The great dual vs quad vs hex gaming benchmarkapalooza

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Lonbjerg

Diamond Member
Dec 6, 2009
4,419
0
0
Most of those games don't support >=4 threads...so what is this supposed to tell me?
 

Lonbjerg

Diamond Member
Dec 6, 2009
4,419
0
0

/offtopic
Could you stop posting replies to me, you trolling is getting boring, thx.


/topic
Again, if you sample group (read: games) dosn't use more than 2 threads tthose charts are of no use.

Try real multithreadded games...like ARMA2.

BREAKING NEWS:
Dual threadded software don't run faster on quad or hexa cores...oh noes!!!
:thumbsdown:
 

2disbetter

Member
Jul 19, 2011
72
0
0
Yep, it is blocked. IE7 actually shows me the red boxes. The other browsers dont. Note that simple text reads much faster, and is never blocked! Now that I've been able to actually look at these pics through a proxy, the big question is: what about 3 cores? Hasnt that been shown to be the critical cutoff point for good performance?


I guess posters should be mindful of all the world's ISP limitations (I'm in Germany and because of my ip's location alot of copyrighted material I could normally see in the states is not viewable here without the help of a proxy) and work proxy limits (government machine, and this government doesn't feel image hosting is a privilege I deserve etc.). Never mind knowing how to use the internet yourself.



To the OP thanks for the thread. Excellent work. I can't believe people are actually complaining about your voluntary work here. Thanks for the hard work!

2d
 
Last edited:

AdamK47

Lifer
Oct 9, 1999
15,168
2,826
126
I went ahead and crunched the numbers to see the percentage increased.

Alien vs. Predator (2010)
Dual: 224.8
Quad: 226.9
Hexa: 227.0
Dual-->Quad = 0.093%
Quad-->Hexa = 0.04%
Dual-->Hexa = 0.98%

Batman AA
Dual: 147
Quad: 170
Hexa: 172
Dual-->Quad = 16.65%
Quad-->Hexa = 1.18%
Dual-->Hexa = 17.01%

Battleforge
Dual: 171.654
Quad: 200.729
Hexa: 210.456
Dual-->Quad = 16.94%
Quad-->Hexa = 4.85%
Dual-->Hexa = 22.60%

Civilization 5
Full Render Score
Dual: 1546
Quad: 2449
Hexa: 2617
Dual-->Quad = 58.41%
Quad-->Hexa = 6.86%
Dual-->Hexa = 69.28%
NoShadow Render Score
Dual: 1943
Quad: 3520
Hexa: 3732
Dual-->Quad = 81.16%
Quad-->Hexa = 6.02%
Dual-->Hexa = 92.07%
No Render Score
Dual: 1939
Quad: 3313
Hexa: 3934
Dual-->Quad = 70.86%
Quad-->Hexa = 18.74%
Dual-->Hexa = 102.89%

Company of Heroes
Dual: 208.0
Quad: 245.3
Hexa: 243.6
Dual-->Quad = 17.93%
Quad-->Hexa = -0.70%
Dual-->Hexa = 17.12%

Call of Juares
Dual: 229.8
Quad: 235.2
Hexa: 237.7
Dual-->Quad = 2.35%
Quad-->Hexa = 1.06%
Dual-->Hexa = 3.44%

STALKER Call of Pripyat
Day
Dual: 159.8
Quad: 169.1
Hexa: 173.3
Dual-->Quad = 5.82%
Quad-->Hexa = 2.48%
Dual-->Hexa = 8.45%
Night
Dual: 184.3
Quad: 194.3
Hexa: 198.1
Dual-->Quad = 5.43%
Quad-->Hexa = 1.96%
Dual-->Hexa = 7.49%
Rain
Dual: 206.3
Quad: 206.3
Hexa: 224.4
Dual-->Quad = 0%
Quad-->Hexa = 8.77%
Dual-->Hexa = 8.77%
Sun Shafts
Dual: 122.3
Quad: 138.3
Hexa: 142.2
Dual-->Quad = 13.08%
Quad-->Hexa = 2.82%
Dual-->Hexa = 16.27%

