My 3dmark dropped 1600 points in 1 month with no hardware changes

[Hammyton]

My 3dmark in january was 9998, but now its 8393!!
I have the exact same system components!
I uninstalled and changed to omega drivers and now i changed back and all of a sudden my 3dmark sucks! I did a clean uninstall of the previous drivers, and I am sure AA and Aniso are off....
Any ideas?
9998
8393
(note i just realized that the 8393 was 10mhz lower memory clock for the g4 than the 9998, so i benchmarked again with the changed, but my score only went up 100 points)....

 

[Ryan]

WHo cares. 3dmark is just a synthetic becnhmark to make you feel good, it's not an accurate judge of preformance.

 

[DannyBoy]

Originally posted by: Hammyton
My 3dmark in january was 9998, but now its 8393!!
I have the exact same system components!
I uninstalled and changed to omega drivers and now i changed back and all of a sudden my 3dmark sucks! I did a clean uninstall of the previous drivers, and I am sure AA and Aniso are off....
Any ideas?
9998
8393
(note i just realized that the 8393 was 10mhz lower memory clock for the g4 than the 9998, so i benchmarked again with the changed, but my score only went up 100 points)....

Sorry to sound sorta blank but, oh well, thats life

 

[Hammyton]

I realize the scores may not actually reflect ingame actual performance, but I do believe a drop in scores may reflect a drop in in-game performance. Even if the benchmarks within 3dmark do not accurately portray real ingame situations, a drop in those scores should signify that I am getting lower performance now than I did when i had a 10k 3dmark score....

 

[Ryan]

Honestly, the frames are going by so fast that you cannot even perceive the FPS loss.

 

[Hammyton]

Originally posted by: rbloedow
WHo cares. 3dmark is just a synthetic becnhmark to make you feel good, it's not an accurate judge of preformance.

If this is true perhaps in a "real world" situation the frames will drop enough to where I will notice the difference? I am just looking for some possible explanations or solutions to my dilemna...

 

[chizow]

Check for background bloatware and other processes in task manager that may be eating up clock cycles. Also try something like the driver destroyers that have been linked a few times (Mem has posted it a few times, search Video or GH for his name as the author). I agree, a 1600 pt drop is nothing to sneeze at, and is certainly a good indication of performance degradation (at least using 3DMark2K1. 2K3 just means you switched your GPU or OC'd it).

Chiz

 

[NOX]

Originally posted by: Hammyton

Originally posted by: rbloedow
WHo cares. 3dmark is just a synthetic becnhmark to make you feel good, it's not an accurate judge of preformance.

If this is true perhaps in a "real world" situation the frames will drop enough to where I will notice the difference? I am just looking for some possible explanations or solutions to my dilemna...

The solution is don't use 3dmark to determine the real world gaming performance of your system, instead use a real game.

 

[BmXStuD]

defrag/ scan for adware/ scan fr virus

 

[Hammyton]

just did all of those thanks for the suggestion tho I do appreciate it...
And from now on I will be benchmarking using FPS's

 

[NOX]

Originally posted by: Hammyton
just did all of those thanks for the suggestion tho I do appreciate it...
And from now on I will be benchmarking using FPS's

That's it, really, use FPS. It will give you the best indication of what your system is capable of. If you want to find out if your system is capable of the newer games, uses games like UT03 with its build in benchmark. Games like AA, where you can use a command line to "stat fps", will display your frames etc...

3dmark, well, is 3dmark, an eyecandy demo.

 

[BoberFett]

I think this guys point was...

He was running the same program he did before, and was getting slower times. If he ran UT2003 and got a 10% lower score than last month, that would be cause for concern. If ripping MP3s was taking longer than it did before, that's cause for concern. Fact is, if doing the exact same thing is slower now than it was a month ago, you might want to check into it. Regardless of the fact that 3DMark isn't an ideal benchmark, it shouldn't change from month to month.

 

[NewSc2]

Originally posted by: BoberFett
I think this guys point was...

