XtremeCPU Benchmark

[WeakSauce]

2400 25sec 3400+ s754 @2.695ghz ddr 490 2-2-2-5

[masterofpuppets]

Well it's a CPU bench, so timings dont matter

[ahmad]

Well it's a CPU bench, so timings dont matter

Can you read or can you read?

He has got a proc at 2.7GHz and you got one at 2.8GHz. Scores are identical, what does that tell you? Sure tells me it aint just clocks here

[AgentGOD]

Today, I will update this program to make it multi-threaded , I might also fix up a few benchmarks. (Remember, after the update, current scores may not be compared).

And yes, as I've mentioned above. Your system RAM is crucial in this test, as it stores the benchmark data into the RAM and reads it off of it (during tests and at the end). Your CPU will be doing all the calculations.

[AgentGOD]

Updated to v.1B

Changes:
-Full dual-core support
-Added test
-Fixed scoring system
-Fixed 2 tests (now work as intended)

Just a reminder: You cannot compare old version scores with the new version. The scoring system is different.

[masterofpuppets]

It's slower with dualcore.. 2680 with it off and 2511 with it on.

[AgentGOD]

Please do remember that when using dual-core mode, it will also use double the amount of RAM at the same time.

Hmm.. I've also noticed a screw up in the dual-core score calculation algorithm. Re-download Version 1B please.

[NickS]

Rig in Sig:

AMD Athlon 64 3200+ Venice s939 - 2510 - Crucial Ballistix 2x512MB - 251MHz 2.5-3-3-8 - 2068 - 29s

Screenshot

AMD Athlon XP 2000+ Palomino - 1664 - Nanya 2x256MB - 166MHz 2.5-3-3-7 - 1250 - 48s

Screenshot

School Rig: Intel P4 2.8GHz Prescott - 2.8GHz - 2x512MB - DDR400 - 204 - 294s

Screenshot ^^Somethings gotta be fuggered.

School Rig: Intel Pentium M 1.6GHz Dothan - 1.6GHz - 2x512MB - 631 - 95s

Screenshot

[AgentGOD]

1: Do not use Dual-core mode when you only have 1 core.
2: Looks like AMD is better at Integer and floating point, just like how the dual-core duel turned out.

[NickS]

Huh, dual core mode?

Nick

[AgentGOD]

Dude... you need to download Version 1B. You have 1A. Told you guys already, you cannot compare different scores from different program versions. That's like comparing 3DMark2001SE and 3DMark06.

[mnewxcv]

Ver 1B

fx51 2475mhz corsair xms @ 225mhz 2-3-2-6 - 2210 - 76sec

[NickS]

Dude... you need to download Version 1B. You have 1A. Told you guys already, you cannot compare different scores from different program versions. That's like comparing 3DMark2001SE and 3DMark06.

Too lazy now.

Nick

[pete5990]

Version 1B

Celeron D 335 at 3.7GHz - 256MB PC4400 2.5-3-3-8 - 258 - 649 seconds

I think that L2 Cache must play a huge part in the second test, because it killed my score.

Test 1 - 5 seconds
Test 2 - 536 seconds
Test 3 - 34 seconds
Test 4 - 71 seconds
Test 5 - 1 second

Maybe the program has something against the prescott core? I've noticed that most of the P4 scores are also pretty low.

EDIT: Actually I've noticed this with my DOS based cpu benchmark, but the difference between AMD and intel isn't as drastic as it is here.

[BlueWonder]

All of the Intel scores are low. Even the dothans are getting ridiculous scores.

[Kunaak]

Well, theres some Improvements, and theres some errors.

1. it's definatly using Dual Cores now.

Single core Score 2727, both versions.
Dual Core Score 3612, V1.b

so thats good.

however, theres a bug now, the clock timer is waaaaay off.
when running single core, it said it took 60 seconds.
when running dual core, it said it took 181 seconds.

in reality, it takes about 30-40 seconds to complete, so it needs alittle reworking.

My Suggestions.

Clearly put the Version in the Title Bar.

that way theres no chance of confusion as to what version someone is using.
helps avoid confusion, helps avoid cheaters using early versions that gave much different results.

If this benchmark is drastically different from AMD to Intel, then maybe it should be made as a AMD benchmark only.

Rework the Clock timer.

from there, I think your onto one hell of a nice benchmark.
just needs alittle work, but thats what we are here for, to help you get it going good

[Obelisk79]

AMD Athlon 64 3200+ "Venice" at 2.3GHz - 1GB PC3200 Geil 2.5-4-4-8 - 230HTT - 2180 XtremeCPU Marks

Xtreme CPU Benchmark
Version 1B
Created by AgentGOD

This benchmarking software is based on integer and floating point perfor
mance on your current PC. The score given at the end of 5 executions will be bas
ed on how well your PC performed.

Dual-core processor? (1 (yes)/0 (no)): 0
Press Enter to begin benchmarking...

Execution 1: Integer sorting...
Execution 1 Finished after 3 seconds.

Execution 2: Solving factorials...
Execution 2 Finished after 7 seconds.

Execution 3: Floating point performance...
Execution 3 Finished after 29 seconds.

Execution 4: Extreme Graphing (simulates games' high intensity)...
Execution 4 Finished after 36 seconds.

Execution 5: Finding Prime numbers...
Execution 5 Finished after 0 seconds.

Your computer scored 2180 XtremeCPU Marks. You can compare your results with others on the forums.
Total time took to complete benchmarks: 77 seconds
Note: Different versions may not be compared!

[AgentGOD]

Nice score for your CPU & Clock

Thank you all for your support. I will be sure to fix up the timers.


Okay, I fixed the timer . It should not be an AMD-sided benchmark. I guess it is just showing that AMD definitely outwitts Intel's prescott/northwood core processors. Let's see when a yonah or conroe takes role

Re-download it.

To code optimization geeks: For those of you who believe C++ headers, think again. I wrote most of this program using C code, except the part where I had to utilize the Windows API.

[Obelisk79]

My score stayed the same with that last update... heh once I figure out how to properly overclock my stuff hopefully I'll be able to post some really nice scores.

[ixtapalapaquetl]

Opteron 170 - 3250 MHz - 2x512MB Mushkin Lvl II - 271MHz @ 2222 - 3906 XtremeCPU Marks - 86 sec

Single core version maxed at 2727 last night; dual core version 1B obviously working now. Disregard CPU temps - sensor is borked. Might do some tweaking and shoot for 4k before the FX-60 on phase guys show up!

[AgentGOD]

Awesome! We get to see how some very high-end CPUs score tonight!

[masterofpuppets]

Dual-core mode:
Opteron 165 - 2800MHz - Kingston HyperX BH-5 - DDR510 - 3360 - 100 seconds

Single-core mode:
Opteron 165 - 2800MHz - Kingston HyperX BH-5 - DDR510 - 2665 - 63 seconds

Really odd, it says it's taking less time in single-core mode.

[Kunaak]

Idea...

in single core mode, the clock is reading 1x.

in dual core mode, it's reading 2x for the clock.

so take the score above...
it says 100 seconds (which is obviously wrong) but if you cut the time in half, then it would be 50 seconds... which seems alot more in line with what one should expect going from single to dual core at the same speed and such.

[Kasparz]

Athlon 64 3700+ - 2888 MHz - Mushkin Redline HP3200 - DDR524 - 2846 XtremeCPU Marks - 59 sec.

Your computer scored 2846 XtremeCPU Marks. You can compare your results with oth
ers on the forums.
Total time took to complete benchmarks: 59 seconds

[AgentGOD]

Damn... nice. That 3700+ is a monster .

Even though that 3.7 GHz one is just overkill