Overclock.net banner
1 - 18 of 18 Posts

·
Cheesebumps!!
Joined
·
3,034 Posts
Discussion Starter · #1 ·


posted bench with same clocks @3.7ghz..it seems i can still squeeze out some juice on the 3870's than the new 4850x2..can someone explain why i get these results??
 

·
Premium Member
Joined
·
9,807 Posts
Have you run a any other benchmarks to see what the difference is.

Try running the Crysis benchmark @ 16xAA. If you don't have crysis download the demo and you can benchmark it using that.
Run the benchmark in CSS the ingame one. I know that your frame rates will be super high, just curious of the difference between the 2.

If you can think of other benchmarks run those. Dont base your results on just 1 benchmark.
Good luck
 

·
Cheesebumps!!
Joined
·
3,034 Posts
Discussion Starter · #7 ·
Quote:

Originally Posted by grunion View Post
Doesn't look like you had the X2 oc'd for the 19,255 run.

Run some game benchmarks with AA and you'll see the X2 pull away, or if you have the advanced 06 version, add AA.
ok will try later..its OC'd to 700/1130 wrong readings from catalyst..but its really OC'd if not OC'd i only get 18k'ish scores..
 

·
Cheesebumps!!
Joined
·
3,034 Posts
Discussion Starter · #8 ·
Quote:

Originally Posted by BlackOmega View Post
Have you run a any other benchmarks to see what the difference is.

Try running the Crysis benchmark @ 16xAA. If you don't have crysis download the demo and you can benchmark it using that.
Run the benchmark in CSS the ingame one. I know that your frame rates will be super high, just curious of the difference between the 2.

If you can think of other benchmarks run those. Dont base your results on just 1 benchmark.
Good luck

will try this one too..

in gaming the 4850x2 is more responsive even on stock clocks but in benching it really sucks..
 

·
Registered
Joined
·
295 Posts
The 48XX series was ment to make up for the lack of filtering capabilities of the 38XX series, so, when running benches like 3dmark etc with no filtering MSAA/aniso. Plus as someone else stated, the other two are running in their own PCI-E lane. So crank up that AA to max on both and see what you get.
 

·
Registered
Ti Power
Joined
·
3,709 Posts
Quote:


Originally Posted by BlackOmega
View Post

Have you run a any other benchmarks to see what the difference is.

Try running the Crysis benchmark @ 16xAA. If you don't have crysis download the demo and you can benchmark it using that.
Run the benchmark in CSS the ingame one. I know that your frame rates will be super high, just curious of the difference between the 2.

If you can think of other benchmarks run those. Dont base your results on just 1 benchmark.
Good luck



I agree. My xfired 2900XT's got crushed by my 4870X2 in Crysis benchmark. It actually has a few PFS higher once I put full AA on it also.
Seems strange but the 48x0X2 cards love the AA.
 

·
Registered
Joined
·
3,096 Posts
Quote:


Originally Posted by Pheatton
View Post

I agree. My xfired 2900XT's got crushed by my 4870X2 in Crysis benchmark. It actually has a few PFS higher once I put full AA on it also.
Seems strange but the 48x0X2 cards love the AA.

It's that GDDR5 it just eat's AA and high resolutions like nothing. Thats why we love it though.
 

·
Cheesebumps!!
Joined
·
3,034 Posts
Discussion Starter · #12 ·
yup seems like that its cranking up with 4gb mems and with AA filtering added..compared to the 38xx series..but the CPU scores still goes low
 

·
Banned
Joined
·
8,570 Posts
sm2 is more cpu bound. You had a stronger cpu or a higher frequency on your 3870 crossfire run. Notice the canyon flight test, which is the test least effected by cpu frequency. The 4850x2 owned the 3870's face. Understanding the benchmark is the most important part of it.

*edit* just noticed you had the same cpu frequency. Crossfire always reduces cpu score, and it could be driver related between the 3870's and the 4850x2. Also, who knows, the 4850x2 make place more stress on the cpu.

What drivers are you running between the runs?
 

·
Cheesebumps!!
Joined
·
3,034 Posts
Discussion Starter · #15 ·
i really need your opinions on this one seriously..

1 more thing does benching in Vista can give a possibility to up or lessen down my scores??
 

·
Premium Member
Joined
·
6,985 Posts
vista should not really effect scores any more (maybe like 1-200 points)
 

·
god loves ugly
Joined
·
4,259 Posts
If it helps any, you can compare your results to mine in my sig. If you can browse my profile, you're more than welcome to look around all of my submissions. My 3870x2 should be able to rule out the multiple PCI-e slot theory.
 
1 - 18 of 18 Posts
Top