He was running the same program he did before, and was getting slower times. If he ran UT2003 and got a 10% lower score than last month, that would be cause for concern. If ripping MP3s was taking longer than it did before, that's cause for concern. Fact is, if doing the exact same thing is slower now than it was a month ago, you might want to check into it. Regardless of the fact that 3DMark isn't an ideal benchmark, it shouldn't change from month to month.

Finally a good response... My first assumption would be that you have a lot of hidden programs running in the background. I know for a fact that my computer's been slowing down lately... looks like high time for a reformat.

 

[Bojangles139]

hehe, i still love the fact that after 80fps, the human eye can't tell any difference's. :Q

brandon

 

[Confused]

Originally posted by: Bojangles139
hehe, i still love the fact that after 80fps, the human eye can't tell any difference's. :Q

brandon

This has been argued MANY times in the Video forum, and every time I've seen it, it has been proved that eyes don't "see FPS", we are always scanning, so we can in fact tell the difference between say 80 and 200.


Confused

 

[NOX]

Originally posted by: BoberFett
I think this guys point was...

He was running the same program he did before, and was getting slower times. If he ran UT2003 and got a 10% lower score than last month, that would be cause for concern. If ripping MP3s was taking longer than it did before, that's cause for concern. Fact is, if doing the exact same thing is slower now than it was a month ago, you might want to check into it. Regardless of the fact that 3DMark isn't an ideal benchmark, it shouldn't change from month to month.

True, however knowing that 3dmark has done this many times in the past for many users, and proving time and time again it was actually 3dmark and not their systems indicates to me that 3dmark can't be trusted and should be treated for what it really is, a Demo. If he had stated that he noticed a substantial drop in performance in all his games then I would think that fact alone would be cause for some valid concern.

3dmark has done this to many users in the past, that?s nothing new. If he had stated anything as your example of ripping MP3's suggested, then I'm more then positive the responses would have been much different.

 

[Dreadogg]

Originally posted by: Bojangles139
hehe, i still love the fact that after 80fps, the human eye can't tell any difference's. :Q

brandon

So what does that say about anything, not many care for peak FPS, but most want constant FPS above 80 in all games. Or to be able to run Vsync with thier 120 hertz refresh rate and have it stay constant, this does make big difference in FPS games.

 

[arcenite]

If you're really that worried about it just redo your system with the original drivers (that gave you the increase). Like everyone else said, it's only numbers. The most important thing is what you see.

Bill

 

[Chad]

Don't sweat these guys... every since nVidia has lambasted the benchmark lots of people have followed their pr bs like little ducks. Most of these people apparently are incapable of seeing your point. In apples to apples, when you lose that many points in 3dMark compared to a previous 3dMark score on the exact same system, something is not quite right and you *HAVE* lost performance somewhere. Granted, 3dMark on it's own isn't that big of a deal, but when you compare your score to a previous score on the exact same system and there is a big change, somethings wrong and you're right for feeling so.

A lot of people will stick their fingers in their ears and go "lalalalalalal lalalalalala can't hear you lalalala lalalalalala" but nevermind them, they probably can't help you anyways.

As for the problem, I think it has to do with those Omega drivers. Try doing an anal uninstall of your video drivers, there are instructions for ATI's and nVidia's cards on rage3d.com forums.

 

[thorin]

You probably loaded up a bunch of crap software that loads a crap load of stuff at startup, fireup msconfig and then edit your windows services setup.

Thorin

 

[Lalakai]

depending on your OS, you could be accumulating alot of bloat as chizow and thorin pointed out. W98 is notorious for never really deleting stuff, and letting useless stuff just keep gathering.

try the defrag and scan recommendation, along with a good scan for virus and spy-ware (lava soft has a decent free spy ware program). If you're still not getting your frames back, try and reinstall your OS to see if that would wipe out some of the hidden bloat.

 

[fkloster]

My 3dmark dropped 1600 points in 1 month with no hardware changes

How about software/driver changes? Performance settings?

 

[mchammer187]

Originally posted by: NOX

Originally posted by: BoberFett
I think this guys point was...

