Measure your Dual/Multi Core Notebook CPU Speed - Bye Bye to SuperPI

NB: Asus G1
T7200 2,0GHz
Time: 41,324s

Vista Home Professional 32-bit

A thing i noticed to the undervolting rightmark CPU clock -users: I got about 0,9sec better result from shutting it down.

 

Lenovo X61 Tablet (L7500 @ 1.60Ghz)
Windows XP Pro w/ SP2
wPrime Time: 56 secs

 

Forcebook Compal JFL92, WinXP SP2 + Hotfix

wPrime 1.62: 34.734s
wPrime 1.55: 34.702s

Why do I get such slow results? Other T9300 got 4secs better. That's big difference.

 

Can someone provide me a wPrime v1.61?

Ok, found it. Sorry

 

Desk Top Computer
OS : XP pro SP3
Intel(R) Pentium(R) 4 CPU 2.26GHz
Core: Northwood B0
Clock: 2259.2 MHz

FSB: 531.6 MHz
L2 Cache: 512 KB

Memory: 1024 MB DDR
Frequency: 132.9 MHz @ 2.5-3-3-6

Mainboard: IBM IBM

32m(speed test) : 150.019 sec

 

Dell Latitude D610
Intel Pentium M CPU @ 2GHz
Windows XP Pro SP2
Time: 92.109 sec

WPrime 1.63 also reported my clock speed incorrectly when using "performance on demand" in RMclock and always reported the stock voltages rather than my undervolted ones.
Here's the full set of tests I ran calculating to 32M:
wPrime 1.63
Undervolted: 94.609 sec
No management: 94.843 sec
Rightmark closed: 92.812 sec

wPrime 1.55
Undervolted: 93.984 sec
No management: 93.812 sec
Rightmark closed: 92.109 sec

 

I got like 12 seconds for my new desktop before I came to work, looks like I have to grab another top spot

I sold the opteron or I could have pushed it higher for #1 on single core too.

 

LOL. nice... whats the specs of your desktop rig?

 

Until Mr. Vicious gets home from work I believe I hold the record now:

That's at stock, on the desktop in my signature. Running Vista Ultimate SP1 x64.

Update: System 2, my alternate desktop. Specs:
Athlon 64 X2 4200+ EE @ 2.42 GHz
2GB DDR2-675 @ 5-5-5-15-1T
GeForce 7600GT @ 575/752
XP Pro SP3

Not record breaking but fairly interesting that it throttles my ThinkPad. Speaking of which, I'll post that one in a minute.

Updato numero dos: System 3, my ThinkPad T61
Specs are in sig. Running Vista Ultimate SP1 x64.

No record breaking going on here either, but it's pretty gosh-darn close to the record T7300 score, so I'm happy.

Reviewing the compiled results table it's interesting that AMD's K10 looks very competitive. Very competitive. This is consistent with some other emerging true-multi-core benchmarks, including POV-Ray real-time benchmarks, LAME MP3 encoder, 7-zip's benchmark, etc., though it defies most traditional benchmarks (3DMark, PCMark, WMP/AAC encoders) that show Intel with a massive lead. I predict that as we get more true multi-core benchamrks - ones that push every core to 100%, not just to 50-70% like the half-arsed PCMark and 3DMark Vantage, the Core 2's pseudo-quad-core architecture and miserable FSB-based memory controller are going to start to show their ugly side.

Hopefully Nehalem clears that up. I may not be too fond of Intel's rather questionable business tactics, but you got to admit the level of competition out there right now is good for the consumer.

 

good ol' fashioned q6600, the new quads are about 15% faster at the same clock, can possibly clock higher, and run cooler. However being that I was building a budget system the q6600 still has the kings crown when it comes to performance per dollar.

 

Wow, looks like the other T5500's on here blew my Toshiba P105 away.

I joined the team on HWbot and submitted my score, then it vanished. Anyway, the 32 test was 55.120 seconds.

Edited with best score after closing tray programs.

 

You can boost scores by doing a few things, killing anything that is eating resources helps some, x64 OS seems to help some in most cases, ram speed/timings have a good impact on this (and I lack here), Im not so worried that I am going to go out of my way to tweak for a better score tho.

 

Actually halfhalo holds the record. It's on page 54. I think it was around 12 seconds?

 

Overview:



32M (Speed Test) Log:



1024M (Stability Test) Log:



Sager NP9262, Stock Speed (Lower Clock Speed And Voltage On The Image due to the benchmark being over) and Cooling, No Undervolting, Windows Vista Home Premium 64-Bit, Arctic Silver 5. ^.^

Edit: Added the screenshots of the wPrime logs.

 

Ah, so he does. I certainly hold the "stock" record, though - beating out the Q6700 by around a second. Not too shabby.

My ThinkPad also nabs the second-highest T7300 spot.

