Windows Experience Index for T61 w/ T7500, 100GB HDD, & Integrated Graphics

I've been getting some real strange numbers for my Windows Experience Index before and after my clean install.

Specs of my notebook:
T61
T7500
4GB Ram
100GB HDD
Integrated Graphics

Can anyone with the same setup and Vista Premium 32 or 64 bit please post their values for the Windows Experience Index?

Initially when i first had 2.5GB's of ram installed on my 32-bit OS, i received a WEI for Memory of 4.8. After i upgraded the memory to 4GB's, my WEI for Memory dropped .1 from 4.8 to 4.7. This is still the case after the clean install of Vista 64-bit.

What's even more puzzling to me is the WEI for Processor decreased from 5.1 to 4.4 after i installed the memory. That's a drop of .7pts. This is still the case with the 64-bit OS.

Keep in mind i've ran memtest for both 2GB modules, 3 passes each, about 1:45 minutes per session per module, and both were error free. WEI are arbitrary values, but there still has to be an explanation as to why both the Processor and Memory index dropped on my notebook.

 

I find it amusing that anyone who is interested in performance would even consider Vista. If you have any desire for performance, run XP...

 

And your comment is related to the topic at hand in what way again?

 

Made me chuckle, very true

 

Still hasnt added anything to the thread.

 

I have the following
T61
T7300
120gb 5400rpm HD
2gb of ram
Integrated graphics

My overall windows experience index is 3.4 and my processor recieved a 4.9. Memory ops per second is at 4.8

Just by eyeballing your specs, there is absolutely no reason why I should be scoring higher than you in these categories.

What type of ram are you using?

 

The WEI score really is pointless. It doesn't affect your ability to run anything in any way, I don't see why you should mind. Unless this is another thing we can chalk up to your rampant OCD (ie. brightness OSD).

 

Have you tried returning your system back the way it was (ie, take out the extra RAM)? If it goes back up, I'd try replacing the original RAM with the new RAM and seeing if it drops the score again (so you know if it's the new RAM or just more RAM that's doing it).

Either way, I'd rely on superpi or some more standard benchmark to compare your CPU against published results from other identical CPUs instead. Ops/sec tests tend to be RAM-agnostic since they keep data chunks small to isolate the CPU performance from other system components.

I guess I'm OCD about this kind of stuff too.

 

Johny has a point - the WEI score is a bit pointless to be honest. I remember reading when Vista was first out that you had to have some kick ass hardware to hit the top score for WEI when something that would score a LOT lower and therefore result in a lower WEI score yet still provide a perfectly enjoyable Vista experience.

 

I'm not after the highest WEI score possible. What i'm trying to figure out is where and why the variance occurred, whether it is hardware or software based.

 

Johnny, i dont know where you feel you have the room to comment on my "OCD'ness" since you have more than a few trivial threads in this forum yourself. The most recent being "How annoying is the 7200rpm harddrive?"

This forum use to be about helping one another with their problems. Now it's become a popularity contest, a home where all the true "OG hardcore notebook thugs" can gather and flame the newer members [Johnny you have dished out your fair share as well].

It's a pretty easy concept guys. If you cant add anything positive to the thread, just dont post at all.

Thanks panteedropper and El Guano for adding substance to the thread!

 

I only label you as having OCD because you labeled yourself as having OCD. It was all in jest, not trying to insult you. When I asked why people were so concerned about the brightness OSD, you admitted it was not really important but you were annoyed because of "OCDness."

See? Hypertrophy is hyper sensitive.

Let me explain why WEI and brightness OSD are trivial. WEI is trivial because there are no substantive implications from it. All it is is some number assigned by Vista. It doesn't affect your computer in any real way. The brightness OSD is trivial because your eyes don't need a meter to detect how bright a screen is. By contrast, a volume OSD is a lot more important because it'll be too late for you if you had to listen to find out your volume was set at 50.

