anyone that invested in Crossfire technology is actually getting 'no better' frame sequencing than a single card, and ALL benching results on Crossfire is also 'a scam', for all we know, AMD could be inserting runt frames everywhere to increase "FPS" on their products. (meaning, all "benching results" from AMD cards can have a GOOD portion of their results being artifacts")... I call that a "major scam".

I mean, a decent driver programmer can 'insert runt frames' at near no loss of GPU performance, and the so called 120 FPS is actually a 100FPS with 20% runts.

I am glad this is brought into the light.
The above claim is LEGIT, and should be known to all. While the test is too expensive, and might not be worth investing into, but until AMD -fixes- this, someone should monitor this, and warns all consumers from investing into AMD crossfire technology.

doing this will make this website more important to people who cares what they money goes into.

for example, the 7990 as of right now, is a outright highway robbery.
 
Last edited:
Did you click the link just two posts ago showing FCAT performance on the new drivers?
 
Did you click the link just two posts ago showing FCAT performance on the new drivers?

yes indeed I have. thanks for showing the link.

I am eager to see them post it soon, and not let the matter drop... (they SAY they fixed it, but it is not 'real' on the consumer level yet.)
what I meant previously was.. if this didn't hit the surface, AMD would have never gotten the intention to fix it...

maybe that's why nvidia brought this up.
 
@trueblack.

Don't mind my sig, I am no AMD fanboy (nVidia/Intel either, lol!).

I've experienced xFire once, with a couple of 5830's, one year ago, before getting a GTX 580.

These were no high end cards, but I can say that performance increase was obvious: you can't say it's a scam.

I also hardly noticed any micro stuttering (most was on BF3).

Sure, I am not a big gamer, casual at most, but when I moved from the dual 5830's to the GTX580, I roughly noticed any difference in games (FPS were roughly the same between the 2 setups, let's say 10% more in favour of the nVidia card).
 
Back