XtremeCPU Benchmark

Quote:

---

Originally Posted by Readme.txt

-= Xtreme CPU Benchmark =-
Version: 1B
Created by: AgentGOD
Homepage: www.ultimate-filez.com
Remember to check forums/my website for new updates!

This program is a CPU/Memory benchmarking tool. It allows you to see your CPU's performance ratio levels (the RAM will also be crucial, as it allocates the calculations into your RAM).

My system has an AMD Athlon(tm) 64 3000+, overclocked to 2.73 GHz, 2 GB of Dual-channel DDR400 RAM.
I've scored 2469 XtremeCPU Marks, and took 68 seconds to complete.

Remember, only high end system may end up in the 3000-4000+ range, so don't be disappointed if you don't.

I'd say around 2000 is an average overclocked AMD system, and around 3000 are the extreme high-end ones on LN2 or DI.


How the program works:
----------------------
It runs through a series of tests, 1: Integer sorting (in order), 2: Calculating factorials, 3: Floating point calculation, 4: Linear graphing, 5: Finding prime numbers.
It calculates the score according to your time. The lower time it took to calculate, the higher your score will be.

---

V. 1B changes:
-Full dual-core support
-Added test
-Fixed scoring system
-Fixed 2 tests (now work as intended)

Screenshot:
http://www.ultimate-filez.com/xtremecpu-myscore.jpg

Download:
http://www.ultimate-filez.com/XtremeCPUBenchmark.zip

Good luck guys, and have fun overclocking!

Note: You may also post your results here.

Format: CPU - CPU Clock - Memory - Memory Type - The score - The time
Ex: Athlon 64 3000+ - 2732 MHz - OCZ Tech 2048 MB - DDR400 - 2222 XtremeCPU Marks - 27 sec.

fx-51-2200mhz-corsair xms regecc ddr400 3-3-3-8- 2000 XtremeCPU Marks - 30 sec

Awesome :) Old school FX-51.

Celeron M 1.5GHz - 2060 [15x137] - Transcend SO-DIMM 1024MB - DDR2-533 - 659 XtremeCPU Marks - 91 sec.

On my laptop w/ single channel ram (don't know if ram is running at the right speed either), does that seem right?

Quote:

---

Originally Posted by vapb400

Celeron M 1.5GHz - 2060 [15x137] - Transcend SO-DIMM 1024MB - DDR2-533 - 659 XtremeCPU Marks - 91 sec.

On my laptop w/ single channel ram (don't know if ram is running at the right speed either), does that seem right?

---

Looks about there. Check CPU-Z and see if your bus speeds are correct.

fx-51-2400mhz-corsair xms regecc ddr400 2-3-2-6- 2307 XtremeCPU Marks - 26 sec


:)

Nice score!

Opteron 170 - 2430MHz - OCZ EB 2x1GB - DDR500 3-3-2 - 2307 - 26

Quote:

---

Originally Posted by ahmad

Opteron 170 - 2430MHz - OCZ EB 2x1GB - DDR500 3-3-2 - 2307 - 26

---

tie :p:

Tomorrow I will fix up the "gaming simulation" benchmark, and also add another benchmark. I will also add support for dual-core processors. But for now, I must go sleep.

Opt 170 @ 2.8ghz, 2GB @ DDR468(OCZ Plat) 3-3-2-5-1T - 2400XSM - 25sec

3700+ San Diego@ 2.6GHz 2x512mb @ 250mhz 2.5:2:2:5: 1T - 28s

Athlon 64 3700+ - 2888 MHz - Mushkin Redline HP3200 - DDR524 - 2608 XtremeCPU Marks - 23 sec.

Your computer scored 2608 XtremeCPU Marks. You can compare your results with oth
ers on the forums.
Total time took to complete benchmarks: 23 seconds

2727 at 3.03 ghz Opteron 170, 22 seconds.

but I notice, this is only using 1 core.
any chance in a dual core version?

Winchester 3200+ - 2200MHz - 1.5gb @ 2.5-3-3-7 - DDR400 - 2000 XtremeCPU Marks - 30 seconds

may i suggest for the next version, the time to three decimal places?

