UPDATED, Part 4 of Benches out! E8400 vs Core i7 with my 5970,5870 and 5850

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Tempered81

Diamond Member
Jan 29, 2007
6,374
1
81
Damn, interesting. My 4ghz E8400 is still within 10% of a 4ghz i7, even in BC2 & Metro. I bet if you benched FC2 the 4g i7 would be 40% faster than the 4g e8400.
 

konakona

Diamond Member
May 6, 2004
6,285
1
0
I got the same PSU here, cool beans :)

Is that OCZ gold 1600 I am seeing in that pic? Mine doesn't seem to like my mobo (TP55). What vtt/vdimm/freq/latency are you using those at? Then again, I had problems with other corsair sticks with that mobo, it just might be defective for all I know.
 
Last edited:

thedosbox

Senior member
Oct 16, 2009
961
0
0
Kinda interesting that the stock E8400 consistently stomps all over the i7 in Metro 2033...
 

SSChevy2001

Senior member
Jul 9, 2008
774
0
0
Thanks for the benchmarks.

Looks like the E8400 held up just fine with a 5870, so I guess my Q9300 still has some life left in it for a new GPU upgrade.
 

Lonyo

Lifer
Aug 10, 2002
21,939
6
81
Kinda interesting that the stock E8400 consistently stomps all over the i7 in Metro 2033...

It doesn't stomp all over it, it does exhibit weird performance though.
Averages are about the same (GPU limited), but the stock E8400 gets a much lower minimum, but also a significantly higher maximum.
It's weird, but it doesn't stomp all over the i7 (and since the performance is equivalent between all the i7's and both E8400 clocks it's not a one off benchmarking issue).

Maybe Metro 2033 does something funny when it only has 2 cores available that results in the weird performance maximum?
The average seems clearly GPU limited, so it doing something funny which has to do wiht 2 vs 4(8) cores might make sense?
 

thedosbox

Senior member
Oct 16, 2009
961
0
0
It doesn't stomp all over it, it does exhibit weird performance though.
Averages are about the same (GPU limited), but the stock E8400 gets a much lower minimum, but also a significantly higher maximum.
It's weird, but it doesn't stomp all over the i7 (and since the performance is equivalent between all the i7's and both E8400 clocks it's not a one off benchmarking issue).

Well, when the CPU in question is over two years old and costs $100 less...

Maybe Metro 2033 does something funny when it only has 2 cores available that results in the weird performance maximum?
The average seems clearly GPU limited, so it doing something funny which has to do wiht 2 vs 4(8) cores might make sense?

Quite possible. Saboteur had problems with quad-core CPU's.


[edit] note, I'm not recommending people go buy a E8400. Just that existing owners may want to do a more thorough cost/benefit analysis.
 

Attic

Diamond Member
Jan 9, 2010
4,282
2
76
Thanks for benchies, I was surprised by CPU results, very good to know and keep in mind when I'm looking at my upgrade options/priorities. :thumbsup:
 

Apocalypse23

Golden Member
Jul 14, 2003
1,467
1
0
I think there is some problem with BFBC2 and the i7, I can't get it to run stable with any setting, everything else is running fine though, I'm working my ass off trying to figure out the bug....The game just crashes after a while...Further testing to continue.

EDIT- Resolved issue by disabling onboard Realtek Sound. It is a BFBC2 only issue.
 
Last edited:

Apocalypse23

Golden Member
Jul 14, 2003
1,467
1
0
What's the name of that behemoth case? I absolutely love it! :)

CoolerMaster Storm Sniper Black Edition

dsc00591w.jpg

dsc00592wo.jpg

dsc00593u.jpg

dsc00594j.jpg
 

jaggerwild

Guest
Sep 14, 2007
430
0
0
I think there is some problem with BFBC2 and the i7, I can't get it to run stable with any setting, everything else is running fine though, I'm working my ass off trying to figure out the bug....The game just crashes after a while...Further testing to continue.

I had a similar issue, my fix was to reinstall DX after updating BFBC2. No issue's after this. hope it helps you!
 

Apocalypse23

Golden Member
Jul 14, 2003
1,467
1
0
I had a similar issue, my fix was to reinstall DX after updating BFBC2. No issue's after this. hope it helps you!

Hmmm, what kind of crashes were you experiencing? Did your entire system crash? Or just the game, my entire sytem crashes, sometimes I get a BSOD...

I'll try doing what you suggest though. Thanks.
 

n7

Elite Member
Jan 4, 2004
21,303
4
81
Really nice job with the benchmarking! :thumbsup:

FWIW, P95 is somewhat of a waste of time when stability testing with i5/i7s.

