|
![]() |
Overclock.net - Overclocking.net > Graphics Cards > NVIDIA | |
SLI indicator problems
|
||
![]() |
|
|
LinkBack | Thread Tools |
|
|
#31 (permalink) | |||||||||||
|
New to Overclock.net
|
thanks for the help and the time/effort.
Sorry about not being clear but when i mean i had physx off, i meant i had it off for both, panel and game settings. However in BMAA, the SLI bar didnt even show up, leading me to believe that I couldnt use that game to do tests for SLI. I wish i can use Vantage but unfortunately, my LCD TV does not have the resolution requirement to run it ![]() **Update:** So i downloaded the clearsky benchmark and here are my results. They are pretty interesting. The first ss is with one card. The second is on SLI. Default settings 1360X658 resolution Enhanced Full Dynamic lighting DX10 No AA
Last edited by youra6 : 2 Weeks Ago at 09:48 PM |
|||||||||||
|
|
|
|
|
#32 (permalink) | ||||||||||||||
|
4.0ghz
![]()
Join Date: Mar 2008
Location: Cali Native in AZ
Posts: 2,577
Rep: 317
![]() ![]() ![]() ![]() Unique Rep: 230
Trader Rating: 1
|
Hmmm ... SLI is worse than single card on Clear Sky? How do the voltages look on each card when you run the bench in SLI mode? Are they both going up more or less together?
Honestly, this resolution is a bit on the low side to be running SLI to begin with ... combine that with not having too fast a cpu when it comes to gaming ... I'm wondering if you aren't seeing CPU bottlenecking on this test as well with these particular settings. Try running it again on extreme, with the AA maxed out, and compare single card with SLI. I think you have the test set up where it's so easy for the gfx card(s) to run it that the CPU is becoming the limiting factor to performance. Honestly, SLI280's is meant for like 1920x1200 or above. And it's definitely best to have at least an e8400@4.0 if you are going to run cards that are that powerful. But if you make the test hard enough on the gfx cards by bumping up the gfx quality, on something like Clear Sky, SLI should differentiate itself even at low resolutions like you're running.
__________________
3dMark06: 26,490 http://service.futuremark.com/compare?3dm06=11655839 Quote:
|
||||||||||||||
|
|
|
|
#33 (permalink) | |||||||||||
|
New to Overclock.net
|
Here are my clear sky at ultra settings @ 4X AA.
I also posted the ss to the voltages during the benchmark. They look normal. The SLI definitely did better (not by a whole lot), although in some tests the SLI had lower max FPS I hope this helps you out. +rep for being the only one helping me!
|
|||||||||||
|
|
|
|
|
#34 (permalink) | ||||||||||||||
|
4.0ghz
![]()
Join Date: Mar 2008
Location: Cali Native in AZ
Posts: 2,577
Rep: 317
![]() ![]() ![]() ![]() Unique Rep: 230
Trader Rating: 1
|
Well, it doesn't help ME out
![]() The voltages do indeed indicate that SLI is working, which is good news. Yeah, see, in order to really benefit from the SLI, you have to get the 'graphics difficulty' level up. Look at the hardest test here, which is the sun shafts test ... You are seeing nearly a 50% increase in avg fps in the SLI mode on that test. That's a respectable amount of scaling you got going on there ![]() The issue here (in terms of why good scaling is hard for you to come by) is the combination of SLI with the lower resolution and a slower CPU ... the architecture of the original phenom's makes them honestly not the best gaming chips. And hardly any games (esp. not stalker!) use >2 cores, so the quad aspect doesn't really help your situation. And SLI ups the cpu overhead as well. Honestly, with that CPU and particularly at such a low resolution ... a large percentage of the time you're not going to see much improvement with SLI. If you make the test hard enough (like you did with the stalker Sun Shafts test, or if you were to run Crysis on all Very High with 4xAA) you will see the gains, but on the majority of tests ... you just won't. Many games could even be worse with SLI on, to be honest. The setup is not really optimal for good scaling is the bottom line. U needz moar pixels!!! One more thing ... with 32-Bit Vista and two 1GB graphics cards ... you are also going to potentially run into some available memory issues with this setup. If you want to run sli280's, and really get something out of that second card, your priorities should go: 1) 1080p or bigger 2) 64-bit Windows 3) A nicely oc'd Phenom II chip (or an i5/i7 )P.S. Max fps is pretty meaningless, really. Also, you should grab the Crysis demo. Like I say, if you run that baby at Very High with 4xAA, you are going to get good scaling even at your current resolution, just like on the Sun Shafts test.
__________________
3dMark06: 26,490 http://service.futuremark.com/compare?3dm06=11655839 Quote:
Last edited by brettjv : 2 Weeks Ago at 02:48 AM |
||||||||||||||
|
|
|
|
#35 (permalink) | ||||||||
|
4.0 GHz
![]() |
Quote:
|
||||||||
|
|
|
|
|
#36 (permalink) | |||||||||||
|
New to Overclock.net
|
I was just worried that my hardware was faulty and that I had no idea what was going on.
__________________After your explanation, I feel a bit at ease, knowing that my cpu and low resolution were the culprit. I just dont want to waste 200 dollars on a new cpu and that not solving my problem. My old ASUS LCD should be sent here soon, so now I can game at a much higher res. I will probably get a 955 when the price drops If AMD comes out with a new chip, do you think it will still be based on a AM3+ architecture or will they phase them out and come out with a new one?
|
|||||||||||
|
|
|
|
|
#37 (permalink) | ||||||||||||||
|
4.0 GHz
![]() |
Quote:
na, i ran at very low resolution on my old system, with SLI doubled my fps in games that used it.
|
||||||||||||||
|
|
|
|
|
#38 (permalink) | |||||||||||
|
New to Overclock.net
|
u have an i7, i have a 9950. You cant even make the same comparison
__________________
|
|||||||||||
|
|
|
![]() |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | |
|
|