4400+ with GSkill 3-4-4-8 at 238 x 11 = 2400

LOL! I'm too ashamed to post my score (201s) lol! Poor me, can't afford anything :( Donations ..

awesome. i wanna try this with my 170 soon. mmmmmm

2500 24sec
165 @ 2.7 2GB g-skill HZ @ 265Mhz

Athlon XP 1900+ @ stock - (1600mhz) - 384 memory "generic" - ddr266 - 1176 xtremeCPU Marks - 51 secs :clap:

:p:

Opteron 165 - 2800MHz - Kingston HyperX BH-5 - DDR510 - 2500 - 24 seconds

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

wow, that is really bad. and you got beat by a celeron M. :p:

At least I'm not the worst around here .. :clap:

Quote:

---

Originally Posted by Getttosmurf

2500 24sec
165 @ 2.7 2GB g-skill HZ @ 265Mhz

---

Quote:

---

Originally Posted by masterofpuppets

Opteron 165 - 2800MHz - Kingston HyperX BH-5 - DDR510 - 2500 - 24 seconds

---

Thats interesting. So timings play a huge role here.. timings and CPU freq (with A64).

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

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

Quote:

---

Originally Posted by masterofpuppets

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

---

Can you read or can you read? :stick: :stick: :stick:

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 :slap:

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.

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.

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

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.

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

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.

Huh, dual core mode?

Nick

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.

Ver 1B

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

Quote:

---

Originally Posted by 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.

---

Too lazy now. :p:

Nick

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.

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

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 :)

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!

Nice score for your CPU & Clock :slobber:

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.

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.

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!

http://img156.imageshack.us/img156/2...39061xd.th.png

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

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.

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.

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

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

Even though that 3.7 GHz one is just overkill :):cool:

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.

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.

Here's my score, stock settings.

http://img.photobucket.com/albums/v4...olie/score.jpg

My everyday settings:

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

Nice benchmark, AgentGOD :up:

Very nice. Destroyed an FX-60 :slobber:

My opty 146 :)
http://img426.imageshack.us/img426/5...ench1fv.th.jpg

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?

http://img114.imageshack.us/img114/1...dpwrmon3wb.jpg

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 :toast:

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

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

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

Interesting, I'll give my 170 a shot :)

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

http://i16.photobucket.com/albums/b2.../XtremeCPU.jpg

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

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.

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

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

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

http://img283.imageshack.us/img283/9...pubench8jn.jpg

Quote:

---

Originally Posted by 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

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

4352

http://i2.tinypic.com/xnzog2.jpg

Quote:

---

Originally Posted by AgentGOD

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?

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

http://img.photobucket.com/albums/v4...treme-test.jpg


:D

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

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

Quote:

---

Originally Posted by AgentGOD

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.

Quote:

---

Originally Posted by DaWaN

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!

Quote:

---

Originally Posted by SillySider03

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.

There is something wrong with the AMD dc scores too. 2.5 GHz seems to be the sweet spot...:D

Quote:

---

Originally Posted by Fred_Pohl

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.

---

Of course it doesn't produce meaningful results for intel CPUs. You'd have to be a biased idiot to believe that.

And I am 99% sure that he did not code it that way. Like I said earlier, I had the same problem with a benchmark that I wrote. After some research, it turned out to be how the program was compiled. I ran in a DOS window, which use both 16-bit and 32-bit instructions. For whatever reason, AMD CPUs are tons faster at processing 16-bit instructions. Although it's fairly interesting, that shouldn't be for a reason to say that AMD is better than intel mainly due to the fact that 16-bit CPUs and OSs are obsolete.

Quote:

---

Originally Posted by derektm

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

4352

http://i2.tinypic.com/xnzog2.jpg

---

??? :eek: WTF, how is that possible :slapass: :confused: :slap: :toast: :stick:

This is my score :fact:
http://img435.imageshack.us/img435/814/37119wv.jpg

heh - inetresting benchmark, here is my score -
A64 2800+ (s.754) - 2000MHZ - OCZ Gold DDR500 512 (2*256 mb, single channel) - ddr500 - 1555 XCPUM - 108 sec

think that score is normal, but time? :confused:

Sticky?

I want to see a conroe on this! ;)

It would be interestimg to see how deep the bias goes!