On the other hand, a question about the loudness of 7200 rpm is very substantive. It helps me and others interested in upgrading from 5400 rpm drives to 7200 ones find out if the difference in loudness is worth considering. A problem with excess noise cannot be chalked up to simply "OSD." Coworkers, roommates, and other neighbors may not want to tolerate too much noise. Some of us like to leave our computers on at night when we sleep. Noise is a very real issue. Brightness meters and WEI are not.

Where is this "flaming newer members" stuff all coming from? Did you bring this up because you consider yourself to be a "newer member"? If you do, then you don't know anything about how this forum "used to be." And what is this "popularity contest" you are referring to? I wasn't invited to this contest.

This forum is plenty still about helping others. See the new FAQ lately? What was your latest contribution?

acaurora and I are adding something positive to the thread. The idea is that you shouldn't worry about your WEI score!! It means nothing and as you already witnessed, it is inaccurate. Maybe you didn't like the tone of my post, but acaurora's was quite neutral to say the least, and you didn't mention him in your "thanks for adding substance." Clearly, when people have opinions that you don't see eye to eye with, you find them "lacking of substance." You're a true original gangsta yourself.

 

I was using OCD as a joke; you were not. You referred to "my" OCD as, and I quote, "rampant."

Your over the top thread titles suggest otherwise. Someone is craving a little too much attention. Didnt you just recently publicly state that you should be a moderator?

WEI means something to me. I''m trying to find the cause of the variance between my original score for the Processor, and my new score. You and the others that just brush off my concern does NOT add substance to this thread. If you wanted to help, which i doubt, you really would have just posted up your scores. None of the added BS was necessary. But then again, you were never really trying to help. This is all just a carryover from the argument a couple of days ago.

I was not referring to myself when i stated newer members. I was talking about the "newer members" who had questions and you just gave them attitude instead of answering their questions, and flamed them in their own thread. Dont deny this fact because the other day in your "we need more stickies" thread you stated you have flamed the newer members in the past.

See comment #2.

 

@ hypertrophy -

I am willing to take your original question seriously and give some constructive feedback.

You are obviously concerned that your system performance and maybe even stability has been effected since your RAM upgrade.

I think I have a possible solution to the problem.

I would suggest using one of the standardized tests such as PCMark05 (and maybe even Superpi as suggested by El Guano). PCMark05 is available for free download. So try this... run PCMark05 a couple time on your current system and write down the results. Then swap the RAM for the original RAM. Then run PCMark05 a couple more times. Again, make note of the results. Also, you might as well run WEI one more time with the original RAM while your there.

If your PCMark05 scores are better with the 4GB RAM then you know the WEI numbers are bogus. Therefore you have nothing to be concerned about. If your PCMark05 scores are lower with the 4GB RAM then the RAM is the culprit problem and you should contact the supplier or manufacturer. You have a lifetime warranty on those sticks so it shouldn't be any trouble.

Good Luck!

 

Hypertrophy, you read WAY too much into my posts or my "over the top" thread titles, always trying to rip out some sort of negative connotation that wasn't intended in the first place. It's not even some sort of accidental misunderstanding because I tell you straight up I was kidding about the OCD or I was satirizing racism. Judging people is perfectly fair, but if someone tries to clarify their intentions to you, you ought to respect what they say and accept that maybe your judgment was inaccurate. You are probably the only person out of the 77,105 NBR members that actually thinks I was being serious saying I should be a moderator. Not before, but now, I am going to be unwilling to help you with anything.

That's the last I will speak on the matter. You can get the last word if you want because I can't see your posts anymore.

 

This isn't healthy. This needs to stop, on all sides.

 

Believe me, it's stopped already. No need to pursue the obvious.

 

Back to the topic please fellas.

I have ruled out the memory as being the culprit to the lowered Processor WEI score. I ran many different configurations last night of the ram, 2GB, 2.5GB, and 4GB, switching from DIMM 0 to DIMM 1 with the older and newest 2Gb ram module [wanted to try that before doing benchmark testing]. All the WEI scores are consistent across the board for 2GB and 2.5GB configurations. Memory score for 2GB was 4.6 and 4.7 for 2.5GB configuration.

The two 4GB configurations also show consistency. The Memory WEI score both scored 5.7 when switching from the newest and older 2GB modules from DIMM 0 and DIMM 1.