Crysis 2
Times Square
Dual: 108.5
Quad: 121.4
Hexa: 125.0
Dual-->Quad = 11.89%
Quad-->Hexa = 2.97%
Dual-->Hexa = 15.21%
Downtown
Dual: 96.8
Quad: 104.0
Hexa: 109.0
Dual-->Quad = 7.44%
Quad-->Hexa = 4.81%
Dual-->Hexa = 12.60%
Central Park
Dual: 105.5
Quad: 109.4
Hexa: 111.5
Dual-->Quad = 3.70%
Quad-->Hexa = 1.92%
Dual-->Hexa = 5.69%

Doom 3
Dual: 188.0
Quad: 195.6
Hexa: 198.8
Dual-->Quad = 4.04%
Quad-->Hexa = 1.64%
Dual-->Hexa = 5.74%

Far Cry 2
Dual: 116.99
Quad: 160.51
Hexa: 168.01
Dual-->Quad = 37.20%
Quad-->Hexa = 4.67%
Dual-->Hexa = 43.61%

Just Cause 2
Dual: 99.05
Quad: 105.26
Hexa: 105.12
Dual-->Quad = 6.27%
Quad-->Hexa = -0.13%
Dual-->Hexa = 6.13%

Lost Planet 2
Dual: 124.5
Quad: 162.0
Hexa: 170.8
Dual-->Quad = 30.12%
Quad-->Hexa = 5.43%
Dual-->Hexa = 37.19%

Mafia 2
Dual: 42.5
Quad: 51.9
Hexa: 60.3
Dual-->Quad = 22.12%
Quad-->Hexa = 16.18%
Dual-->Hexa = 41.88%

Metro 2033
Dual: 71.33
Quad: 79.67
Hexa: 79.00
Dual-->Quad = 11.69%
Quad-->Hexa = -0.85%
Dual-->Hexa = 10.75%

Resident Evil 5
Dual: 176.5
Quad: 213.7
Hexa: 219.3
Dual-->Quad = 21.08%
Quad-->Hexa = 2.62%
Dual-->Hexa = 24.25%

Shogun 2
Dual: 126.438
Quad: 139.537
Hexa: 140.75
Dual-->Quad = 10.36%
Quad-->Hexa = 0.87%
Dual-->Hexa = 11.32%

Grand Theft Auto 4
Dual: 61.37
Quad: 65.27
Hexa: 65.54
Dual-->Quad = 6.35%
Quad-->Hexa = 0.41%
Dual-->Hexa = 6.79%
 
Last edited:

Voo

Golden Member
Feb 27, 2009
1,684
0
76
I guess posters should be mindful of all the world's ISP limitations (I'm in Germany and my ISP blocks a butt ton of content from the states, etc.) and work proxy limits (government machine, and this government doesn't feel image hosting is a privilege I deserve etc.). Never mind knowing how to use the internet yourself.
Uh no I doubt that your ISP is doing that (that would have some severe consequences in german law).

Now that your employer blocks specific URLs that's nothing out of the ordinary I fear - heck I had work where for security reasons there was no outgoing internet connection at all (now that's boring :( ).
 

m3t4lh34d

Senior member
Oct 23, 2008
203
0
0
Be happy the guy even did the tests at all... you guys are complaining about graphs?
It's enough of a PITA to do all of those tests, do you really think adding the extra headache of manually typing all of those numbers into excel would be worthwhile for the dude?

Be grateful ffs
 

Idontcare

Elite Member
Oct 10, 1999
21,118
58
91
It is not thread crapping to point out the fact that at least 5% of the people who load this thread arent going to be able to see wtf is supposed to be here. Pointing that out reduces the possibility of seeing more of it in the future. Stuff that is posted here should *just work*. Period. No ifs and or buts. It's not my problem if it dont work. Stick with what works.

For the record it IS thread crapping.

What you had to say here could have just as easily been communicate via pm to the OP, there was no reason to state it publicly.

It is your problem if stuff doesn't work. Your false sense of entitlement is the source of your disappointment. Align your expectations with reality and get over it.

If you can't see it, have alerted the OP to it via pm, and they have not pursued addressing it, then simply move on. Simple enough.

This is not a pay-for-use forum, you are getting exactly what you paid for. But we do have posting guidelines and thread-crapping is not acceptable.
 

