L2 cache comparison roundup - E7200 vs E8500 plus QX9770 - lets take a look

[PcCI2iminal]

Dont know if this has been posted, but here is a comparison between the L2 cache .


http://img155.imageshack.us/img155/6715/45nmfotohr7.jpg

in some apps llike Cinebench,Wprime,etc the duals will see a performance penalty in comparison to a quad but is interesting to see the quad vs dual in single tasks like Pi,Pifast,etc.
Well, how does cache (3mb vs. 6mb vs. 2x6mb) affect the Core2 performance ?
lets take a look and enjoy :)

Core2 3.17GHz
E7200 @ 334x9.5 vs E8500 @ 334x9.5 vs QX9770 vs 334x9.5
Team Group DDR3 1600 @ 1333MHz Cas6 6-6-15 (see memset at Aquamark)
Asus P5E3 Premium 0402
9800GTX default clock
Forceware 175.16 high performance with no LOD

Aquamark fps
http://img155.imageshack.us/img155/8139/aquaim5.jpg

3D06 3dmarks
http://img529.imageshack.us/img529/7697/3d06pg3.jpg

3D06 fps
http://img148.imageshack.us/img148/5371/3d06returntoproxyconiy6.jpg
http://img529.imageshack.us/img529/2228/3d06fireflyforestnc1.jpg
http://img155.imageshack.us/img155/4227/3d06cpu1redvalleyli8.jpg
http://img148.imageshack.us/img148/1466/3d06cpu2redvalleylo1.jpg
http://img148.imageshack.us/img148/6505/3d06canyonflightaf1.jpg
http://img529.imageshack.us/img529/2011/3d06deepfreezeae2.jpg

Cinebench
http://img529.imageshack.us/img529/2301/cinebench1cpupr7.jpg
http://img529.imageshack.us/img529/3135/cinebenchxcpuse4.jpg
http://img529.imageshack.us/img529/305/cinebenchmultiprocessorvd9.jpg

Cpumark
http://img148.imageshack.us/img148/3456/cpumarkqo8.jpg

Pifast secs
http://img529.imageshack.us/img529/1124/pifastfw2.jpg

Everest read MB/s
http://img148.imageshack.us/img148/668/everestreadnr7.jpg

Everest latency ns
http://img148.imageshack.us/img148/4436/latencygh1.jpg

Fritz Chess
http://img148.imageshack.us/img148/1513/fritzchessbw5.jpg

Super Pi 1m secs
http://img529.imageshack.us/img529/1234/superp1ok6.jpg

Winrar KB/s
http://img155.imageshack.us/img155/4674/winrarzk1.jpg

WPrime 32m secs
http://img148.imageshack.us/img148/315/wprime32mzf6.jpg

WPrime 1024M secs
http://img529.imageshack.us/img529/3787/wprime1024mut3.jpg

the screens
E7200
http://img155.imageshack.us/img155/3316/3d06jt0.th.jpg (http://img155.imageshack.us/my.php?image=3d06jt0.jpg)http://img155.imageshack.us/img155/7808/aqua176412wg4.th.jpg (http://img155.imageshack.us/my.php?image=aqua176412wg4.jpg)http://img529.imageshack.us/img529/2391/cinebenchih8.th.jpg (http://img529.imageshack.us/my.php?image=cinebenchih8.jpg)http://img148.imageshack.us/img148/3221/pieverestwinrarop6.th.jpg (http://img148.imageshack.us/my.php?image=pieverestwinrarop6.jpg)http://img148.imageshack.us/img148/4470/pifastcpumarkfrtizwprimfd9.th.jpg (http://img148.imageshack.us/my.php?image=pifastcpumarkfrtizwprimfd9.jpg)

E8500
http://img155.imageshack.us/img155/9681/3d06th4.th.jpg (http://img155.imageshack.us/my.php?image=3d06th4.jpg)http://img148.imageshack.us/img148/5871/aqua198361gp6.th.jpg (http://img148.imageshack.us/my.php?image=aqua198361gp6.jpg)http://img148.imageshack.us/img148/581/cinebenchxk5.th.jpg (http://img148.imageshack.us/my.php?image=cinebenchxk5.jpg)http://img529.imageshack.us/img529/2460/pieverestwinrargx9.th.jpg (http://img529.imageshack.us/my.php?image=pieverestwinrargx9.jpg)http://img155.imageshack.us/img155/1058/pifastcpumarkfrtizwprimgt6.th.jpg (http://img155.imageshack.us/my.php?image=pifastcpumarkfrtizwprimgt6.jpg)

QX9770
http://img148.imageshack.us/img148/9336/3d06hn6.th.jpg (http://img148.imageshack.us/my.php?image=3d06hn6.jpg)http://img529.imageshack.us/img529/3889/aqua199785bt5.th.jpg (http://img529.imageshack.us/my.php?image=aqua199785bt5.jpg)http://img148.imageshack.us/img148/192/cinebenchjn1.th.jpg (http://img148.imageshack.us/my.php?image=cinebenchjn1.jpg)http://img155.imageshack.us/img155/3936/pieverestwinrarfw0.th.jpg (http://img155.imageshack.us/my.php?image=pieverestwinrarfw0.jpg)http://img155.imageshack.us/img155/7393/pifastcpumarkfrtizwprimwq0.th.jpg (http://img155.imageshack.us/my.php?image=pifastcpumarkfrtizwprimwq0.jpg)

[chinaguy]

Well a E7200 is on my shopping list for my HTPC. Looks like the extra cache doesn't make all that much difference for the extra $ for the E8500!

