Tigger69
Ati HD3850 256mb @ 770mhz core/1008mhz mem
2911

http://img.techpowerup.org/080110/Untitled.jpg

one more.. this one would show 944 core in the CCC setting..

http://www.cavecom.com/pics/fanfix-.jpg

trog

boy this is a very ATI bias benchmark .

not really, I guess it utilises one aspect that ATI cards excel at. the 2nd place is an 8800 gt

its a pure a grafix card bench.. just like the original 3dmarks were..

its not helped along by massively overclocked intel cpus.. upping my cpu speed by 300mhz is good for about two points.. which is why i havnt bothered doing it..

its scored well on my amd system too.. pretty rare for a bench to do that.. except a memory bandwidth one.. amd still holds the crown there..

trog

Poor CPU scaling probably is likely due to the nature of the benchmark object - the CPU only "sees" 2 objects in the scene:

- the toroid
- one "hair" of fur which is cloned by the GPU on-the-fly forming an impression of an even fur cover

And the bench is probably heavily optimised for R600 architecture. That doesn't mean it's biased - just optimised to use all available resources.

Well I still think it is heavily ATI BIAS . JMO .

whats odd is its open gl.. i always thought nvidia was better at that.. i know it heats the card temps up more than anything else.. its also one of the few things that loads my gpu at 100%.. 2006 dosnt.. at least not at its default resolution.. at higher ones it probably does..

trog

I agree I upped my x17 EE from 3.4 to 3.8, and no help,even gave a shot w/667 bootstrap nada, but the BLISS from 630 to 680:rockout:

Random Murderer
HD3870 @ 931/1305
3549


Sorry trog, I'm taking back the 3870 crown.

http://img.photobucket.com/albums/v6...le666/o3-1.jpg

candle_86

6600GT @ 540/1150
311
http://i36.photobucket.com/albums/e4...86/sssssss.jpg

High score at that res :eek:

http://img.techpowerup.org/080116/rawrnewrecordz.png

GPU doesnt work it crashes my system when I try and launch it :(

yeah, don't run at high res with fur, its really glitched out at high res so its best to stick on the default res

what about rivatuner then

try to push the card harder, I have hit nearly 3170 with my 3850, and will go higher when censuspc finally decides to send out my cooling equiptment I bought, I hate those guys so much.

try pushing the memory, the 3850 gets bottlenecked by its memory alot

My cpu is holding me back. I have my core clocks higher than HD 3870 speeds.. I cant get my mem any higher it just locks up.

I dont know how my computer did worse compared to that of mandalore's when he was on a dual core opty and 2900XT, but here is mine:

CrAsHnBuRnXp
XFX 8800GT 512MB 725/1000
2811

Your system is like 1000x better than mine how did I score higher :wtf:

CPU speed means near nothing in fur, I have seen no benefit overclocking or underclocking my cpu using it

its all done by your card and nothing else on your system.. thats why if you want to try and make sure your card is stable then you do the stability test on fur...

That doesnt explain why a 2900XT beats my 8800GT overclocked though.

Yea.. My HD 3850 shouldnt beat your 8800GT in anything. This benchmark is WHACK!

It tests your Card and that only... if you look at my two posts.. It ran at stock first with a non flash bios and no mod.. Then I flashed a new bios, did the v mod, and went up almost 400 points

Well still, my card should not be losing to a 3850 let alone a 2900XT.

2900XT is very strong in this bench, stronger that the 3850. It's very easy for one to overtake an 8800GT. I'm guessing this bench is benefited by the number of shaders in ATI cards.

yea, seriously crash, stop frickin whining.
this isn't a popularity contest.
the fact of the matter is that the only 8800gt, nay, the only nvidia card in the top ten was volt modded and heavily overclocked.
ati cards are just better at THIS PARTICULAR BENCHMARK.
calm down, unbunch your panties, and go run aquamark3.

That is right. It works better with a card that has the better shader. 3850/3870 is a better card... now if we were looking for what makes the Nvidia cards do great, then we would have a bench where Nvidia cards own.... This is a shader test tool.. sorry for your luck Crash, but thats all it is...
Ocing is all about who can get the best out of their product... Plain and simple.

My burned 8800GT never came to life. Turned out the GPU had shorted when the mosfet fried. Anyways, here's my new toy:

Asus 8800GTS 512MB @ 802 / 1998 / 1053 (2106)
3433
http://img.techpowerup.org/080117/FUR3433_c.jpg
No vmods yet but GPU is watercooled. At stock freqs GTS scores a bit over 2800.

Well one, Im not really whining as I was just trying to figure out why the hell i was getting low scores compared to that of ATI cards which are much lower in performance compared to that of my card.

Two, I could care less about the top nvidia card.