For all test, the Processor WEI score was 4.4, down from the original 5.1 when i first received my system (keep in mind i have a T7500 2.2Ghz cpu). From what i can tell by others who have reported their Processor WEI score, this is not the norm. How did this variance occur? Is it one of the registry patches? Is it a problem with a driver? That i'm still trying to figure out.

 

I've been planning to upgrade to 4GB, while researching on Intel site I noticed that the GS/PM965 supports "2 SO-DIMMs / up to 4GB Max System Memory @ 533 MHz": http://compare.intel.com/pcc/showchart.aspx?mmID=28116&familyID=7&culture=en-US

I guess when you installed 4GB, the memory speed drops to 533MHz instead of 677MHz, that might be the reason you are seeing the performance index drop on the memory. But it is a bit weird that it impacts your CPU, unless that Vista's CPU index score is not all CPU depended (As it should be), but also impacted by the memory bandwidth.

You can quickly confirm that using cpu-z: http://www.cpuid.com/cpuz.php
I would recommend that you check in cpu-z the following:
CPU: Core Speed, Multiplier and Bus Speed
Memory: Frequency and FSB':'DRAM
Try comparing 2GB vs. 4GB

Please let us know what are your finding.
Thanks

 

Hmmm, that's interesting. Yeah, I use cpu-z as well. Very lightweight free utlity that will give you a "real-time" display of CPU / RAM etc. You could use that or check in the BIOS.

 

Hi Hypertrophy,

Can you please have a zip up the following files from your PC:
%windir%\performance\winsat\winsat.log
%windir%\performance\winsat\datastore\*.xml

The first file contains error information during your assessment. The second set of logs are XML data stores of your previous assessments. I will have a look and see if I can diagnose your issue. I am wondering if you are having some memory configuration issues. Please private message me for logistics.

Thanks,
Quais

 

If you want to measure your raw memory bandwidth in Vista simply executed

WinSAT mem

In an elevated command prompt.

 

At this point, I would look at what an average T7500 system on the CPU WEI. I can't really tell if you're saying it's lower than average, or just lower than before.

If it's lower than average, I agree with the others in that you should try testing against the other benchmarks (Sisoft Sandra and 3DMark CPU both give good synthetic scores that should isolate your CPU). If these apps tell you that your CPU is on-par with the average around the 'net, I think you're fine. It might be possible that the first WEI score that shipped with your system was some kind of fluke...

If OTOH the other benchmarks show your CPU performing lower than average, I'd start looking at software first, and potentially nuking as many "unneeded" processes as possible - thinkvantage, sound, quicktime, adobe, wifi, etc. drivers as well, and see if that affects your score.

My hunch is it's all software and vista related. There's not much room for hardware issues that merely dink your performance rather than smoke your entire system.

 

I just ran a Super Pi test to 2 million digits and i received a time of 01m 38.779s. I believe that is WAY below average for the T7500 2.2Ghz processor.

I tired running a PCmark05 test, but upon completion it stated there was an error so i had no score.

Edit: 3Dmark05 failed as well.

 

According to CPU-Z, under the Memory tab, under timings, my 2GB modules are only running at 332.5MHz. This is for a total of 4GBs of ram installed, 4022MB's listed in CPU-Z.

 

I ran the command and it stated "Memory Performance 3405.90 MB/s."

 

This is interesting, it actually means that that your memory is operating in 667MHz. DDR stands for double-data-rate, which is twice as effective than the actual memory clock speed (332.5MHz x 2 = 665MHz, which is in the range of 667). So in this case your memory doesn't drop to 533MHz, as Intel stated.
You are only seeing 4022MB of ram, because the integrated graphics is using part of the total RAM (4096MB - 4022MB = 74MB).
Could you post screen shots of the CPU and memory tab from cpu-z? If not, what do you see under CPU Tab:
1) Core Speed
2) Multiplier
3) Bus Speed

And under the Memory tab:
1) Channel #:
2) Frequency
3) FSB':'DRAM

Thanks

 

Core Speed -1196.6MHz
Multiplier x 6.0
Bus Speed - 199.5MHz

Channel# - Dual
Frequency - 332.5MHz
FSB : DRAM - 3:5

 

Here's the print shots:

 

^^^Good Point.


