ThinkPad Benchmark Thread

I don't know about the rest of you, but I often find myself quite curious about the performance of my system compared to similar set up ThinkPads.

In other words I would like to check whether or not my machine is on par, slower or faster. The information can serve as basis for diagnostics or it can serve as useful information for people considering buying your exact configuration.

I know that there are some excellent reviews out there of the T61p, X61, X61s, etc showing benchmark results, but they don't cover every configuration, i.e. 5400RPM HDDs vs. 7200RPM HDDs.

Hence, I suggest you post your benchmark results using the same set of programs (or as many as possible) as NBR recommends, i.e. HD tune, PCmark05, SuperPi, 3Dmark05/06. As many as you can. Windows Experience Index would be nice as well.

Please post an overview of your system specs and if convenient please also post screenshots.

I'll start.

-ThinkPad X61
- 2.0Ghz T7300 Core2Duo
- 3GB PC5300 RAM (2GB Kingston module + pre-installed 1GB Hynix module)
- 120GB 5400RPM HDD (Hitachi 5K160)
- GMA X3100
- 12.1" 1024x768
- NO Intel Turbo Memory
- NO ReadyBoost device


PCmark05: 3938 (No screenshot. Uninstalled after running benchmark)

HD tune: See attached

SuperPi: 0 min, 58 sec. See attached

WEI: 3.4 overall. See attached

 

Nice idea.

I would suggest these who have the T61p to run SPECviewperf 10, since I'm curious how the Quadro FX 570M performs in 3D modeling applications, and that's what this card is designed for at the and.

http://www.spec.org/gwpg/gpc.static/vp10info.html

It's sad that you can't find proper reviews for workstation models these days. (to be taken as a criticism).

It's about 600MB and the the tests usually take about an hour.
http://www.spec.org/gwpg/downloadindex.html

 

Thinkpad T61
- 2.0Ghz T7300 Core2Duo
- 2GB PC5300 RAM (1GB Crucial + pre-installed 1GB)
- 80GB 5400RPM HDD
- nVida 140M
- 14.1 SXGA+
- Factory install Vista Home Premium


Windows experience posted below.
3D Mark2006 1385
PC Mark2005 4678
HD Tune - Shown below.

 

Good suggestion Dreamer.

Do you own a ThinkPad as well? If you do I'd like to see your results.

 

Wow, your WEI harddisk score is really impressive, considering it is 5400RPM. Which make and model is it?

 

I do but it isn't new (Z60m.. somewhere in my house), so there is no point.

 

Seagate Momentus 80GB - 5400rpm - 150MBps according to PC Wizard 2007. It was stock from Lenovo.

PCMark05 score updated btw: PC Mark2005 = 4678

 

T60
T7200 - Core 2 Duo 2ghz
1gb 667mhz ram
120gb 5200rpm hdd
15" ips
ATI X1400 128mb

3DMark 05 - 2874

Haven't had a chance to run any other benchmarks.

 

Great idea subscribing to thread so i can do it when I get home

 

T61p
3GB (Stock Lenovo RAM, 1GB + 2GB chips)
Vista Ultimate
Hitachi 200GB SATA
1GB Turbo Memory
T7700 (2.4 GHz)

Vista Windows Experience Index scores:

Processor 5.3
Memory 4.8
Graphics 5.9
Gaming Graphics 5.4
Primary HDD 5.4

 

My goodness, on the stock T61 with the Intel IGP, it scores twenty percent higher for gaming than for Aero? Umm, wow. I find that extremely unlikely to reflect reality. . .

. . .ah, hang on it's the NV 140m. Well, okay that makes the gaming score make more sense, but the Aero score of 4.0 still seems awfully odd to me.

Edit: 4.0 apparently isn't considered bad at all. For the Aero test its actually testing video bandwidth.

http://windowsvistablog.com/blogs/windowsvista/pages/458117.aspx In depth discussion of the tests and results.

 

Hey laptopguru, what are you doing with that Lexar 16GB SSD I see in your sig?

I'm thinking about dedicating my turbo memory 100% to readydrive and getting a 4GB SDHC card for readyboost.

 

Well the thread started out right and then it lost momentum. Anyone else? Would especially like to hear from X61 owners.

 

-ThinkPad X61s
- 1.6Ghz L7500 Low-Voltage Core2Duo
- 2GB PC5300 RAM (2 1GB Corsair modules)
- 120GB 5400RPM HDD (Hitachi 5K160)
- GMA X3100
- 12.1" 1024x768
- Vista Business x64
- NO Intel Turbo Memory
- NO ReadyBoost device

WEI: 3.5

SuperPI (2M): 1 min, 11 sec.

