I just bought Sapphire 4870 and on my system it scores 10700 on 3dmark06. My previous card ( 3870) scored 9500. This as much as i know is very low score for the 4870. Any ideas why might that be? My processor is amd 5600+ and the psu is CoolerMaster real power 520w,( which is sufficient, if the psu is not broken in some way). Is it possible that i have a faulty card?
------------------------------Coolermaster RC1000 COSMOS; ASUS P5Q-E; e8400 with Coolermaster Hyper 212; Sapphire ATI HD4870; Coolermaster Real Power M520 PSU;2x2GB Apacer DDR2 800;HDD Samsung Spinpoint F1 750gb + Sygate 150GB
If you didn't overclock at all it sounds normal i think. that processor is kinda bad
Yes i thought that might be the case, but my friend on his rig( core2duo on 3.0ghz) had 13600 points. That is almost 3000 difference. Isn't that a bit too much? edit: That is with his 4870, and he lives in another country so i can't check mine on his system.
Message edited by LeonRa on 07-14-2008 at 07:48:44 PM
------------------------------Coolermaster RC1000 COSMOS; ASUS P5Q-E; e8400 with Coolermaster Hyper 212; Sapphire ATI HD4870; Coolermaster Real Power M520 PSU;2x2GB Apacer DDR2 800;HDD Samsung Spinpoint F1 750gb + Sygate 150GB
Reply to LeonRa
I think 3D mark is Nvidia biased. But even if it wasn't, don't worry about that stupid score, it's not worth much when it comes to gaming as it doesn't tell the whole story. Just forget about those numbers and play.
If you start getting worse FPS and quality then you should, THEN you come back and tell us something's wrong.
OK thanks folks. I'll definitely change the mobo and the cpu. I just thought something is wrong because the scores were so close. I still don't think they should be so close just because of the cpu.
its' your processor dude, very slow. I run a oc'd 8800gt and an oc'd q6600 and get ~14.5k. Hang on to that 4870 and save up for a new mobo/cpu
Correct.
OP: I have the same Sapphire card but I have a X3350 (yorkie quad) @ 3.2GHz. I pull in 15K with 3Dmark06 at stock settings. 3Dmark is very dependent on your CPU, so I would look at gaming performance more so than 3Dmark scores with an older CPU. Both of NVs and ATI's new cards this round are monster GPUs that are going to be sensitive to how good your CPU is.
That is not normal for the 4870, the CPU is holding it back.
Correct.
OP: I have the same Sapphire card but I have a X3350 (yorkie quad) @ 3.2GHz. I pull in 15K with 3Dmark06 at stock settings. 3Dmark is very dependent on your CPU, so I would look at gaming performance more so than 3Dmark scores with an older CPU. Both of NVs and ATI's new cards this round are monster GPUs that are going to be sensitive to how good your CPU is.
Why isn't it normal? With his processor/video card that kinda benchmark is normal.
The 4870 can do better than that at stock speeds. It's performance is being hindered by the CPU. If bottle necking our graphics performance with a slow CPU is normal then that is your cup-o-tea.
Also, OP, did you do a full clean up of the drivers before installing the latest CCC? I wonder if that might help out a bit.
The 4870 can do better than that at stock speeds. It's performance is being hindered by the CPU. If bottle necking our graphics performance with a slow CPU is normal then that is your cup-o-tea.
Also, OP, did you do a full clean up of the drivers before installing the latest CCC? I wonder if that might help out a bit.
Yes, thats the first thing i did, but it didn't help. Just in case i'll try the card in a different system and let you know what happened.
why does everyone think that if you have an amd your cpu limited. i know they're slower than c2d's cuz i have an amd. plus its a single card so the cpu wont decrese preformance that much. but 3dmark is very cpu dependent and i only get like an 1800 cpu scor on my 5200+ so ur 3dmark will be lower but in real world gaming it shouldn't make a big difference at all.
why does everyone think that if you have an amd your cpu limited. i know they're slower than c2d's cuz i have an amd. plus its a single card so the cpu wont decrese preformance that much. but 3dmark is very cpu dependent and i only get like an 1800 cpu scor on my 5200+ so ur 3dmark will be lower but in real world gaming it shouldn't make a big difference at all.
The cache size is abysmal on the 5600+ (L1 Cache: 128KB+128KB, L2 Cache: 2 x 512KB) vs what is on a Wolfdale E8400 (L2 of 6MB). I am not saying that is only difference or what is the problem but it does play a factor. What you say is not true. Better CPUs bring better GPU performance. That is the entire half-baked sales pitch NV was trying to throw at people that is just not true. There is a point to be made between quad cores and dual cores and how the games utilize them but that is off topic.
These benchmarks tell it all. I am not trying to say Intel is better than AMD but looking at the deviations between CPUs....they matter when it comes down to FPS.
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.