I think everything in cpu-z looks OK.

Aside from numbers, how is the real life performance. Does windows boot in a reasonable amount of time. (I think it should boot in less than 2 minutes) Do applications load fairly quickly. Have you done any burning or video encoding/rendering? Does DVD video play smoothly?

Also, you might have a look at Black Vipers Vista Service Configuration to see if you can knock out any more processes. I would stick with the settings under "safe" at least to start out with.

Black Vipers Vista Service Configuration

 

Wait a second - 1200MHz is certainly not right for your CPU speed. Your multiplier should be 11x, not 6x. My T7300 scores a 4.9 on the WEI and can do SuperPi to 2M in roughly 1 minute, even in Vista. Something's wrong here, and it's possible you're overheating and thusly the processor's downclocking to compensate. It's also possible that the processor just isn't properly waking to full power. What is your current power setting? If you're on "Max Battery" or equivalent, that could be the source of your problem. Try setting it to "High Performance" and rerunning all the benchmarks. Also, go to Google and find a program called CoreTemp. Make sure your CPU isn't ever approaching the TJunction temperature (should be 100C) - if it is, send your notebook back for repairs! Few things are worse for a computer than too much heat.

 

My settings were set to High Performance directly after doing the clean install, so all test were done with that setting.

I googled CoreTemp and it appears it's not compatible with Vista. I tried opening the good'o Intel TAT, but it does not open. I dont think it is supports Vista 64-bit. Any other CoreTemp monitors that you can think of?

Edit: Ran Super Pi with Maximum Performance setting and still received a time of 01m 38.436s for accuracy to 2 million.

Ran another WEI score test and the Processor still reads 4.4. The Memory score has dropped back down to 4.7.

It's feels a little laggy. Boot times are pretty reasonable though. Applications take a bit of time to load. I havent tried watching any dvd's yet.

I think i'm going to stay away from the services.msc configurations for now. I want it as "stock" as possible so i dont change too much variables in the equation.

 

Hey chill, just expressing an opinion.

 

@ lowlymarine - Could it just be power management? Have you looked at your numbers on cpu-z? They are real-time... meaning that if power management is doing it's job, it is throttling down the cpu to conserve engery (thus giving a false low reading). You should be able to run super pi with cpu-z open and watch those numbers climb.

However, if you are getting much better scores with super pi than hypertrophy then their must be a problem.

 

Seriously, that was three days ago. Everyone has chilled out since then

 

I just ran another Super Pi test with CPU-Z running. The Core Speed was holding steadily around 1197.1MHz-1197.2MHz.

 

I agree with stallen everything seems right with CPU-Z, as far as I can tell.

Santa Rosa or even older Centrinos can throttle down their clock multipliers based on CPU load. This is not a false reading.

But I sure yet that this is a software issue, from what I understand the WEI score changed by only changing the memory module on the same system. The same services should run under different memory configuration! Yet, shutting down services might improve his performance results.

hypertrophy, I had to read your previous posts again and now I'm a bit confused, in your first post:

I understand that you are comparing Vista 32 bit with 2.5GB Vs. Vista 64 4GB? And:
1) Upgrading to 4GB drops your memory index to 4.6:
2) Your CPU score drops to 4.4 - Have you seen the same score drop in 32 and 64 bit?

And here:

Here you were testing on Vista 64 only?
1) But this time in the 4GB configuration your score is 5.7, so it is actually higher??
2) CPU score stays 4.4
What seems to be consistent is that your CPU index score is lower under 64-bit compared to 32-bit - This might be indeed a software issue.

On the other hand, I realized that your memory bandwidth test results seems to be very low. Unfortunately I haven't received my T61p laptop, yet, so I don't have a good reference, but WinSAT on my desktop is showing 4015.78 MB/s and this is for DDR 400 dual channel. In order to validate WinSAT I tested my memory bandwidth with Sisoft sandra which returned 3998MB/s - Both are in the same range.
If you take a look at the Samsung Q70 (also Santa Rosa) review on Notebookreview.com ( http://www.notebookreview.com/default.asp?newsID=3806) you'll see that memory bandwidth on this system scored ~4300MB/s!

