6625W users please test

Hi 6625W owners,

From what I've seen on this forum, it does seem that the BIOS which handles ACPI functions and cpu/front side bus speed a bit ... well I wouldn't say faulty, but rather not as matured..

So I would like to test something basic, if you would.

http://superpi.ilbello.com/ -- The superPi of course, first making sure that the CPU is performing on par with others. Please state your Model (or real clock frequency if possible)

For example, figures from Mobile01.com ( http://5i01.com/topicdetail.php?f=236&t=367162 )

Dell XPS M1330 (2.0GHz Intel Core 2 Duo T7300) 0m 58s
Lenovo ThinkPad X61 (2.0GHz Intel Core 2 Duo T7300) 1m 01s
Lenovo 3000 V200 (2.0GHz Intel Core 2 Duo T7300) 0m 59s
Lenovo ThinkPad T61 (2.00GHz Core 2 Duo Intel T7300) 0m 59s
Dell Inspiron e1705 (2.0GHz Core 2 Duo) 1m 02s

So it'd be great if 6625W can do it under 58s, while if it's something like 1m 02s plus then maybe there could be rooms for improvement.

http://www.wprime.net/download.html -- Similarly for wPrime as well

I came across some results( http://www.tabletpcreview.com/default.asp?newsID=980 ), which suggested to me that zeptos do seem to have an issue with it..

I've extract the interesting bits shown below..

Notebook / CPU wPrime 32M time
Samsung Q35 (Core 2 Duo T5600 @ 1.83GHz) 46.274s
Zepto 6224W (Core 2 Duo T7300 @ 2.0GHz) 45.788s
Samsung X60plus (Core 2 Duo T7200 @ 2.0GHz) 44.922s
Zepto 6024W (Core 2 Duo T7300 @ 2.0GHz) 42.385s
Samsung Q70 (Core 2 Duo T7300 @ 2.0GHz) 42.218s
Systemax Assault Ruggedized (Core 2 Duo T7200 @2.0GHz) 41.982s
HP Compaq 6910p (Core 2 Duo T7300 @ 2GHz) 40.965s

We can easily see 6224W is 'slower' than T7200 which suffers from a lower bus frequency when compared with T7300..

I just hope it's not the case for 6625W now, seeing how it has updated BIOS all the time etc.

Would be great if we can get to see some real figures...
What do you guys think? Any better ways to test the functionalities of BIOS to make sure it's working as it should and efficiently?

 

T7500 2.2GHz

SuperPI 2M: 1m 02s
Wprime 32M: 42,13s

I'm disappointed to see that Lenovo and Dell with T7300 scores better at lower MHz

But I must add, that I haven't installed the new BIOS, which enables IDA support! So the SuperPI "could" be better!

 

well, I dont know wich of the superpi calculations those results are from, so its hard to test...

 

I'm responsible for a number of those results so I will comment on the testing conditions: Usually on a new computer with any unwanted software removed and before more is installed. Also background software such as anti-virus was closed down.

Today the best my 6024W can manage is:

SuperPi: 1m 04s
wPrime: 44.499s

However, 1 have about 10 applications open and Task Manager (processes page) shows the system idle process bouncing around between 89% and 94%. That means that various processes are taking 6% to 11% of the CPU time. Getting these other processes down to 2% - 3% would improve the test scores.

IDA should help the SuperPi score. However, today I'm not seeing any hint of the CPU using the IDA mode, perhaps because the other processes are using the 2nd core enough that the IDA doesn't cut in.

I don't think there is anything inherently slow with the Zeptos (apart from the tardiness in getting the IDA supported in the BIOS). It's all the background processes which affect these benchmark scores.

John

 

T7700 2,4ghz

superpi 2M -> 56,902s
wPrime 32M -> 46,25s

I did have some applications open...


now without other programs running and rmclock disabled:
superpi 2M -> 56,579s
wPrime 32M -> 43,406s

 

well, I have the 011 for 6625wd, and RMClock running, but my processor doesnt ever go to IDA when running superpi, and my results are like 1m 18s with the T7100, but something isnt quite right, becos the processor isnt fully used. And I had quite a few background programs running

 

IDA is a joke... core 2 must be COMPLETELY idle before it kicks in. When is it completely idle? almost never

I'll try SuperPI and Wprime again later, with all running apps closed

 

Also IDA isn't officially supported in the Core 2 DUO CPU's until the Penryn according to some sites online.

