• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

**THE NVIDIA DRIVERS THREAD**

Those frame latency times are superb Paul. What the hell is PCper doing with their testing methodology? They have it all over the place lol.

Maybe Ryan got his 295x2 in place of a favourable review and this is probably why he has avoided my question on why Grid 2 and 780Ti's only got 75 fps average in his tests and yet my stock system (everything at stock) scored 88 fps. I assume the 780Ti's are faster than Titans as well.
 
7928f1a59016c3d4029046ddff039f63.jpg


So yer, I looked at another review and scratched my head again at the results. They ran the test at 4K resolution with all settings maxed. The 295x2 gets a very respectable 46 fps but the 780Ti SLI only gets 43 average :( I thought something was up and decided to see how my Titans got on in this bench...

No overclocks on anything and all settings maxed, I got 48.6 (I think) average and recorded the bench on my phone. I will edit this post to add the vid here (poor phone recording sorry) but you can clearly see that the 780Ti's results are massively down from what should be higher than my results.

It is getting impossible to trust any review sites, as looking about the web, so many results look off.

Edit:

My recording and frames of the Thief bench and nice to know that with my whole computer at stock, it is faster than R9 295x2 with Mantle in Thief :cool:


46 fps for the 295x2 and 48.7 fps for the SLI Titans (Titans 6% faster) :cool:
 
Last edited:
I think reviews on both sides seem to be off for most stuff:(
Its annoying and often makes them completely useless

You could generally take review sites with a pinch of salt and I expect to see a bit of bias shown to the Manufacturer giving them a GPU for free to test but recently it isn't a pinch of salt needed but a friggin truck load. Bench's are so far off you might as well not bother reading them.
 
First up...Tomb Raider! Benched using the OcUK bench thread settings on single and dual 780's @ 1110/1600 (back on stock bios!)

Single card..

TR_SG_FT.jpg

2 Way SLI

TR_SLI_FT.jpg

You'll noticed single card has a higher overall latency of ~13ms, with SLI sat at ~7ms.
BF4, frametimes taken from my 335/337 comparisons, both dual card, will do single card 337 shortly.

335

BF_FT_335.jpg

337

BF4_FT_337.jpg

Latency is slightly down but its that minor its in the margin of error.

Single card to follow, if you can make sense of the above, please do.

Edit: Single card:

BF_SG_FT_337.jpg

I must stress single card felt insanely smooth while averaging a healthy 98fps.

Where did you get that frame tool PG?
 
I've not updated the drivers for my 770 since November, are there any particular improvements to expect with BF series or the Metro games?

These new wonder drivers seem to be aimed more so at SLI users but there have been some big gains since November in BF4 and I expect (don't know officially) that Metro LL has some further optimisations.

Worth installing these new Beta's and if they are not good for you, roll back to the one's you found worked well.
 
Threw the new drivers on. Not run any benchmarks yet and only tried BF4 which I lock to 120FPS.

Not had any problems, and hard to tell if smoothness is placebo or a result of the new driver giving more consistent 120FPS lockout.

Might run some benchmarks when I get the chance, if I can be arsed :D
 
Ok I did some more frame time testing at 1440p on BF4, won't upload screenies yet, as, well, I can't be bothered cropping and uploading them just now.

BUT

Single card = 17ms, wavely line things all over the shop

Dual card = 8ms, line almost perfectly straight.

So while the PCPer graphs posted earlier DO have some truth to them, it only extends as far as single cards, I would say Ryan is having SLI issues of another kind if his frame times are that far out.
 
Ok I did some more frame time testing at 1440p on BF4, won't upload screenies yet, as, well, I can't be bothered cropping and uploading them just now.

BUT

Single card = 17ms, wavely line things all over the shop

Dual card = 8ms, line almost perfectly straight.

So while the PCPer graphs posted earlier DO have some truth to them, it only extends as far as single cards, I would say Ryan is having SLI issues of another kind if his frame times are that far out.
Aah, thats why you said no shooting then, bit o testing yourself.:D

Btw mate, whats the program your using to test with?
 
FRAPS, check the box to record frame times, then open them up with FLACalculator, click the little graph icons to bring those up.

FRAPS is purely is can accurately time the run with its time limit cut off, otherwise for round recording in BF4 I use the perfoverlay.framefilecapture 1 thing.

Oh and a single 780 full ultra @ 1440p holds up very very well, frame range in shanghai was ~50-80fps
 
Back
Top Bottom