GeForce 8800GT 512meg - Whose getting one?

[seremtan]

ok, got the XFX version of this (i.e. same card without the minor OC), and my god does it make a difference: crysis demo @ high quality as optimal settings, 1280x960, looks awesome. same with COD4 and UT3, r0XX0rs my b0XXers

big fucking card too

[creep]

Glad to see another gamer pulled back from this SLI silliness.

[Canidae]

I got a BFG GeForce 8600GTS recently because I'm quite happy with a medium sized ePenis (but I did measure it using 3DMark as soon as the surgery was complete)

[creep]

3dmark has been completely irrelevant for around 6 years. Why do people still run it?

[shadd_]

er to see if everything is working properly. if your score is 50% less than other systems with the same hardware something must be fucked.

[Massive Quasars]

Is 3dmark still a reliable benchmark?

[creep]

Considering that 3dmark uses synthetic benchmarking, and most of the calc is done on the gpu, and tailored/adjusted for specific gpu... that's seems that by the time you notice the problem there, it would be apparent in plenty of real-world computing.

[shadd_]

sure. when i get new hardware i run 3dmark and check the ORB(online result browser)to see if everything is hunky dory.

[shadd_]

Considering that 3dmark uses synthetic benchmarking, and most of the calc is done on the gpu, and tailored/adjusted for specific gpu... that's seems that by the time you notice the problem there, it would be apparent in plenty of real-world computing.

no there is cpu test as well. and certain gpu tests reflect cpu. i know you have an 8800gt. with my q6600 at 3.6ghz my sm2.0 score would be higher than yours with an x2 6000+. our sm3.0 score would be the same because their not as dependant on cpu.

it gives mucho info if you know how to look for it.

[creep]

whatever info it gives cannot be trusted.

Real benchmarking is the only thing that matters. That means benchmarking with applications you'll actually use, not pre-determined routines that exploit specific instructions to manufacture scores on certain hardware.

I understand your point completely, there is a lot to be said for software that tailors to specific hardware in terms of finding defective parts... but when that software masquerades itself as an all around benchmark it is corrupt.

[shadd_]

so if your 3dmark scores are where they should be for your particular hardware, its not a good indication that everything is running properly?

no it's not an all around benchmark. thats why the futuremark has PCmark as well.

[creep]

No, that's exactly what I said. A synthetic benchmark may help you find a defective component... but by the time it does, the defect would have been apparent in other areas. And at that point, the purpose of the synthetic benchmark has been removed.

[shadd_]

i think i see where you're coming from creep. you could have 2 gt's in sli with an x2 4000+ and get say 15000 in 3dmark. on another system with one gt and a qx9950 getting 15000 points too. yeah it's misleading because the qx9950 is obviously a better gaming machine. to the average person they wouldn't know better.

the answers are in the details though. with a bit of knowledge you can piece things together.

edit: i started typing this before your last post. in some ways you're right, others not. [/nerd]

[creep]

In the end it boils down to a choice:

I have test my hardware with software that is designed to exploit its strengths, and get a score that relates to other identical harware.

Or I can test my hardware with software that I will actually be using, and gives me a performance measure against all hardware running that software.

[Scourge]

I got a BFG GeForce 8600GTS recently because I'm quite happy with a medium sized ePenis (but I did measure it using 3DMark as soon as the surgery was complete)

Have an 8600GTS myself and it seems to be doing just fine. It runs the crysis demo pretty well, and anything else I've thrown at it. Granted maybe not on the highest resolutions, but close enough.

[Canidae]

3dmark has been completely irrelevant for around 6 years. Why do people still run it?

It is until Creep brings us the new tablets down from the mountain.

I use it once every few years to compare to my results from my just being pulled existing card so it works for me.

So tell us what measuring stick I should be dropping my shorts to on a more frequent basis (to entice me to needlessly upgrade more frequently because it make me feel limp) and I'll take a look at it

[seremtan]

I got a BFG GeForce 8600GTS recently because I'm quite happy with a medium sized ePenis (but I did measure it using 3DMark as soon as the surgery was complete)

Have an 8600GTS myself and it seems to be doing just fine. It runs the crysis demo pretty well, and anything else I've thrown at it. Granted maybe not on the highest resolutions, but close enough.

so the 8600 isn't too bad? i usually go up thru mid-range cards (5600, 6600, 7600) but this time went for the top(ish) end, mostly because i'd heard the mid-range GF8 range wasn't up to much

[mac]

i'd heard the mid-range GF8 range wasn't up to much

you've heard right.. crysis on a 8600 is a joke.. low settings low resolution where it looks even worser than farcry. anything less than a 8800 is wasted money.. prolly the next iteration the 9xxx series will have again some decent mid range cards

[Scourge]

i'd heard the mid-range GF8 range wasn't up to much

you've heard right.. crysis on a 8600 is a joke.. low settings low resolution where it looks even worser than farcry. anything less than a 8800 is wasted money.. prolly the next iteration the 9xxx series will have again some decent mid range cards

I don't know what you're talking about. I run it on high settings.

[Hannibal]

just FYI, all the 8000 series cards are glitchy cunts on XP with unreal 2/3 engine games. It's a well known problem, nvidia fixed it for Vista, but r leaving us XP fags out in the cold for almost a year now.

[shadd_]

I don't know what you're talking about. I run it on high settings.

everyone has a different threshold for acceptable frame rates. imo 40fps or higher is acceptable.

right now the mid-range price/perf champ is the 3850 512meg from ati. oc's like a mofo apparently too.

[seremtan]

i haven't had any problems with UT3 at all, running at high quality @ 1600x1200. sounds like you need the latest beta drivers that fix the UT3 glitch (whatever it is)

[Pooinyourmouth]

I'm running the original 8800 GTS 640meg card, but I've been able to overclock the shit out of it. It came stock at 513MHz core, and memory was at 792MHz. I have my core at 642 and memory at 980 now. I'm getting 10 to 25% gains in the newer games like UT3 and Crysis. Older games don't seem to care about the overclock much, but that doesn't really matter since most have a frame rate well into the 100's already. Best part is that I'm still using the stock cooler, and it's still so quiet I can't hardly hear if my computer is even on.

Oh and I've been playing UT3 on Xp with no problems. They've had a driver out for a while now that fixes the locking up issue.

[JooKed]

They released a 1 Gig version of the 8800GT. I don't even think it gives more than a 5% performance increase, but I guess if you used it in SLI, that gig would be helpful.

[Hannibal]

i haven't had any problems with UT3 at all, running at high quality @ 1600x1200. sounds like you need the latest beta drivers that fix the UT3 glitch (whatever it is)

My main concern is ut2004 on xp...which hitches like a mofo on several stock maps now even though my 6800 ran it like butta with high settings. There's a pretty huge thread over at the nvidia forums about it. The Nvidiatechnonerds have even stated that they're aware of it. It may not affect you and Tom and Randy, but quite a few of the rest of us are stuck with it. The workarounds for the ini file are doable but turn the game into a visual turd. The latest XP beta driver has no impact on the problem for ut2004.