 

John is very knowledgeable

He actually posted this in another thread ftp://download.intel.com/design/mobile/datashts/31674504.pdf

It's the Core2Duo data sheet, according to that in order to have IDA activated 3 criteria must be met.

-Exposure, via BIOS, of the opportunistic frequency as the highest ACPI P state.
-Enhanced Multi-Threaded Thermal Management (EMTTM).
-Intel Dynamic Acceleration Technology mode and EMTTM MSR configuration via BIOS.

All these seem to point to BIOS, I am suspecting the latest N11 some users are complaining about the noise etc, is possible due to bug in BIOS where it never let the CPU goes into low power mode?

Anyway, I have not purchased a machine yet, so it's pure speculation on my part... hence the reason for this post to learn from the ones that have the machine to test.

Thanks a lot for the above that have replied!

 

Could be, but we dont get an answer

 

T7500 2,2ghz

superpi 2M -> 57,128s
wPrime 32M -> 70,042s

So that's quite weird

 

background programs...

oh, btw, having IDA enabled, sets the processor going to 1.1750V (when it should be 1.0000V) and my temps while gaming went to 90C...
But I think this is either RMClock or BIOS issue

 

No, the Wprime score is because you haven't set the number of threads. Look in advanced settings and change that number to 2, i'm sure you'll get a MUCH better score

 

Thank you!

superpi 2M -> 57,128s
wPrime 32M -> 41,062s

 

:yes:

Actually, I have in the past, on some Santa Rosa notebooks seen the IDA speed jump shown by CPU-Z briefly go up to the IDA speed.

However, it's a bit like looking for some of the sub-atomic particles: Blink and you have missed it. As already noted, it's rare for the second core to be idle for very long.

Perhaps someone can configure their Windows (MSconfig, Boot, Advanced options or, if available, in the BIOS) to use only one core and see if that gets the IDA to run for longer since the other core will be always off? (I'm too busy at the moment).

John

 

well with the 011 and rmclock beta2 I get my T7100 to go to 700MHz in SLFM, so that works

 

But what about the fans, did undervolting help on the noise? in Bios 011

 

well they turn on a bit less now when its in SLFM, depending on what your doing. And I have mine undervolted to 9.0x 0.9875V

 

I guess it's only useful for one single application (ie. games) which doesn't support SMP and taking up all the cpu time, that way the 2nd core may have a chance to drop all the way back down to p-state thus enabling IDA..

 

I spotted evidence of IDA today. I had CPU-Z open and occasionally the speed was going up to 11x (on 6024W with T7300). This was under fairly light load. However, I am not seeing signs of the SLFM which I normally see. Another BIOS glitch depending on how it feels when it wakes up? (This is without RMClock running).

John

PS: I've just spotted SLFM (=0.9V on my T7300), so both modes are working.

 

what is IDA exactly, cos I havent seen any signs of it, other than setting my processor running at 1.170V when I have 9.0x set at 0.987V and IDA is 1.275V. I got a much hotter lappy whn gaming

edit: I have IDA disabled right now

edit2: I just ran a stress test with TAT and locking the processor in IDA state, temperatures went to 91, where I shut down the test.

IDA wPrime 32M : 50.414
IDA off wPrime 32M : 50.773 (with temp maby 25C lower)

quite good, only 7 sec slower with my 1.83 than a 2.4GHz

oh and I had background programs

 

IDA (Intel Dynamic Acceleration), you wont see the difference in WPrime because its multithreaded, so IDA wont kick in. You may see it SuperPI, if your second core is completely idle.

But 91 degrees? WOW... something must be wrong, overclocking one core by 200MHz shouldn't be causing insane temps. ?

From Wiki:

 

its running on 1.175V. But I have started to suspect that somethings not completely right with my cooling, as other say their idle temp with default temps are lower than my undervolted one

Think I have to take a look inside it

 

When COMPLETELY idle (@ 1197MHz), my temp is 42-43... You can't get as low as that?

 

I think the Intel TAT gives the CPU a hard time, probably beyond real life conditions. It is a tool for testing thermal performance.

John

 

Intel TAT is always 5-6 degrees hotter than RmClock!

Don't know which program is reading correct temps?

 

more likely TAT, becos thats Intels own temp measurement program.