[belial]

thx so much for this test, been looking for something like this

good to know that games almost don't take advantage from the extra cache

[fng77]

Very interesting. Thank for your effort.

[fiveprime]

Have you run any demo's of games? Whats the real world performance difference like?

[CERO]

Nice comparison :up:, however would had like to see some comparsion in some Games if it was possible

[PcCI2iminal]

thx guys
Crysis ,Call of Juarez DX10 & Company of Heroes will be tested soon

:)

[philbrown23]

this is cool! it shows how much the cache actually does.

[MomijiTMO]

Thanks for the review. I so want a E7200 [or a 7300 by the time I get around to buying it].

[conzymaher]

Great comparison :)

But, Graphs that don't start at 0 are hard to interpret at a glance, this one for example:

http://img155.imageshack.us/img155/8139/aquaim5.jpg

I was like OMG WTF CACHE CAKES, then I realized the graph didnt start at 0 :)

[shine121]

nice comparison,just realize the difference between L2 cache
E8500 just want i need

[down8]

It would be nice if the graphs crossed at 0, so ppl could make a better judgment.

But, awesome information, nonetheless!

-bZj

[STEvil]

Graphs at 0? Why, didnt you learn to read them in school?

If you cared to look at the scale used starting at 0 would make some really really poor graphs, hence why the numbers are included (graphs are just quick visual references to most anyone who knows how to use them anyways).

[Solarfall]

excellent comparison PcCI2iminal.. i bet that alot of people will find this really helpful. :up:

[WhiteFireDragon]

so the only reason the quad wins in most is because it has double number of cores instead of double the cache right?

[Miwo]

Thank you for the writeup, lots of good info :) .
If its not too much to ask, do you think you could include some game benchmarks where you show MIN / MAX FPS?

You can buy an E8300 for $30 more than an E7200, I think the extra cache is worth it. You lose the 9.5x multiplyer, but i think 8.5x is fine if you can tweak your board well enough

[conzymaher]

Graphs at 0? Why, didnt you learn to read them in school?

If you cared to look at the scale used starting at 0 would make some really really poor graphs, hence why the numbers are included (graphs are just quick visual references to most anyone who knows how to use them anyways).

I think you'll find graphs started at 0 in school, Graphs for advertising / marketing start elsewhere...

If the graph starts at 0 you can look at the bars or whatever and say "hmmm that one is about 10% faster." What use is a graph as a "visual reference" if it implies that the e8500 is twice as good as the e7200

Also stop being so :banana::banana::banana::banana:y, you were incorrect...

[MadMan007]

It would be nice if the graphs crossed at 0, so ppl could make a better judgment.

But, awesome information, nonetheless!

-bZj

I agree, it's great to see a user comparison but the graphs skew the perception - 'marketing' vs 'real performance difference.' Scores as a percentage along with the absolute scores would give a more realistic impression.

[systemviper]

I see all these people hawking the E6xxx, E4xxx and E2xxxx chips and they all are
asking for $100, + OR - $20 and i think to myself, i could get a.....

Intel Core 2 Duo E7200 Wolfdale 2.53GHz 3MB L2 Cache LGA 775 65W Dual-Core Processor

for $126 delivered..... and it makes me wonder.....
New or Old, Think I take the New....

[Lestat]

dont make your graphs look like ther eis such a HUGE difference in the scores, its really confusing visually. the scores are very minor, make the graphs show the differences are only minor.

[Glow9]

The game benchies still going to come out?

[Leeghoofd]

Found this in a german hardware mag :

http://users.pandora.be/OAP/cache.JPG

[Glow9]

Found this in a german hardware mag :

http://users.pandora.be/OAP/cache.JPG

eh? 2fps difference seems pointless to upgrade, I'm kinda eh about this graph

[BNDMOD]

Very nice info bro..... this afternoo i'll read all this info slowly..... it's very interesting!

[Leeghoofd]

eh? 2fps difference seems pointless to upgrade, I'm kinda eh about this graph

It's at only 1.8Ghz sir, if you use the CPU's at 3ghz gaps will be bigger

[Glow9]

If they are al at 1.8, why would they different at 2, or 3? Seems like if there is differences they should be sorta consistent? No? I dunno. I guess I'd figure technology in the chips would be the determining factor *shrug*

[Leeghoofd]

If I ran a game on my 100% Oc'ed E2140 at 3.2Ghz and would try the same game on my E6600 or even Q6600 the gameplay would feel totally different... beleive me I have found out on a LAN myself... 1Mb cache is too low for gaming, from 4mb on differences start to become less noticeable...