 

Thanks Sp00n.

Was wondering about your X3100 driver - which version do you use? Version 7.14.10.1244 supplied by Lenovo or the latest generic driver from Intel?

 

I am using the version supplied by Lenovo. I've tried using the generic driver from Intel but it always prompted me a permissions dialog every time I booted up the computer.

 

I love this. This is exactly what I wanted to measure. Albeit not a scientific comparison, it is interesting to see the difference between i.e. 2 HDDs of the same model/specs but with differences in results.

For example Sp00ns HD tune result showed CPU usage of 13.8% vs. my 1.9%. Also burst rates are quite different. I know we have i.e. different processors, but I didn't expect the HD tune difference to be so big. Could it be something else causing it?

Does anyone know what triggers the CPU usage and burst rates in HD tune? In other words which components have the biggest effect on the aforementioned two things?

 

How can you do?

 

I'm using it to share data between my two work laptops (T60p and Dell M65)and my personal laptop (T61p)at the moment, but plan to eventually install an OS on it and make it my portable environment. I don't think it will ever be the primary boot device in this box, but you never know...

 

I'll download the SPEC benchmark and run it my T61.

I don't know who is the admin of notebookreview, but it seems it would be beneficial to add a section to the forums where users can submit raw scores for benchmarks and have them stored in the database, then you wouldn't have 3 million threads with people posting stuff. Would also be searchable, sortable, and could be used as a reference to their reviews.

 

I agree Otter. That was what I initially thought as well. But I was too lazy to contact the admin or the moderators. Also, I was pessimistic about them making this into the aforementioned.

Anyone else think it would be a good idea to make a sticky or make a benchmark section?

 

Imho, this thread should be a sticky, the other suggestion is worth thinking about as well.

I can forward this to an admin if you wish....

 

Thanks Dreamer. I'd appreciate it if you did. I think you'd have more leverage considering you having been on NBR for longer than myself.

 

This is really a great idea and something I'll commit to working on so people can better share information, but I can't promise any timelines because I'm pretty busy right now. There are definitely some challenges to doing this and keeping the database "clean" and easy to use. Another thing we'll be working on soon is a way for users to easily post pictures of their notebook setups that are associated to their profiles. There's a number of threads in the forums of people begging for pictures and posting pictures of their systems.

 

Great news Andrew! I'll be looking forward to the future benchmark section.

 

I don't know the costs, but there are companies ( maybe VBuletin itself ) that will create custom plugins for you.

What I would do is add a table along the lines of:

users_laptops:
laptop_id = int
foreign key <vbulletin_user_id> = int
laptop_attribute_id = int
value = varchar(256)

users_bechmarks:
foreign key laptop_id = int
foreign key <benchmark_id> int
value = varchar(256)

laptop_attributes:
laptop_attribute_id = int
attribute = varchar(256)

laptop_benchmarks
benchmark_id = int
value = varchar(256)

Then you would have a page that joins all of that together
Then add a page to the user control panel to 'register my computers', then users could add computer configurations to their member profile.

Users have laptops, and laptops have attributes.
Laptops have benchmarks, and benchmarks have values.

The user interface would be the hardest part, I hate UIs!


I am just brainstorming as I type, but the backend shouldn't be too bad. trying to cover all the bases where users can mess it up....... years!
As long as it stays linked to the user id, and you provide a method in the user CP to delete, edit, and add laptops and then scores it it could then stay transparent so that messageboard admins also have that ability.

 

Great seeing some help there Otter.

In the meantime, where are all the benchmarks? I cannot possibly be the only one with the interest in seeing benchmark results. Come out of the bushes people and spread your message

 

I have a question about SPEC:

What is the preferred method? I see single thread, 2, 4 thread runs. Which resolution? it lists 1280x1024 and 1600x1200.

Is there a way to get it to aggregate the results? I ran it on 1,2,4 threads but it seems to erase previous runs even though it lists a column for them in the results page.

 

These features are new, check out the readme.
"C:\Program Files\SPECopc\SPECViewperf\SPECViewperf10.htm"

Otherwise, you can run what you want, but the point is that if want to compare the results with others, it should be apples-to-apples comparison.

Here is the new table with the new methods: (still no notebooks), and the old version.

Btw, Vista has an effect on performance, so don't make a direct comparison...

 

Thanks I have never used this one before so I didn't know which was the baseline ><.