2disbetter

Member
Jul 19, 2011
72
0
0
Uh no I doubt that your ISP is doing that (that would have some severe consequences in german law).

Now that your employer blocks specific URLs that's nothing out of the ordinary I fear - heck I had work where for security reasons there was no outgoing internet connection at all (now that's boring :( ).


I didn't go into detail, but I was mainly referring to copyright blocks and things of that nature. You're right germany in general is very open with their internet traffic.

2d
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
Great work, Adam! I hope that some of you out there who are thread crapping (and some who are not) will in the future put forth even a small fraction of the effort that Adam put into this thread. From my own (very limited) personal experience with benchmarking, he put a LOT of effort into this and deserves our thanks.

I thought that the Civ5 results were interesting, I expected much better scaling from 4-6 cores. Perhaps civ is taking advantage of HT so much that adding 2c/4t has limited returns.
 

apoppin

Lifer
Mar 9, 2000
34,890
1
0
alienbabeltech.com
ABT has done similar review. I think the general conclusion was with a single GPU a dual core was enough. Jump to SLI/Crossfire a quad is necessary.
It depends on the game. The newest games are taking advantage of more than two cores, however. And it is always a good idea to have a fast Quad if you run multi-GPU although it is not yet mandatory.

Be happy the guy even did the tests at all... you guys are complaining about graphs?
It's enough of a PITA to do all of those tests, do you really think adding the extra headache of manually typing all of those numbers into excel would be worthwhile for the dude?

Be grateful ffs
Sure - if he was posting this for a tech site evaluation, graphs are a must. However, he is doing what we used to do here before we put ABT together. It is great work and i'd love to get back to it myself (as soon as i finish this damn 4-month S3D/3D Vision evaluation, i am testing CPU platforms and SLI vs CF at high resolutions - 1920x1080>2560x1600>5760x1080).

Actually, it is just as easy to put the results into an Excel chart - as you bench - as it is to type them out; actually it is easier!
:\
 

guskline

Diamond Member
Apr 17, 2006
5,338
476
126
Adam, you put a lot of time into this analysis. THANK YOU!
BTW, I'm pleased as punch with the 2500k@4400 running crossfire 5850s
 

Patrick Wolf

Platinum Member
Jan 5, 2005
2,443
0
0
All this really shows is the core scaling on these 17 games using the same Gulftown CPU. It doesn't account for changes in microarchitectures which is the data potential upgraders really need.
 

toyota

Lifer
Apr 15, 2001
12,957
1
0
All this really shows is the core scaling on these 17 games using the same Gulftown CPU. It doesn't account for changes in microarchitectures which is the data potential upgraders really need.
yeah two cores of Gulftown with HT would still match or beat almost any other cpu no matter if that other cpu had 2, 3 or even 4 cores. even without HT, it would run rings around a typical dual core cpu.
 

Shmee

Memory & Storage, Graphics Cards Mod Elite Member
Super Moderator
Sep 13, 2008
7,381
2,414
146
adam, great job! Curious, I have a very similar system to yours. asus p6x58d premium, 960x @4.2, my blck is lower though, im running 155x27=4.2. memory is about 1860ddr. are you overclocking your uncore? what is it running at/what multi are you using? mine is at 20x=3118 Mhz for uncore freq.

It would be nice to know what settings you are using, for your overclock; bios revision, settings, CPU stepping/batch/stats.

Just curious, you have a very similar system, was wondering how far you could oc yours and what it took. Mine could do about 4.4, not sure it was 100% stable, plus I didnt think the vcore was low enough. also, what are temps like?
 

AdamK47

Lifer
Oct 9, 1999
15,168
2,826
126
Most of that information is in the CPU-Z screenshot in the first post.

Other than that...
The Uncore/L3 is at 3400MHz with a 17x multiplier.
The QPI/Uncore voltage is at 1.325V.
The IOH voltage is at 1.2V.
The Memory is at 2000 with 9-10-9-24-1T timings at 1.60V.
The P6X58D Premium BIOS is 1402.
The P6X58D Premium has been modified with an auxiliary 12V molex connector for the PCI-E.