• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Surprising "benchmark" result. Can you explain?

Soldato
Joined
24 Feb 2003
Posts
3,142
Location
Porthcawl and Southampton
Over many years I've run a kind of personal benchmark on every computer I've built.

It's a very complex spreadsheet of aircraft noise modelling done by the friend of a friend back in 1997.

For reference the first record was a Pentium 2 350MHz that took 8mins4secs to complete the calculations. (office 97, Windows NT)

Other points are a Duron 1.1GHz =70 secs ( Office 2000,XP Pro)


I haven't built a computer for years but just finished a budget build just for fun from parts from MM.

The previous record was my overclocked Athlon XP-M at 35 seconds ( Office 2k, XP Pro)

So I ran it on my new build, an E5200 at 3.0GHz which is the highest stable clock at standard voltages.

I know that the mitigating factors are a much heavier OS ( 7 H.P.) and Office (2010 Beta) and that a simple maths test will not showcase the improvements of a multi core processor BUT

I was gobsmacked when the test came in at 37seconds


Can anybody explain WTF!?!?!?!?!
 
if the file doesn't contain anything sensitive why don't you let a few other people have a pop at it and see how the results tally
 
I'd be interested in seeing the algorithm used, is it multicore capable?

Also Windows and all apps made by Microsoft gets bloated with every incarnation.
 
Written in '97. I shouldn't really publish is because it doesn't belong to me. It was given in confidence.

It's just always provided a real world measure of performance increase.

I'll install an older version of Office and maybe Win XP to try out.
 
Well, an interesting perspective is the combination of CPU power fighting with the bloat of more modern equivalent software and operating systems. As I said I'll try with XP and Office 97/2000 but my original post still stands.

As I said above I'm still really surprised that in spite of the OS and new Office that the E5200 was so slow.
 
It would be nice it MS did a "lite" version of W7 with all the bloat removed. Could be a nice performance boost without any upgrade to hardware :)
 
I think its more to do with the code than the processor. If the code is not taking advanatge of the extra features of the new processor then there will be no difference me thinks.
 
You should test it under same OS/Office version.

Otherwise its really not comparable.

+1

Seems a little unfair to run such differing software versions. Microsoft are well known for soaking up additional processing power of the new generation of CPU with more bloatware in thier new version of office.

Also as others have mentioned, the code you are using certainly won't support mutithreading etc.
 
A little update.

Windows 7 but with Office 2003 dropped the time to 36 seconds. Still very disappointing.

However, I did an install of XP Pro (nlited) on a new partition to dual boot and put Office 2000 on that one.

Result........26 seconds. Now that's more like it!!!!!!
 
Haha, i bet you could go lower if you ever reached the original OS/Office.

But still, just goes to show how much bloat gets added through the various releases.
 
Indeed.

Another update......Win 7 with Office 2000=29seconds.

So the biggest factor in the slowdown is office.


On the flipside, I was pleased to see Super-pi being much quicker in Windows 7(64) than XP Pro(32).
 
Small story:

Person had a Celeron 466 + 128MB + 6GBHD + Win98SE
Ran it for years and was perfectly happy on its speed and it did her just fine.
Only a couple of years ago, she bought herself a new PC

Celeron 3.06 + 512MB + 60GBHD + XP Home

She was so upset at the awful speed of her new PC that she put it away and brought her old PC back from under the stairs.

3 weeks ago, she was still using her 466 + Win98

---

Moral of the story would be I suppose that newer and supposedly faster CPUs dont really make newer PCs any faster because of the even more bloated Apps that come with them.

Try installing Win2000 ( Seems a good compromise I recon as it runs on both old and new hardware ) and the same version of all the software on each PC
 
Back
Top Bottom