I can't wait to get Linux on my T61, I want to run HPL. My P4 gives about 3Gflops. I am very curious what the new 'Cores' give. Considering it is 64 bit, and 2 dual cores, I am hoping to hit 10Gflops, but that maybe a stretch.

I'll let it run the SPECs again tonight and post the results Sunday. Assuming I can stop playing Silverfall long enough to do that!

Edit:
I am on XP so my results should correlate with theirs.

 

Below are the results of the Viewperf 10 runeverything.bat. This ran them at 1280x1024.
Please not I do not use my Laptops monitor, I use a 22" widescreen set as the primary, and the laptop monitor off. The monitors native resolution is 1680x1050.

The tests were run over night with no other applications running. Typical processes were running ( e.g., firewall, battery monitoring ... core services you would have running on a regular basis )

Of the 2GB RAM I have the tests never utilized more than 1GB ... this could be notable for frugal workstation buyers. (1053148 max used to be specific ). The system cache on the other hand is using 1512260 ~ 1.5GB This is not RAM that is in use... e.g., the system is storing stuff in the memory since it isn't being used, but if it a program needs memory it will take from the system cache. This is important because the system cache speeds things up, when you click a menu, if that data that is in the menu is in the cache it does not have to look it up from the disk. You can view your statistics for this in Task Manager under Performance->Physical Memory, the Available is how much memory is available to applications, System Cache is how much of that available the system is storing stuff in since it is free.

It is interesting to see the results, on a 2 core system you see a huge performance gain from threading... In the 3DStudio and Catia tests you see double the performance at 2 threads, versus the proe, sw, tcvis tests were you see practically no benefit. Interestedly when going to 4 threads on a 2 core system you see small gains on Catia, Maya, Proe, and Sw, while 3Dstudio and Tcvis decline. This was my expected results, as very efficient threading will max out when the number of threads reaches that of available CPUs.

I maybe interpreting some results completely wrong, I have 0 experience with any of the programs it tests. I have done programming on distributed systems using MPI and clusters so I do have some experience with massively parallel jobs ( programs using 100s or thousands of CPUs at once ).

And before anyone comments....
I did not add the enormous amount of whitespace between the tables. It seems to be a result of using HTML in Vbulletin as there is no whitespace in the HTML I submitted.

If you would like to view the results in the native format HTML pages please download the attached ZIP, it contains the 'results' folder zipped in its entirety. This also will show you more detailed results as the below tables contains links to specific test scores and putting all of that into a post would have driven me mad.

rename the file from .zip.pdf to just .zip since zip files aren't allowed over 100k and its 300k with all images stripped.

I'll probably run the 1600x1200 test over night, but will only post the zipped results its alot of trouble to hack out the HTML.

Viewperf 10.0

T61 Nvidia Quadro NVS 140M
Viewset	Composite	Mulitsample Performance
3dsmax-04	3.36	up to 8x
catia-02	3.92	up to 8x
ensight-03	3.43	up to 0x
maya-02	12.19	up to 0x
proe-04	5.12	up to 8x
sw-01	5.10	up to 16x
tcvis-01	1.03	up to 2x
ugnx-01	1.50	up to 2x

Threads Enabled (viewperfMTsummary.html)

T61 Nvidia Quadro NVS 140M
Viewset	Composite 1 thread	Composite 2 thread	Composite 4 thread
3dsmax-04	3.46	8.58	7.57
catia-02	4.04	7.45	8.40
maya-02	11.68	15.81	17.72
proe-04	5.30	6.66	7.20
sw-01	5.55	5.07	5.66
tcvis-01	1.11	1.85	1.77

Full Scene Antialiasing Enabled

T61 Nvidia Quadro NVS 140M
Viewset	Composite no AA	Composite 2X	Composite 4X	Composite 8X	Composite 16X
3dsmax-04	3.36	3.37	3.30	3.12	2.39
catia-02	3.92	3.96	3.97	3.66	2.72
ensight-03	3.43	3.07	2.81	2.31	1.27
maya-02	12.19	10.34	10.25	10.11	5.90
proe-04	5.12	4.90	4.96	4.72	3.42
sw-01	5.10	5.08	5.12	5.60	4.84
tcvis-01	1.03	0.93	0.88	0.65	0.59
ugnx-01	1.50	1.39	1.27	1.13	1.01