@n0elia, no, have never got down to those temps after the laptop have been used for a while. My cores are at 600Mhz .8500V, and temp is over 50C all the time...

 

Well, my temp readings are from RmClock, and if you're using Intel TAT, our temps are most likely the same

 

well I couldnt undervolt my lappy without RMClock, and btw my TAT and RMClock temps are the same.

 

How come, you couldn't undervolt?

 

This is with FF2 with 3 tabs open, itunes playing, and MSN

 

well heres mine

wmp playing, browsing + background apps

and yes I can undervolt, but I cant undervolt without RMClock. Get it?

 

heres mine, seems to run a bit cooler!

 

A bit! What processor do you have and what applications did you have running?

 

More importantly, stay indoor during winter!

 

well yeah, I can get my laptop processor run at 15C inside, but thats in a non heated house in -20C

 

its a t7500 (2.2ghz) i was just running winamp and web-browser, it does fluctuate slightly from 39-43 when opening and closing apps etc. and general stuff. Goes up to about 55-60 when full load (gaming and encoding etc).

I was lucky with my e6600 2.4 core duo desktop CPU too That runs about 6 dgress hotter from idle to full load.. overclocked to 3GHZ and still runs at 34 idle ;p

 

and me... only itunes and ie running...
temp is pretty normal for light work...

 

Woot, you run 0.8500V on 8.0x ??? Stable ?

 

Frick!
My settings reset to default again! Actually I'm at 8.0x = 1.175... not
.850V as shown... not sure why it's showing that... i didn't even notice
until you pointed out.

my 8.0x should be 1.025... stable... but it keeps resetting (but not all
the time...)

 

how the fack can you have gaming temps of 55-60???? Mine is 75-80C?????

Im seriously starting to suspect that something isnt quite right with my cooling

 

Not sure is this the right size, but ran SuperPi in Ubuntu Gutsy with the command "./super_pi 20" and got this:
Started super_pi run : Sat Dec 8 15:56:45 EET 2007
Start of PI calculation up to 1048576 decimal digits
End of initialization. Time= 0.292 Sec.
I= 1 L= 0 Time= 0.768 Sec.
I= 2 L= 0 Time= 0.876 Sec.
I= 3 L= 1 Time= 0.876 Sec.
I= 4 L= 2 Time= 0.872 Sec.
I= 5 L= 5 Time= 0.872 Sec.
I= 6 L= 10 Time= 0.876 Sec.
I= 7 L= 21 Time= 0.876 Sec.
I= 8 L= 43 Time= 0.868 Sec.
I= 9 L= 87 Time= 0.876 Sec.
I=10 L= 174 Time= 0.872 Sec.
I=11 L= 349 Time= 0.872 Sec.
I=12 L= 698 Time= 0.928 Sec.
I=13 L= 1396 Time= 0.884 Sec.
I=14 L= 2794 Time= 0.872 Sec.
I=15 L= 5588 Time= 0.868 Sec.
I=16 L= 11176 Time= 0.864 Sec.
I=17 L= 22353 Time= 0.852 Sec.
I=18 L= 44707 Time= 0.820 Sec.
I=19 L= 89415 Time= 0.756 Sec.
End of main loop
End of calculation. Time= 17.309 Sec.
End of data output. Time= 0.108 Sec.
Total calculation(I/O) time= 17.417( 0.608) Sec.
------ Ended super_pi run : Sat Dec 8 15:57:03 EET 2007

That's with T7500 and 2x1Gb Memory.

 

You need to run the 2M test, now your running a shorter test

 

Yup, that was too small.. Ran again with the right command "./super_pi 21"
Results:

