Will it make a big difference if ...

[kevvvvv]

... I upgrade my P4 2.4 1gb Ram machine to a hyperthreading 3ghz 2gb Ram machine?

Or should I wait for cpus to move up another gear?

My current machine was professionally configured, and while it's okay I'm still running out of resources and wondered if a 3 gig hyperthreader would make a worthwhile difference.

I use loads of midi and cpu sucking VSTIs and samples.

thanx = kev

[DevonB]

What FSB on your 2.4ghz? 400, 533, or 800? If it's 400fsb, and you go to 800 on that 3ghz (I'm assuming 3.0 800fsb, and not a 3.06 533fsb), you're going to see a VERY nice jump indeed.

Devon

[Alive In Chernobyl]

Hyperthread slows many many programs. You add the extra time to assemble and disassemle multiple cpu thread for one cpu. This takes speed from the computer. The clock speed will help.

[Muff Wiggler]

my personal rule of thumb is hold off upgrading 'till you can triple your power

has served me well over many years and many upgrades, and avoided me spending money to get a depressingly small bump in power for my hard earned cash

[quincy]

I did a little research on hyperthreaded processors for work, was thinking of building a new server. The general vibe i got from my look into it was that hyperthreaded P4's are good for multiple taks on a machine, very smooth etc, but don't make a huge amount of difference(for the money) with a single dedicated task.

So depends on how you work i suppose. If you use a single host it may not be worth the money, but if you use many different sample editors, different hosts at once or whatever, it may make a decent difference.

Hope this helps. Let it be known i am not a hardware expert, just what i gathered from what i read.

[DevonB]

Hyperthread slows many many programs. You add the extra time to assemble and disassemle multiple cpu thread for one cpu. This takes speed from the computer. The clock speed will help.

Depends on the app. I think Kev is on Cubase? That can take advantage of hyperthreading, and I've seen a 10% boost of power.

Devon

[Alive In Chernobyl]

The increase is to the cpu meter only. If you do test with stability then at lower cpu meter reading the system become unstable.

This is because of the idea I try to describe earlier. The host sees 2 processor and splits the data for it. This data still is assemble for one cpu. The assembling takes more cycles. Host thinks it is saving time because it does not see the lost cycles. When the number crunching comes to the wall you will see much lost power. The program will act like it does at 100% even if you are really only at 80%.

[kevvvvv]

Alive, Devon, Quince, Muffie ... thanx v much for your ideas and thoughts.

My current PC is over 2 years old now, so it won't have the latest fast boards.

So it seems a new machine will be a little faster, but the hyperthreading isn't something that SX and all my plugs won't take best advantage of.

I guess I'm impatient. We haven't seen the usual doubling in power in the last 2 years that Intel and Co usually produce.

thanx again

[DevonB]

Alive, Devon, Quince, Muffie ... thanx v much for your ideas and thoughts.

My current PC is over 2 years old now, so it won't have the latest fast boards.

So it seems a new machine will be a little faster, but the hyperthreading isn't something that SX and all my plugs won't take best advantage of.

I guess I'm impatient. We haven't seen the usual doubling in power in the last 2 years that Intel and Co usually produce.

thanx again

Hey Kev, you didnt' answer which FSB you had. If you're at 400, and you do jump to an 800, you ARE going to see a pretty significant jump. My friend did the same jump you did, from a 2.26 to a 3.0, and he's quite happy.

Devon

[kevvvvv]

Aahh

thanx = kev

[Alive In Chernobyl]

Fast centrino will double your power. Centrino is much faster than clock speed.