Sorry for being a Numskull, but can you please clarify a bit more the tests procedures you preformed?

Thanks

 

I believe it, referring to the Processor WEI score drop, occurred right around the same time as when i upgraded my memory while i was still running vista 32. When i initially received my notebook, Vista 32, the Processor WEI score was 5.1. Before my clean install of Vista 64bit, i had already been seeing the lowered Processor WEI score of 4.4. The test i've ran testing the different memory density configurations with different dimm locations were all ran under Vista 64 bit after the initial clean install. 2GB configuration showed Memory score of 4.6, 2.5GB showed score of 4.6, and 4GB configuration showed a score of 5.7.

The WEI score test ran today is using Vista 64bit, but i did yet another clean install from when i received the 5.7 Memory score. The Memory score now went back down to 4.7 with 4GB's of memory and i am unable to repeat the 5.7 score with this clean install. The processor score with the latest clean install still is at 4.4.

I've also tested the theory that speed step might be the problem. With CPU-Z opened, i switched back and forth from maximum battery life to maximum performance. I did this multiple times and the core speed never read any higher then 1197.

I ran the Super Pi test earlier with CPU-Z running, and the core speed never reached anything higher then ~1197. If speed step was enabled, when the system is stressed with a benchmark test, shouldnt i see core speed numbers great then 1197MHz?

I just got off the phone with tech support and, off the record of course, he believes i should look into the Intel Speed Step software.

Edit: I would also like to make a note that i have the newest BIOS v1.14 installed on this computer. I installed it after the latest clean install. The previous BIOS version was v1.07. Processor WEI score test of 4.4 was the same for both BIOS versions.

 

Well, there is something else to take into consideration. Does super pi test for dual core. I really don't know. You might not be seeing a bump in the number in cpu-z because you are looking at one core while super pi is stressing the other. Just a thought.

For some reason with these new santa rosa processors cpu-z doesn't allow to toggle between processors in the drop down menu (it is greyed out). On my desktop PC with dual core I can select under the CPU tab "processor selection" and see information on both cores.

Perhaps you could set the affinity in the task manager. Open super pi -> Open task manager -> Find super pi under the processes tab -> right click -> select only one core -> then run cpu-z and super pi simultaneously. Try this for both cores. Let's see if your processor throttles up for either core.

If you don't show that one of those cores is throttling up. I'm at a loss. Perhaps a power management driver issue???

 

It reports "Access Denied" when i click on Set Affinity.

I've been looking through the recent notebook reviews w/ the santa rosa platform, and their Super Pi times are much lower then mine even with slower speed chips (mostly T7300). Even if Super Pi doesnt test for dual cores, shouldnt my times be comparable, or even lower then their's since i have the T7500? Enabling or disabling the Speedstep in the BIOS has no affect on the CPU-Z core speed readings i'm receiving.

 

Stallen has a point, Super PI is single threaded and I believe that in Core 2 Duo each core has its own throttling. But from my experience, when one CPU is 100% utilized it should have some impact on the other core. A non changing 1197 seems a bit strange. As an alternate test, try:
1) Open task manager and leave it on the performance tab
2) Open CPU-Z
3) Then run Super PI twice simultaneously.
At this point your computer will be very sluggish. Make sure that both cores are 100% utilized, and take a note on the core speed shown in CPU-Z. You can stop at this point the Super PI tests.
If CPU is still 1197, then it seems that tech support is leading you to the right direction. If we assume that CPU-Z is not lying and the higher index score in Vista 32-bit indicates that CPU multipliers are throttling up properly under high load, then I would guess that something is wrong with the power management / Intel Speedstep on your Vista 64 bit.

Have you installed all available updates for windows Vista and the Lenovo power management drives and utilities?
Unfortunately, I am currently laptopless, so I don't remember all of them by hart.