Graphics Hardware Configuration
Graphics Accelerator	Nvidia Quadro NVS 140M
Total Graphics Memory	128MB
Display Manufacturer/Model
Display Resolution	1280x1024
Display Size/Technology	22" LCD
Display Refresh Rate	60
Swap on Vertical Retrace
System Hardware Configuration
Processor Type	Intel T7300
Number of Populated Processor Sockets	1
Cores per Processor Socket	2
Primary Cache per Core (KB)
Secondary Cache per Socket (KB)	4096
Tertiary Cache per Socket (KB)
System Memory (MB)	2048
Memory Type	DDR2
Memory Speed	667
Memory Configuration	2x1GB 667MHz DDR2 SODIMM ECC
Disk (GB)	120
Disk Interface	SATA
Disk RPM	5400
Software Configuration
Operating System	Windows XP Pro SP2 w/Dual Core patch
O/S Type	Windows
Compiler Name
Compiler Version
Window System
OpenGL Version
OpenGL Renderer
OpenGL Vendor
Driver Version
Viewperf Version	10.0
Viewperf Executable	Standard
Price, availability, etc
Price	$1600 USD
System Class
Test Date	08/25/2007
General Availability
Submitted by	Lenovo
Comments

 

Well, I never realize that you don't have a workstation card... your card doesn't have hardware OpenGL acceleration, that's way the result are so low, it's done in software, so basically you are mainly stressing your CPU with this benchmark...

Anyway, thank you for your effords, very throughout, thanks.

As for the whitespaces, you should edit out all whitespaces/new rows etc. between the tags but it doesn't matter... I know that because I posted tables before.... so don't bother it's annoying.

 

Thank God! I have the same config, was ready to poke my eyes out last night thinking something was wrong with the painful 3d rendering at 1-3fps! Considering how bad the 140M does on these tests, we might have been better served with the x3100 card rather then beliveving we got something that might actually handle the load.

 

At little of topic of what you guys are talking about, but still related to the title of the thread...

I am interested in overall system performance. Has anyone used PCMark05? I haven't yet but I plan on giving it a run. If anyone is interested I'll post it here (it'll probably be a few days before I get to it though).

Also, I'm not familiar with this benchmark, but I might give it a run as well... http://www.passmark.com/download/pt_download.htm It has a free 30-day trial. That's long enough for me to run it a few times!

 

Check my post in the beginning of this thread, I ran PCMark05. SCORE: 4678

 

Here is the PCMark05 score:
4362 - see attachment for the details.

I probably could have done better on the PCMark, but that benchmark is rather stupid IMO. You can change a few things and get completely different scores. Not to mention the way it tests is rather flawed. It uses the Media 9 encoder which is known to suck. They should use make a benchmark that involves ripping a DVD, encoding it using Xvid and transcoding the audio along the way. Would be much more meaningful.

 

What benchmark do you suggest for overall system performance; something that test a little bit of everything?

 

I downloaded this to see whether or not my score was comparable to yours...same config. But after the testing was done, a screen pops uo and says ckick here for your online score in orange...when I click on it, I get a an error....and of course no score ??? Did you have any trouble? I assumed it was because it was a trial? Was not able to change any options at all..I downloaded from their site. any suggestions? What can you do with a trial download? change options? get scores...etc

 

My PCmark05 score is 3938 as you can also see in my first post. Not bad for an ultraportable.

 

It may be a stupid question but did you have internet access when you clicked for the online score?

The first time I ran PCmark05 I got the same error and it seems it was because I wasn't connected to the internet.

 

I assume so, I ran the program right after downloading it. Maybe during the test the network connections is temp disabled? I will retry. thanks

EDIT:

OK worked this time: Braddd looks like you have your system tweeked pretty good, 1 less gb ram & 5400 but better score
4273 PCMark2005
1464 3D Mark2006
Funny:

Below is a rundown of the main components you have in your current system.
Components Your System
Processor Intel Core 2 1995 MHz
Physical / Logical CPUs 1 / 2
MultiCore 2 Processor Cores
HyperThreading N/A
Graphics Card NVIDIA Quadro NVS
Graphics Driver NVIDIA Quadro NVS 140M
Co-operative adapters No
DirectX Version 9.0c
System Memory 3072 MB
Disk Space 95.4 GB
Motherboard Manufacturer Lenovo
Motherboard Model 6459CTO
Operating System Microsoft Windows XP
Is Your System Ready for Windows Vista™
Unfortunately, your system does not meet the requirements for Windows Vista™. See the table below for details.
Components Your System Minimum Required
CPU Intel Core 2 2.0 GHz 32-bit (x86) processor 800 MHz
System Memory 3040 MB 512 MB
Graphics Chipset NVIDIA Quadro NVS With 512 MB Memory DX9.x compatible With 64 MB Memory
Hard Disk Drive 95.39 GB Capacity 15.36 GB Capacity

 

I think you want to click 'no' or something like that. The online score is optional. Once you decline it, it takes you back to the main window where you scores will now be shown on the right side.