 

Nevermind, I'm a dumas.............Score for the X205 listed in my sig:

32: 29.654
1024: 944.658

 

Ok I am home!

I'll beat this later

 

And here I was expecting some amazing feat of overclocking, and instead you just used the wrong version of wPrime. Care to try again with the stable version? I was really eager to see how my Phenom 9850 matched up with the Q6600, but 1.63 won't run on my desktop. On the plus side if I use 1.63 on my ThinkPad I capture the T7300 performance crown by over a second and a half!

 

Says 2.4ghz but it was at 3.6ghz

I googled wprime and thats the version I found. I dont know how it can be the wrong version...

 

Here's a direct link to 1.55, the latest "stable" version:
http://www.wprime.net/?q=download&f=wprime_155.zip

From the wPrime homepage:

So I figure we want to follow the same general rule here and all use 1.55 as well. 1.6x has been acting up for a lot of people, giving weird results - I got it to run on my desktop finally but the threads were calculating at drastically different speeds (2 and 4 finished before 1 and 3 even got to 75%) and my time almost doubled from 1.55.

 

This is on v1.63. Penryn T9300. This is when the notebook is on idle. I couldn't be bothered to close everything just to get a score a second better.

http://img393.imageshack.us/img393/5396/wprimeor0.png

 

I read what was posted about 1.55 vs 1.63, downloaded 1.55 on my older Toshiba and lost a second on the 32 test. I noticed that on both tests, core 2 seems to lag anywhere from 1 - 2 seconds behind core 1.......is this normal?

 

I noticed that too. Core one reaches 100% like half a second after core two (according to the CPU meter). I know during the test, core two usually got just under a second lower than core one. Core one usually finished at like 29.xx seconds, whereas core two finishes at 30 seconds or so (for me), thus lowering my score.

 

Specs in sig. Not too shabby for a stock 5550. Cant wait for the 9300 to get here tho!

 

X9000 ES in D630....3.8GHz x19.0 Vista Ultimate x64 (temp was above 80*C without the cooler.. )
Have sold it now, unfortunately never bothered to take a screenshot of CPUz..

View attachment 20568 View attachment 20567

 

Celeron M 420 1.6GHz (OC - 2.0GHz.. )....Acer Aspire 3680
2GB RAM....XP SP2 x86-32
91.874 seconds.. (32M - Speed Test)
A bit ashamed to upload the screenshot..

 

Ok retested with 1.55

11.7

 

Is that testing 4 threads or just 2?

 

I can bet a million bucks on that being 4. It's just impossible to get 11 seconds with two cores....

 

Right, and I'm new to all of this stuff......but I remember reading something somewhere about having to set the number of threads if you're running Vista. Was just wondering if running 2 threads on a processor like his (Quad Core) might give him a lower time. Not accusing him of cheating (he's a Jarhead, probably couldn't cheat if he wanted to anyway ) but just wondering if by failing to make the change it could have given that time.

 

Man....11 seconds at a clock speed of 2.4GHz (does the FSB and L2 cache also come into play here..??)....That has definitely got to be 4 threads..
I wasn't able to push my cpu any further than 14 secs..

 

it was 4 threads @ 3.6ghz the speed step always downclocks me back to 2.4 when not under load so it reads the cpu as 2.4ghz.

Just for the record I tried 2 threads and got 22.7 first run.

 

yeah I know thats a pain....I had a similar problem too, unless I disabled EIST/C1E in the BIOS, check out if you have that option, though my BIOS was custom made....

 

I dont think its a problem, I want that function. Most say turn it off for stability but if I am stable I have no reason to turn it off.

I just ran a few sweeps of prime95 and under 100% load for all 4 cores im stable and temps are great.

I just droped my ram from 5-5-5-15 (safe) to 5-4-4-12 (epp) so it shaved off abit more time, and I think the cpu would go higher still, but 3.6ghz is a great 24/7 setting and I am not too concerned with taking my time for a max for a few benchmarks.

So with the timings lowered it looks like 11.67

 

I'm so happy right now, I forgot that AS5 improves over the first 200 hours. Before overclocking (when I completed my build, my cpu loaded at about 42C, and now, after a day of use, and after overclocking my cpu loads at 36C with a stock intel cooler!
My E6420 OC'd from 2.13ghz to 2.8ghz:

 

AMD X2 RM-70, 3GB ram,
40.4 secs on 32M

This is in a Toshiba Satellite P305DS8828 (new "PUMA" platform, with ATI Radeon 3100 graphics card)

 

Here are the score for the Turion X2 Ultra ZM-86...

 

Laptop #1-Fujitsu E8210 Intel Core2 Duo T7400 Windows XP Pro SP3.
wPrime 32m -37.828 secs.

