XtremeCPU Benchmark

[Obelisk79]

Something of note... as I mentioned in a previous post, I'm new to this over-clocking thing. As I explore the limits of my hardware I have been documenting in a notebook my benchmark scores. I see that this particular benchmark is a better indicator of overall system performance for the CPU+MEMORY. This test is highly dependant on mem clock.

Eg. I posted a score of 2183 @ 2300mhz with my mem clocked at 230mhz
I then enabled a 166mhz divider reduced my ht frq to 4x and raised my HTT
to 240mhz aka 200mhz mem clock and 2400mhz cpu. I then post the test
and score 2073.

Just food for thought. I still like the proggy however.

[AgentGOD]

Yes, this program is VERY CPU and Memory dependent, because allocating data to memory and reading + calculating + storing, are CPU & Memory operations. It gets very intensive. Either the CPU or RAM slows down, then the overall benchmark will become slow.

[grahamwoolcott]

Here's my score, stock settings.

[bachus_anonym]

My everyday settings:

Opteron 146 - 3504MHz - 2GB Corsair PC4400PRO @ DDR500 5-2-3-2.5-1T - 3293 XtremeCPU Marks - 51 sec

Nice benchmark, AgentGOD

[AgentGOD]

Very nice. Destroyed an FX-60

[neoman]

My opty 146

[Daddyjaxx]

I'm confused with this program. Why is the time that it took mine to finish was longer than most, but my score is higher?

[AgentGOD]

It runs 2 tests (on seperate threads) simultaneously and then at the end adds the score to find the total score. That is why it may appear to run slower.

But other than that, awesome score

[AgentGOD]

Let's see how an Intel Conroe(r) would score...

[masterwoot]

Opty 146 - 2900 MHz - OCZ 2GB - DDR480 - 2584 XtremeCPU Marks - 65 sec.

[dr.house]

Opty 170 2800 Mhz, G.Skill at 280 HTT and 4x LDT 4/5 divider. 3498 XtremeCPU Marks 96 Sec.

[alpha0ne]

Interesting, I'll give my 170 a shot

[alpha0ne]

Opteron 170 - 2800 MHz - G.Skill HZ 2048 MB - DDR560 - 3731 XtremeCPU Marks - 90sec

This is with a number of app running in the background, i'll try again later with just this benchie running to see how much variation



Its going to be pretty hard to beat 51s Bachus !!!

[PORTUGAL]

i have a problem,i can run the program but at the end the program close imediatly,i reduced my overclock but problem is the same.

[AgentGOD]

Strange. Are you sure you're running an appropriate operating system?

[BigDov]

Athlon 64 3200+ Newcastle @ default 2.0ghz
2gb g.skill zx

1886 marks, 89 seconds

will report back when my new bios chip for my other motherboard arrives, and I can get through the 210mhz barrier this current board has

[OmegaMerc]

2.75ghz Opt 170 / 2gb OCZ plat DDR500 / 3.3.2.8.1t - 3445 XCPU Marks 98 seconds to complete.

[BigDov]

Athlon 64 3200+ Newcastle @ default 2.0ghz
2gb g.skill zx

1886 marks, 89 seconds

will report back when my new bios chip for my other motherboard arrives, and I can get through the 210mhz barrier this current board has

Just remembered this.... ran it again at 225mhz/2.25ghz

2210 marks, 76 seconds

[derektm]

Opteron 170 @ 10x250, gskill 4000hz, 1:1

4352

[Fred_Pohl]

Let's see how an Intel Conroe(r) would score...

What is your agenda here? I ran your BM on my 3.0 P4C @3.6G and scored 277 XtremeCPU marks in 1229 seconds. In 3DM01, 03, 05, 06, Q4, etc. my system equals a A64-3500+ system and in PCM05 and FEAR it beats an A64-4000+ system. WTF dude?

Conroe is 20-25% faster than X2, clock for clock, in almost every benchmark. Are you hoping to provide a BM where reality takes a backseat to AMD fanboyism?

[IndianScout]

X2 3800 @ 2.5ghz 2.5/3/3/7

[mtzki]

It is nice to play with this one...good length etc. May become popular if it scales well.

X2 4400+ - 3158 MHz - Mushkin Redline 2*512MB - 263MHz, 2-2-2-5 - 3861 XtremeCPU Marks - 87 sec.

[Fred_Pohl]

Let's see how an Intel Conroe(r) would score...

Obviously you subscribe to PT Barnum theory and judging by the posts thus far, you're correct.

Nice job of creating an AMD only synthetic BM that has no basis in reality.

[Fred_Pohl]

This benchmark doesn't seems to be fair for netburst......

Bingo! That was obviously the author's intent. AgentAMDisGod obviously went out of his way to deliberately write this BM to run slow on netburst.

My 3.6G P4C system matches A64-3500+ systems in gaming BMs (using identical vidcards) and A64-4000+ systems in PCM04/05, encoding, etc, yet scores less than 300 in this joke of a BM!

[Fred_Pohl]

Do you mean a joke as opposed to the benchmark you wrote, or in general? Why not try to offer some help, or point out what you feel the issue is, rather than calling the bench a joke. And to be honest, I don't think the authors intentions were to make a bench that favors one chip over another, but I may be wrong. Either way, I like it and my scores rock - Boo Yah!

And could you explain the P.T Barnum theory please?

This BM is totally broken with Netburst, the author has been made aware of it and hasn't even responded. IMO his apparent apathy about that fact indicates that he either doesn't care or wanted it that way. If he wants to make a BM that only runs on AMD cpus, that's fine but he should at least advertise it as what it is. An AMD only BM.

It seems to me that in order for this BM to produce scores that are ~700% low for Netburst, it would almost have to be intentionally written that way.

P.T. Barnum said that there is a sucker born every minute and IMO that adage applies to anyone who believes that this BM produces meaningful results for all x86 cpus.