LMAO a xeon E5450 is a monster

[Vengy]

Hello

Ive been playing arround with the 771 to 775 mod. And what i see is scary. I have a 6 core am3 6100 cpu in another system OC to 3.5Ghx with a MSI 970 Gameing mobo. And a ABIT IP35 Pro with a Xeon E5450 4 core cpu OC to 3.6Ghz. SO here it is I have been messing around to see what is the fastest i can get the Xeon not thinking there would be any chance that it would beat the shyt out of the am3 cpu. But after the test started and finished i was shocked. The AM3 cpu scored a 354 and my Xeon scored a 430 and im also able to get the Xeon to OC at 4.1Ghz but cant seam to make it stable. Works fine for half the test then here comes the crash lol. But still can any1 tell me why the Xeon at 4 cores beats the am3 at 6 cores.

 

[EarthDog]

Welcome!

not thinking there would be any chance that it would beat the shyt out of the am3 cpu.

You started off on the wrong side of the boat...

But still can any1 tell me why the Xeon at 4 cores beats the am3 at 6 cores.

Not sure what test you used, or what the systems hardware is like as it wasn't listed, but it was likely based on heavy floating point/integer calculations which due to the bulldozer architecture, leaves a lot to be desired.

 

[Vengy]

This was just done for some fun till i saw how much better the Xeon was than my Am. I used cenebench to test the CPUs im just not understanding how the Xeon seam tobe the better CPU when it is alot older. And if i was to go with the 8 core fx-8320 would it match or pass the Xeon?

 

[chrisjames61]

The XEON also retailed for $1,000 vs a $100 cpu. And you talk about scores. Scores in what? Is it a modern well threaded benchmark you ran?

 

[EarthDog]

Cinebench, surprisingly. Heavily multithreaded..no idea about modern if he ran 11.5 or r15. Still though in the story would be the same.

I think he is comparing clock for clock. The fact that it costs a grand 6 years ago has nothing to do with its performance clock for clock in this rendering benchmark... yet I still get your point, lol!

 

[chrisjames61]

Cinebench, surprisingly. Heavily multithreaded..no idea about modern if he ran 11.5 or r15. Still though in the story would be the same.

I think he is comparing clock for clock. The fact that it costs a grand 6 years ago has nothing to do with its performance clock for clock in this rendering benchmark... yet I still get your point, lol!

The 6100 was weak. I would think a 6300 at 4.5 GHz would trounce that XEON in most workloads. A better test would be a four core Kaveri at say 4.5 GHz cpu vs that XEON.

 

[RJARRRPCGP]

The 6100 was weak. I would think a 6300 at 4.5 GHz would trounce that XEON in most workloads. A better test would be a four core Kaveri at say 4.5 GHz cpu vs that XEON.

I bet an FX 8350 at stock decimates an OC'ed 775 CPU. My FX8350 without a doubt decimates my Q6600 in x264.

 

[EarthDog]

I'd like to see those tests at the same clocks to see what IPC is like...

 

[Waza]

FX cpus are strongest doing heavy integer calculations. That is compared to their own floating point performance. Their IPC still sucks tho.

 

[xrror]

Hello Ive been playing arround with the 771 to 775 mod.

I want more pics/bench/anything/information on this. I posted about the socket adapters for 771 to 775 in the Intel CPU area long ago, and pretty much nobody seemed to give any care about it

Also yea, you probably should have posted this in the Intel CPU area, though on the other hand those guys don't seem to care about anything older than Sandy Bridge

The 6100 was weak. I would think a 6300 at 4.5 GHz would trounce that XEON in most workloads. A better test would be a four core Kaveri at say 4.5 GHz cpu vs that XEON.

Welp, wish me luck in retrieving my tortured 775 rig to test this hypothesis. You've got me curious now.... I suspect the 775 rig will win this bet, but ... I gotta see if the thing even still works first. C0 stepping xeon on DDR2 board at 4.2Ghz.... errrr, there's a reason it got retired from Rosetta, mostly cause... how did it EVER work ;p

 

[q9650]

this is my Q9650 @3.69GHz

 

[poco242]

Here is my 8350 at a bunch of different clock speeds and different benchmark programs. He must have been using Cinebench 15 because of the numbers.

 

[q9650]

Here is my 8350 at a bunch of different clock speeds and different benchmark programs. He must have been using Cinebench 15 because of the numbers.



View attachment 175807

What were the single core scores for cinebench R15??

 

[poco242]

Never did any. I guess I didn't know I could, or how to go about it.

 

[poco242]

I just figured it out and tried it. On all 8 cores I got 703. On a single core I got 109. This is at 4.6ghz. It is lower than I got 2 years ago with the chart I showed, but I am also running Windows 10 now so I am not sure if that makes a difference. Also the OC profile I have saved for 4.6 may be slightly different than the one I used back then. Either way it should give you what you were looking for at that clock anyway.

 

[q9650]

yeah that's what i though. i was just wondering because i know overclocking does a lot for multi-core but not as much for single core scores. i get 104 single core with my Q9650 @3.7GHz but my multi-core pales in comparison to your FX-8350, lol.

 

[poco242]

AMD is not great at single core. But then again, I don't do anything I use it for that is CPU intensive with just one core so I am fine with that.

 

[Dolk]

Overclocking affects 100% of everything in the CPU.... it does not create bias benefits.

 

[q9650]

When i overclocked my CPU i had an increase of 40% in single core performance and 49% in multi-core performance. I know they wont end up being exactly equal but 9% seems like a big difference to me.

 

[TucoPacifico]

Running xeon E5450 @ 4104 MHz on ASUS P5Q3, 8gb DDR3 @ 1460 MHz, getting 6300 CPU score in Passmark 8 + R9 390 graphic adapter. CPU cooler is Noctua NH-D14.

Skyrim ENB + 112 mods smooth mostly 50-80 FPS, before OC (@3 GHz) had like 30-45 FPS. For it's age and price the old xeon is a beast.

If anyone needs my voltages I can post them. Running stable Prime95 small data heat generator @75C core max for 30 minutes, have not tested further.