Started super_pi run : Sun Dec 9 16:10:58 EET 2007
Start of PI calculation up to 2097152 decimal digits
End of initialization. Time= 0.592 Sec.
I= 1 L= 0 Time= 1.656 Sec.
I= 2 L= 0 Time= 1.884 Sec.
I= 3 L= 1 Time= 1.932 Sec.
I= 4 L= 2 Time= 1.920 Sec.
I= 5 L= 5 Time= 1.944 Sec.
I= 6 L= 10 Time= 1.908 Sec.
I= 7 L= 21 Time= 1.892 Sec.
I= 8 L= 43 Time= 1.912 Sec.
I= 9 L= 87 Time= 1.884 Sec.
I=10 L= 174 Time= 1.900 Sec.
I=11 L= 349 Time= 1.904 Sec.
I=12 L= 698 Time= 1.900 Sec.
I=13 L= 1396 Time= 1.900 Sec.
I=14 L= 2794 Time= 1.888 Sec.
I=15 L= 5588 Time= 1.888 Sec.
I=16 L= 11176 Time= 1.920 Sec.
I=17 L= 22353 Time= 1.876 Sec.
I=18 L= 44707 Time= 1.872 Sec.
I=19 L= 89415 Time= 1.784 Sec.
I=20 L= 178831 Time= 1.664 Sec.
End of main loop
End of calculation. Time= 39.422 Sec.
End of data output. Time= 0.212 Sec.
Total calculation(I/O) time= 39.634( 3.084) Sec.
------ Ended super_pi run : Sun Dec 9 16:11:39 EET 2007

That was without any running processes, then again while surfing and listening to music it was 41,2sec.

 

superpi 2M -> 55.411
wPrime -> unknown, couldn't get program to run with beta or stable version.

 

you need to run wPrime as administrator, or it will give an error

 

Woops, too much 'glogg' this weekend... here's the scores:

superpi 2M -> 55.411
wPrime 32M -> 37.098

(used 2 threads with wPrime, correct?)

 

Correct

 

Hi All, Thanks for all the testing results.

Judging from the limited samples below, and the variance shown plus the fact that Intel's IDA kicking in from time to time depending on different BIOS ACPI definition it is no wonder that the time varied so much.

The low overhead of the linux also shined through in this comparison, as we can see the time's drastically shorter when compared with other times with Microsoft eye-candy sucking up cpu cycles in the background.

Since the fastest times are on-par with the average manufacture, I think it's safe to say Zepto's BIOS do not have major faults of any kind, and as suggested in this thread all sorts of advanced features do kick in, except not as obvious probably due to engineer's taste and perspective.

So, I've put down my order for a 6625W This is a very nice forum indeed, I am lucky to have found this place! (The price actually dropped by couple of quid since I've put in my order, and Zepto's UK representative was ever so helpful and did great customer service!)

Thanks all Now I'm just anxiously waiting for it to arrive...

Code:

superPi
=======
Dell Inspiron e1705 (2.0GHz Core 2 Duo T7300) 1m 02s
Lenovo ThinkPad X61 (2.0GHz Core 2 Duo T7300) 1m 01s
Lenovo 3000 V200    (2.0GHz Core 2 Duo T7300) 0m 59s
Lenovo ThinkPad T61 (2.0GHz Core 2 Duo T7300) 0m 59s
Dell XPS M1330      (2.0GHz Core 2 Duo T7300) 0m 58s

John Ratsey      6024W (2.0GHz Core 2 Duo T7300) 1m 04s
n0elia           6625W (2.2GHz Core 2 Duo T7500) 1m 02s
CLB-NL           6224W (2.2GHz Core 2 Duo T7500) 0m 57.128s
mngm             6625W (2.4GHz Core 2 Duo T7700) 0m 56.902s
ChristopherGrant 6625W (2.4GHz Core 2 Duo T7700) 0m 55.411s
laellis          6625W (2.2GHz Core 2 Duo T7500) 0m 39.634s

wPrime
======
Samsung Q35		(1.8GHz Core 2 Duo T5600) 46.274s
Zepto		6224W	(2.0GHz Core 2 Duo T7300) 45.788s
Samsung X60plus 	(2.0GHz Core 2 Duo T7200) 44.922s
Zepto 		6024W 	(2.0GHz Core 2 Duo T7300) 42.385s
Samsung Q70 		(2.0GHz Core 2 Duo T7300) 42.218s
Systemax Assault  	(2.0GHz Core 2 Duo T7200) 41.982s
HP Compaq 6910p 	(2.0GHz Core 2 Duo T7300) 40.965s

mngm		 6625W  (2.4GHz Core 2 Duo T7700) 0m 46.25s
John Ratsey	 6024W	(2.0GHz Core 2 Duo T7300) 0m 44.499s
n0elia 		 6625W	(2.2GHz Core 2 Duo T7500) 0m 42.13s
CLB-NL		 6224W	(2.2GHz Core 2 Duo T7500) 0m 41.062s
ChristopherGrant 6625W	(2.4GHz Core 2 Duo T7700) 0m 37.098s

 

You didnt add my test results :'(