Three, they need to have a benchmark that is not favored for one particular brand in cards.

Four, I was never uptight in the first place and since the makers of Aquamark 3 were bought out a few years back, there are no new releases and thus it does not work on any machine greater than XP.

not in the fur bench they aint.. which is why ati cards score higher..

trog

Hence me stating that there needs to be a graphics benchmark that can run under Vista that does not favor one brand over the other.

that's what WE'RE saying, the test doesn't favor ati, the 2900's and 3800's just fare so well because they have so many stream processors.

Vista is just to new to have that happen yet... yeah, we all knew that it was coming out, but what would you do? develop for a os that we know runs and works or a os system that don't have a clue that runs, and will be buggy if we do?

I'm with you on vista, love it... but it will come. wait and time will tell

Correct, and if a developer would program a game using this many shader ops, the ATI cards would embarrass the nVidia cards. ATI was banking on this happening when making their design, but it hasn't panned out that way. Things would be a lot different in the gfx market if it had.

HD3870 @ 985/1287 Thank you pencil mod!:D
http://img.techpowerup.org/080117/Fur3707.jpg

dude thats awesome. whats cooling that monster?

Just a regular water cooling setup. This card beg for more voltage...;)

nice one but i recon u might well kill your card.. mine got very hot at 1.5 volts.. but i recon your score will do for a while..

i would scrape some lead of now u have it.. get the other benches first thow.. he he he

and wile is right about the way games are written.. the 2900 was a little advanced for em..

trog

oooh, nice..

guess once my pump arrives ill post my other scores ;)

what temperatures were you getting?

Just because I don't see any x800 class cards in the scores here is my x800XL@440/540.

http://img.photobucket.com/albums/v2...rbenchmark.png

Newtekie, are you trying to get the lowest scores!?!? lol.. looks good and sweetness.

rofl it beats the X1650PRO :roll:

X800 goodness!!

back in the day it rocked the shit outa everything...

i remember modding it from a pro vivo to an xt platinum edition, rock solid. its what i had when i first got HL2. Its daft how far the cards have come along in comparison :laugh:

but to the x800 :toast:

my 7800GS beat it and cost about the same when i was looking (agp wise) so 7800GS goodness for me :rockout:

It isn't surprising when you look at the specs of the cards.

The x800XL has 16 ROPs and 16 Shaders, the x1650Pro only has 4 ROPs and 12 Shaders. The x1650Pro is clocked higher though, so it does edge out the x800XL ever so slightly in Shader operations, but that isn't enough to make up for the lack of ROPs. Add to that the fact that the x1650Pro only has a 128-bit memory bus, compared to the 256-bit bus of the x800XL, and it is no wonder the x800XL wins.

hehehe...

if only my original TEC hadnt melted it, then id show you some! :p :roll:

The gpu run at 38c at 1.51 volt:wtf:
Do you think its safe to go 1.6 -1.65 if the temps are ok? (i have an antec spot cool on the pwm)

safe is relative word dude.. u are the guinea pig u tell us.. he he

i know what mine did at 1.5 volts so your cooling must be bloody wonderfull to keep it at 32 C.. u sure u are measuring it right..???

trog

ps.. its your load temps i am on about.. 32 C seems way too cool for 1.5 volts..

Im using ati tool and everest to mesure the core temperature, on stock cooler it was in the 90s.
At 839 stock voltage on water cooling it stay at 38 100% load and 985 mhz 1.51v still 38...
the board is slightly warm to the touch. So the temp mesuring seem to work.

NastyHabits
6600 GT
152

Drat. I'm not the slowest.

http://img.techpowerup.org/080119/Fur benchmark.jpg

hi everyone, my first post here in the forums. This bench caught my attention and here are my results.
HD2900XT 1GB 832Mhz Core
http://img.techpowerup.org/080119/furbench.jpg

reading the same sounds wrong to me dude..

the stock air cooler auto targets a certain temp.. i cant see how your water does that.. all i know is 1.5 volts generates a hell of a lot more heat than 1.34 does.. so u should see a temp increase of some sorts..

i use rivatuner monitor to give me temps read outs..

my results were at 1.34 volts.. the stock air cooler runs at less than 40% fan speed to maintain its 90 C target under full load.. the fur benchmarks heats things up fully..

at 1.5 volts the stock air cooler ran nearly 90% leaf blower style and the temps went up to 102 C..

with its fan running leaf blower stye the stock cooler isnt a bad cooler.. 1.5 volts did kinda overwhelm it thow..

okay water is better but 32 C at stock and the same at 1.5 just dont seem right to me..

my temps were with the card running around 830 core.. what they would have been like up around the 1000 core i cant imagine..

trog

ps.. i see u say 38 C.. my 32 must from an earlier post..