He was running the same program he did before, and was getting slower times. If he ran UT2003 and got a 10% lower score than last month, that would be cause for concern. If ripping MP3s was taking longer than it did before, that's cause for concern. Fact is, if doing the exact same thing is slower now than it was a month ago, you might want to check into it. Regardless of the fact that 3DMark isn't an ideal benchmark, it shouldn't change from month to month.

True, however knowing that 3dmark has done this many times in the past for many users, and proving time and time again it was actually 3dmark and not their systems indicates to me that 3dmark can't be trusted and should be treated for what it really is, a Demo. If he had stated that he noticed a substantial drop in performance in all his games then I would think that fact alone would be cause for some valid concern.

3dmark has done this to many users in the past, that?s nothing new. If he had stated anything as your example of ripping MP3's suggested, then I'm more then positive the responses would have been much different.

still if you had a system that got 3000 points in 3dmark 2k1 at default settings surely you cant do any gaming and i think everyone would agree that obviously you have a problem if it was previously 10000

3dmark may not be extremely precise but IMO it is still a good indicator of system performance

1600 points is a signifcant drop to me

 

[NOX]

Originally posted by: mchammer187

Originally posted by: NOX

Originally posted by: BoberFett
I think this guys point was...

He was running the same program he did before, and was getting slower times. If he ran UT2003 and got a 10% lower score than last month, that would be cause for concern. If ripping MP3s was taking longer than it did before, that's cause for concern. Fact is, if doing the exact same thing is slower now than it was a month ago, you might want to check into it. Regardless of the fact that 3DMark isn't an ideal benchmark, it shouldn't change from month to month.

True, however knowing that 3dmark has done this many times in the past for many users, and proving time and time again it was actually 3dmark and not their systems indicates to me that 3dmark can't be trusted and should be treated for what it really is, a Demo. If he had stated that he noticed a substantial drop in performance in all his games then I would think that fact alone would be cause for some valid concern.

3dmark has done this to many users in the past, that?s nothing new. If he had stated anything as your example of ripping MP3's suggested, then I'm more then positive the responses would have been much different.

still if you had a system that got 3000 points in 3dmark 2k1 at default settings surely you cant do any gaming and i think everyone would agree that obviously you have a problem if it was previously 10000

3dmark may not be extremely precise but IMO it is still a good indicator of system performance

1600 points is a signifcant drop to me

I'm not sure I understand the 10000 to 3000 example you make. 1600 point is nothing for this program; I?ve seen times were people have reported 3000 point deduction, yet all there games play the same. I don?t understand why people put so much emphasis on a benchmark that has really no substance in a real world situation. Case and point http://subscriber.anandtech.com/messageview.cfm?catid=27&threadid=951695&highlight_key=y&keyword1=3dmark.

I would suggest to Hammyton to uninstall 3dmark01, reinstall it and run a test, and see what kind scores he gets after the reinstall. Will that change anything, yeah maybe his 3dmark scores. Will it change his real games, absolutely not!

 

[BoberFett]

Originally posted by: NOX
I'm not sure I understand the 10000 to 3000 example you make. 1600 point is nothing for this program; I?ve seen times were people have reported 3000 point deduction, yet all there games play the same. I don?t understand why people put so much emphasis on a benchmark that has really no substance in a real world situation. Case and point http://subscriber.anandtech.com/messageview.cfm?catid=27&threadid=951695&highlight_key=y&keyword1=3dmark.

I would suggest to Hammyton to uninstall 3dmark01, reinstall it and run a test, and see what kind scores he gets after the reinstall. Will that change anything, yeah maybe his 3dmark scores. Will it change his real games, absolutely not!

Do you think computers have get PMS?

If identical hardware configurations running using identical software configurations have different scores, then they're not identical. Something is different. Computers don't have personalities. They don't get tired. They do what they're told as fast as they can. If something changes between one process and an identical process run later, then something changed.

We're not arguing the merits of 3DMark as a benchmark. We're discussing the fact that something changed on his system. What's so hard to understand?