Laptop #2-Fujitsu E8410 Intel Core2 Extreme X9000 Windows XP Pro SP3.
wPrime 32m -27.671 secs.

Laptop #3-Sony Vaio VGN-NR310E Intel Core2 Duo T7500 Windows Vista Home Premium SP1.
wPrime 32m -35.987 secs.

 

v1.63 using T9300. Undervolted, stock speeds, 2 threads. Couldn't be bothered to shut down all programs.

http://img61.imageshack.us/my.php?image=m1530we6.jpg

32m time: 30.264 sec

 

AMD Athlon(tm) 64 X2 Dual-Core Processor TK-55
Core: Tyler
Clock: 2250 MHz
Vcore: 1.15 V
FSB: 1000 MHz
L2 Cache: 512 KB
windows vista x64
Memory: 1024 MB DDR2
Frequency: 375 MHz @ 5-5-5-15

32M speed test
best score 39.562 sec

tried with xp x64 and shaved 0.01 of the time lol

 

Intalled SP3 for my XP machine and dropped 5 full seconds off the time. No other changes made.......went from 55.120 down to 50.188 (ok, so not 5 FULL seconds, but close enough for gov't work)

 

I went up from 11.67 when I was FSB 400 x 9 for 3.6ghz and ram at DDR2 800 5-4-4-12

to FSB 450 x 8 for 3.6ghz and fram DDR2 900 5-4-4-12

I dont understand how overclocking the ram & fsb didnt cut my time down some.

 

did the test again this time on both
windows XP x64
1G ddr2 running at 750MHz
athlon X2 duel core running at 2250MHz

32M test = 39.593 sec
1024M test = 1269.046 sec

 

11.217 Is my new lowest. 3.8ghz on the q6600

 

I keep getting "unexpected error" when I try to run wprime. The program automatically quits on me before it even gets a chance to run. Specs in the sig. Any reason for why this is happening?

EDIT: I'm using v1.55 of wprime.

 

"unexpected errors" means you are probably using Vista.

Note 1: (for all Windows versions w/ multiple user accounts) you need to have an account with Admin privileges (right-click on program or link and Run As Administrator) to run this benchmark.
Note 2-a: (for Vista) you need to manually set 2 threads under the advanced settings. Otherwise the times will be doubled.
Note 2-b: (for Vista) you might need to disable the User Account Control & any user passwords if you get an "Unexpected Error..."

 

Well Note 2-b did the trick for me. It's kind of odd to see there aren't very many penryns listed under the hall of fame.

 

there are a lot, I havent gotten around to update it yet.

 

Asus EEE 901
Windows XP Home SP3
1.6GHz Intel Atom with Hyperthreading
2GB OCZ DDR2-800 ram
12GB SSD



2nd time after closing some things and turning on High Performance mode:

 

This is the kind of crap that drove me nuts when I was running Vista Home Premium. No matter that I disabled the UAC - numerous times, I would attempt to delete a folder or move files around in Documents and get a prompt something to the effect that...'I didn't have permission to take such and such and action'.

Switching topics..

My 'old' laptop nearly top of the line when I got it, a Chembook 2056 [1.7 GHz Intel Pentium M735 (Dothan - 2048 L2 Cache) 128 MB Dedicated Video Ram - (Ati Radeon 9700 Pro w/ 4X AGP 3D Graphics Accelerator)60GB HD (5400 RPM) 512 MB DDR333 (PC2700) SDRAM (512 MB x 2)] took on wPrime and performed the task in a blazing 119 sec.

I just put together a desktop comprising of an Ultra XBlaster Mid-Tower with 650 Watt Power Supply/ Intel Core 2 Quad Q6600 2.40GHz/ 2GB DDR2 PC6400Seagate Barracuda 7200.10 250GB HD - 16MB, Serial ATA-300/
XFX GeForce 8400 GS Video Card - 512MB DDR2. As you can see, I spared no expense on the video card.

In any event, just for yuks I ran xPrime in both 32M and
1024 M. 32M yielded 19.1 sec. 1024M yielded 566 sec. This unit is running XP Pro.

These scores are mitigated by the fact that my copy of XP, well...Uncle Bill doesn't recognize it as legitimate copy thus, I am not able to download Microsoft hotfix 896256 ( http://support.microsoft.com/kb/896256), which supposedly addresses the problems that all notebooks and/or desktops - unless configured correctly will not support dual cores in Win XP.

Definitely time to get another notebook. All of my USB ports are not functioning anymore thus I have to use a PCMI/USB Cardbus in order to transfer data.

Mr. Vicious XUSMC's review of the C90S barebone kit sounds like and intriguing way to go especially since I had success 'building' the Ultra XBlaster from components. And then there is the Asus M50VM-A1....All there ready to go. xPrime score pretty decent. Decisions, decisions.

Dick