Can you post the complete scores?

I would suspect things like 5400 vs 7200 RPM drives would make a notable difference.

 

In PCmark05 they sure do

Windows Experience Index is a different story though. It seems you get better harddisk scores if you have a bigger capacity drive. That's why a 4200 RPM 200GB drive scored higher than my 5400RPM 120GB drive. One really needs to take WEI with a grain of salt.

 

This is why a benchmark that attempts to make an entire system boil down to a number will always be flawed.

What do you value: Speed or Space? Where is the dividing line between enough space and enough speed?

PCMark values speed, Windows values space.

Practically speaking people value both, but in different concentrations. I value space more than speed, so I have a 120GB 5400RPM drive. Sure it takes longer to load stuff, but I don't care because I go through GBs like they are Cheeto's on my chest. And because I use programs that generally load once, and then you use them for hours without incuring the load time. So the speed of the disk is only a notable difference once. If I frequently compiled massive projects, where the disk speed is the limiting factor, I would be more happy with the faster drive.

 

Sorry if this has already been asked.

I was just going to run PCMark05 on Vista. It prompted me to install DirectX 9.0c. Vista comes with DirectX 10. I don't really want to downgrade. I opted not to install it, so PCMark will not run without it.

It seems like I have read of a few people having trouble when they installed Directx 9c on Vista. What did you guys do? Install it then re-install DirectX 10?

 

I don't get it, Braddd has 5400 vs 7200, 1 gb less ram, his Vista vs my XP pro and everything else the same ..yet still beats my score????? His post on beginning page Could I have something wrong with my system?

Below is the result details of your submitted project.
Main Test Results
System Test Suite 4297 PCMarks
CPU Test Suite N/A
Memory Test Suite N/A
Graphics Test Suite N/A
HDD Test Suite N/A

Detailed Test Results

System Test Suite
HDD - XP Startup 7.35 MB/s
Physics and 3D 178.61 FPS
Transparent Windows 497.72 Windows/s
3D - Pixel Shader 51.13 FPS
Web Page Rendering 2.68 Pages/s
File Decryption 60.03 MB/s
Graphics Memory - 64 Lines 702.05 FPS
HDD - General Usage 4.43 MB/s
Multithreaded Test 1 / Audio Compression 2029.25 KB/s
Multithreaded Test 1 / Video Encoding 333.87 KB/s
Multithreaded Test 2 / Text Edit 126.57 Pages/s
Multithreaded Test 2 / Image Decompression 26.81 MPixels/s
Multithreaded Test 3 / File Compression 4.82 MB/s
Multithreaded Test 3 / File Encryption 26.63 MB/s
Multithreaded Test 3 / HDD - Virus Scan 31.08 MB/s
Multithreaded Test 3 / Memory Latency - Random 16 MB 8.21 MAccesses/s

Below is a rundown of the main components you have in your current system.
Components Your System
Processor Intel Core 2 1995 MHz
Physical / Logical CPUs 2-Jan
MultiCore 2 Processor Cores
HyperThreading N/A
Graphics Card NVIDIA Quadro NVS
Graphics Driver NVIDIA Quadro NVS 140M
Co-operative adapters No
DirectX Version 9.0c
System Memory 3072 MB
Disk Space 95.4 GB
Motherboard Manufacturer Lenovo
Motherboard Model 6459CTO
Operating System Microsoft Windows XP
Is Your System Ready for Windows Vista™
Unfortunately, your system does not meet the requirements for Windows Vista™. See the table below for details.


Components Your System Minimum Required
CPU Intel Core 2 2.0 GHz 32-bit (x86) processor 800 MHz
System Memory 3040 MB 512 MB
Graphics Chipset NVIDIA Quadro NVS With 512 MB Memory DX9.x compatible With 64 MB Memory
Hard Disk Drive 95.39 GB Capacity 15.36 GB Capacity

 

Stallen, you can install the specific DirectX9 file that PCmark05 is asking about, without any issues. The file is a component which is needed by PCmark05 - it will not replace any DirectX10 files in Vista.

In the next PCmark version it will be compatible with DirectX10.

 

Any benchmarks with FX 570M yet??

 

Which ones are you looking for?

I was able to get 3700+ 3dmark06 on stock clocks in Vista.

 

Has anyone been able to install the hacked GeForce drivers for the FX570? Do those help the gaming performance any?
Tim