depends what your CPU is running at now. but even if you do get 200-300 more Mghz, I doubt you'll "feel" any difference. I'd recommend adding more ram before ditching your 2500+
0
Straight_ManGeeky, in my own wayNaples, FLIcrontian
edited March 2004
Right Kanez.
Bud, to make a CPU upgrade really worthwhile, figure you would need a 3000+. Reason is, eyes do not see that minor a change. The Barton 2500+ rocks with a Gig of RAM, DDR333 or thereabouts, and might benefit from DDR400 on certain boards that can run faster RAM than CPU base rate....
Not as fast as my P4, but the silly Electron Microscope cpuspeed thing insists THAT is running at 3.428 true GHz (3511 MHz)-- the Abit GAME ACCELLERATOR (the latest BIOS for the IC7-Max3 is what has that in it) is OCing it more than I am, as it is in reality a 2.8 GHz CPU which I set as a 3.2 GHz CPU. To that I just say, W00T!!!
I'll agree with all the others. The standard rule-of-thumb these days is that you won't see a noticeable difference with any upgrade less than 400MHz.
What I try for is a minimum true CPU MHz rate that is 1.3 times or more than what I have now. Then I see it, and I know what I am looking for. For a user, or enthusiast who is not going to just use benching to see changes, I say a 1.5X CPU MHz increase is a good minimum. The P4, I went from a base of 1.8 Willamette which ran at about 2.1 GHz max, and stepped up to a 2.8 Northwood which runs at 3.2-3.4 GHz depending on what benchmark you throw at it. Prof, your 400 MHz increase is close to my 1.3X minimum given a Barton 2500+ starting point, excellent advice.
The nice thing is, the Game Accellerator in the Abit BIOS is affecting Gromacs folding significantly (much better)-- it ramps up the CPU such that SSE2 and SSE things are accellerated, which is what Gromacs WUs ARE doing. I don't game, but the BIOS'd Game Accellerator helps speed up Gromacs WUs, the layered vector text work I do (Desktop Publishing, things like whole brochures in CorelDraw 12), and that kind of thing. That DTP work and working with vector graphics format Photo files make use of SSE2 also (FPU side of CPU).
But, faster speed RAM, fastest board will take with current CPU, and up to a GIG of it, would be good for any extremist's multipurpose box, and probably cheaper than the CPU upgrade alone would cost. Second, that RAM, if you choose right, will work with better CPU later also.... So, RAM now, maybe CPU later as prices drop on next higher increment that would give visible benefit. WAITING on CPU would lead to price drop on same, probably, PC3200 RAM is probably near bottom of price curve so it is more economical than CPU as well as best idea. Figure more, faster RAM, is more likely to help box be more stable while doing more things at once, and also likely to be a carryforward investment unless you jump whole generations in a relatively short time.
You already have a damn good system right now. I really don't see any reason to upgrade myself. Save your money for something else, or buy some new speakers or something. I see no need in wasting money on a upgrade that you really will not notice.
If you absolutely must upgrade something, maybe get a full-fledged raid controller (comes with it's own memory, not like the lightweight raid built into some MB's) and a couple screaming-fast HD's.
But I agree with the others. 99% of the people in the world would LOVE to have the system you have now. :thumbup
Comments
Bud, to make a CPU upgrade really worthwhile, figure you would need a 3000+. Reason is, eyes do not see that minor a change. The Barton 2500+ rocks with a Gig of RAM, DDR333 or thereabouts, and might benefit from DDR400 on certain boards that can run faster RAM than CPU base rate....
Not as fast as my P4, but the silly Electron Microscope cpuspeed thing insists THAT is running at 3.428 true GHz (3511 MHz)-- the Abit GAME ACCELLERATOR (the latest BIOS for the IC7-Max3 is what has that in it) is OCing it more than I am, as it is in reality a 2.8 GHz CPU which I set as a 3.2 GHz CPU. To that I just say, W00T!!!
John D.
What I try for is a minimum true CPU MHz rate that is 1.3 times or more than what I have now. Then I see it, and I know what I am looking for. For a user, or enthusiast who is not going to just use benching to see changes, I say a 1.5X CPU MHz increase is a good minimum. The P4, I went from a base of 1.8 Willamette which ran at about 2.1 GHz max, and stepped up to a 2.8 Northwood which runs at 3.2-3.4 GHz depending on what benchmark you throw at it. Prof, your 400 MHz increase is close to my 1.3X minimum given a Barton 2500+ starting point, excellent advice.
The nice thing is, the Game Accellerator in the Abit BIOS is affecting Gromacs folding significantly (much better)-- it ramps up the CPU such that SSE2 and SSE things are accellerated, which is what Gromacs WUs ARE doing. I don't game, but the BIOS'd Game Accellerator helps speed up Gromacs WUs, the layered vector text work I do (Desktop Publishing, things like whole brochures in CorelDraw 12), and that kind of thing. That DTP work and working with vector graphics format Photo files make use of SSE2 also (FPU side of CPU).
But, faster speed RAM, fastest board will take with current CPU, and up to a GIG of it, would be good for any extremist's multipurpose box, and probably cheaper than the CPU upgrade alone would cost. Second, that RAM, if you choose right, will work with better CPU later also.... So, RAM now, maybe CPU later as prices drop on next higher increment that would give visible benefit. WAITING on CPU would lead to price drop on same, probably, PC3200 RAM is probably near bottom of price curve so it is more economical than CPU as well as best idea. Figure more, faster RAM, is more likely to help box be more stable while doing more things at once, and also likely to be a carryforward investment unless you jump whole generations in a relatively short time.
John D.
2500+ with sk-7
abit nf7-s
1 GIG pc4000 kinston hyperx
ati 9800 pro
seagate 160GB sata
But I agree with the others. 99% of the people in the world would LOVE to have the system you have now. :thumbup