Use LinX or IBT...it'll find stability issues MUCH faster & will fail on settings P95 passes on.
 

blanketyblank

Golden Member
Jan 23, 2007
1,149
0
0
Interesting numbers. I wonder what happened in Dirt 2 Croatia for the 5970 at 1.0 to beat out the 5970 on 2.0 slot. I guess it might be the benchmark being unreliable with the cars doing different things as people have said before.

Also with Metro 2033 I see the huge maximums with the e8400 and the same average as a bad thing since it means the game will be alot choppier and the game dipped below the average more often or went lower to compensate. It is interesting though why that happens with the core 2.
 

killster1

Banned
Mar 15, 2007
6,208
475
126
love the post but do you think maybe the psu on the pcie2 board is holding it up? or the fact it has half the ram? seems like should be matching for best results.. also case looks cool but doesnt look any bigger than a antec 1200, is it?

I also think the metro 2033 is interesting i have 4890 crossfire and q6600 stock and runs every game perfect at 1900 res or 1650 if 1900 isnt a option. if you had a quad q9xxx or q6xxx be interesting to see the pcie1/2 vs i7 (im even tempted to loan you a spare q6600 i have heh)
 

Apocalypse23

Golden Member
Jul 14, 2003
1,467
1
0
Really nice job with the benchmarking! :thumbsup:

FWIW, P95 is somewhat of a waste of time when stability testing with i5/i7s.

Use LinX or IBT...it'll find stability issues MUCH faster & will fail on settings P95 passes on.

Thanks, I finally solved the BFBC2 issue by disabling onboard sound. And here are my final 4.0ghz prime stable and Linx stable numbers. I ran both these programs for roughly 30 min all together. Coolermaster Storm Sniper Case fan speed was at 23%.
BIOS Core Voltage level: Standard

Prime and Linx Load:
i7load24ghznewlinx.png


Idle after stress test:
i74ghzidle.png
 
Last edited:

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,001
126
Thanks for taking the time to do all these benches Apocalypse23. Now stop benching and start enjoying that machine as a gamer. :)

It's nice to see that a dual core still has plenty of oomph in the games you tested. I imagine there are games that it would fall behind significantly, but in the games you tested it's still a very good CPU.
 
Last edited:

Apocalypse23

Golden Member
Jul 14, 2003
1,467
1
0
Damn, interesting. My 4ghz E8400 is still within 10% of a 4ghz i7, even in BC2 & Metro. I bet if you benched FC2 the 4g i7 would be 40% faster than the 4g e8400.

Yes unfortunately I don't have that game any more, I'd have to see if I can get a copy to run the tests.

I got the same PSU here, cool beans :)

Is that OCZ gold 1600 I am seeing in that pic? Mine doesn't seem to like my mobo (TP55). What vtt/vdimm/freq/latency are you using those at? Then again, I had problems with other corsair sticks with that mobo, it just might be defective for all I know.

Posted above in pics. I ditched the OCZ Gold as it wouldn't run past 1400Mhz...It is rated at 1066 by OCZ and shipped as 1600 but also doesn't have any XMP profile. I instead used Crucial RAM if you look at the prior posts. :cool:

Thanks for taking the time to do all these benches Apocalypse23. Now stop benching and start enjoying that machine as a gamer. :)

It's nice to see that a dual core still has plenty of oomph in the games you tested. I imagine there are games that it would fall behind significantly, but in the games you tested it's still a very good CPU.

Thanks and I'm having a lot of fun running these tests, I still have the 5970, and hybrid CF tests to come. :D
 

Apocalypse23

Golden Member
Jul 14, 2003
1,467
1
0
Okay I conducted the following tests:

5850+5870 (worked well in 3dMark Vantage, was even able to overclock each card individually and run in CF, however, games like Dirt 2, Metro 2033, Unigine Heaven all crashed due to some unknown reasons, it's most probably a CF issue with a 5850 and a 5870, I've therefore given up on testing a 5850+5870 for the time being.

bench111.png

bench222.png

bench333.png

bench444.png


5970 results coming shortly, I also noticed that my case can not fit a third card as I need an additional expansion slot on the case...I'd have to try to take the mobo out and run the 5970+5870+5850 CF test at a later date.

Notes: Crysis had very minor glitches of micro stuttering, these were very minuscule but I could tell that it was a bit of an issue. BFBC2 and Dirt 2 ran as smooth as silk, it was so perfect I couldn't believe it.
 
Last edited:

happy medium

Lifer
Jun 8, 2003
14,387
480
126
Is it possible to do the e8400 @ 3.9 with the crossfire setup and compare it to the i7 system?

It's nice to know my Q9550 will push a crossfire 5970 and 5870 crossfire setup. Thanks.