 

Is the power management driver/utility even necessary? That energy star 4.0 option is brand new. Doesn't Vista have something built-in for laptops? Can those be uninstalled? If so, maybe give that a try. At least temporarily until you find out if your CPU will throttle up.

 

I kept getting an error message reading "Not Convergent in SQRO5 when i tried to start a second Super Pi test. It wont let me run two test at once.

I only have the Power Manager installed, not the ACPI Power Management driver that's Energy Star 4.0 rated. I believe vista does have it's own power management because i could have chosen from Balanced, High Performance, and Power Saver mode even before installing the Power Manger driver.

Edit 2: When windows started up for the first time after the clean instlal, it automatically did a WEI score test. My Processor was still rated at 4.4. I dont think it's related to the Power Manger driver from the T61 driver matrix.

 

Well, maybe not but you don't know for sure until you try. There are uninstall instructions for the power manager driver on the download page. It's worth a try. Worse case scenario is that it doesn't make a difference and you can just re-install it. I don't know what else it could be unless there is something wrong with the processor itself.

Here is an x61 review I just found that talks about the Vista power management conflicting with Lenovo power management (albeit not in the same way you describe). There are also some super pi scores to look at. http://www.notebookreview.com/default.asp?newsID=3765

 

I uninstalled the Power Manger driver, rebooted, then ran a Super Pi test with CPU-Z open and it still hovers around 1197MHz. It didnt work.

 

Hi, I've got a T7500 C2D, and the WEI gives me a 5.1 (4.8 on 2gb ram). I'd like to add its not a T61 but an HP 6910p. SuperPi to 2M places is around 56 seconds.
Something is seriously wrong with your CPU or you need to install chipset drivers? There is no way that you should be stuck at 1.2Ghz
My CPU downclocks to 800Mhz (8 x 100) using the latest CPU-z, and also downclocks the ram to DDR2-333 (166Mhz).
So a) make sure chipset drivers are correct and b) if it continues you probably got a dud.

 

You might want to try also RightMark CPU Clock Utility (I personally haven't used it much): http://cpu.rightmark.org/download.shtml. With this tool you can check both core clock speeds, even control speedstep and other settings. But if you if you are going to experiment with the settings, then make sure that you read the manual: it comes with a WARNING! and it might not even work under Vista.

Or even wPrime, check this thread: http://forum.notebookreview.com/showthread.php?t=123570

 

I asked my friend to help me out with the problem and we were able to
find the problem, here's what he wrote up:

Here are a few tests that I ran today, to rule out possible software/driver issue, I loaded ubuntu linux using a live-cd and then was able to run two processes at the same time on both cores and watch the cpu clock frequency in /proc/cpuinfo go all the way correctly to 2200Mhz for both cores. The bogomips also reported by linux correctly matches the other dmesg output I have found in google. This confirms the hardware works, it is a Windows issue.

Back in windows, I looked at the windows task manager and it is confirming that the cpu frequency is topped at 54%. If I use the "presentations" power mode, then it is pegged correctly at 36%. The frequency can scale down but it is topped at 54%.

Then I thought that it could be a bug with the power management driver, so I noticed that the battery was not inserted in the computer because it was removed since the laptop is mostly run on AC outlet (to preserve the battery). Immediately when re-inserting the battery, the cpu clock went to 100% and then everything is running faster now. For example, Super Pi test runs at 54s instead of 1m38, AND the WEI score is now back to 5.1. So it all makes sense now and the WEI score does indeed reflect the cpu speed calculated based on the frequency it ran at.

So the conclusion is that there is a logic error in the power management code that doesn't handle properly the use case where the battery is removed and the laptop is running on AC. It is not clear if it is a windows and/or lenovo driver issue or both, but at least there is a workaround.

Edit: Just wanted to add a big thank you to everyone who stuck it out with me during this thread and who tried to help.

 

Wow that's interesting!?! So if I understand correctly. Everything is fine and works at top performance as long as both the battery is installed AND it is plugged into AC. Is that right?

I guess I probably would have never even run mine without a battery. Even if you use AC all the time it is nice back-up power.

Anyway, I'm glad i you got it figured out!

 

Congratulations! that's great news.