I often do multi-tasking on my laptop, ie having a video rendering in the background, or encoding MP3s, while still using Office applications in the foreground.

And I notice sluggisgness in my 1.7 Dothan.

I have a chance to get a very good price of a 2.0 Dothan A190, or for about the same price I could get a p4 3.4 or an Athlon 64 3400 or 3700.

So which, in a multitasking environment as described, would perform best?
Gaming is not important here.

Cheers

Steve

Amd64 !!!!!!!!!!!!!

Sorry, I guess I should have phrased my question a little more carefully.

Does anyone have objective benchmarks or hands on experience in a multitasking environment?

there have been a few reviews of multitasking. The p4 comes out on top in that respect. In regards to multitasking and video/audio encoding, the p4 is still better.

http://www.extremetech.com/article2/...1478694,00.asp

That, and some others I've seen, seem fairly convincing for the P4 over the AMD. I wonder how a Dothan goes.

Steve

For multitasking, Dothan is not as good as AMD64/ P4. Test result shows.

Hardly. While the Dothan is a great chip (I'm typing this on a Dothan 1.7), it's not as good as a P4 at multitasking.

The P4s have hyperthreading, which gives it superior multitasking ability. Even if it's not faster with hyperthreading than other processors overall, a P4 in HT mode can run two threads simultaneously instead of resorting to pre-emptive multitasking on the operating system level. Because of this, interrupts can occure on both threads simultaneously; the system will operate much smoother. Keep in mind smoother is not the same as faster. It may not be faster, it will simply be more responsive under heavy multitasking.

Exactly. In this respect, P4's have the advantage because of the Hyperthreading. AMD 64 does not have a clear advantage on multitasking.

dothans will be worse than amd 64 in multitasking. remember, the amd 64 has really fast access to memory with very low latencies.

I wish the P4's had more memory. Having the core share resources with two threads hurt's its performance. You’re talking about a chip that was originally designed to process a single thread. Then you have it process two threads with the same amount of resources.

Also, the P4 will run hotter, which could result in throttling. It won't be any good if you are multi-tasking and all of a sudden the P4 decides to slow down because it is getting hot. There goes your performance edge.

Too bad we didn't have Cell available for Laptops! (16 Cores!)

Well, there's only one way to find out for sure, and I'll make money out of it either way thanks to the joys of over-priced laptops in Australlia.

I'm on a trip to the States and have nabbed a 2.0 Dothan (A190) and a 3.06HT P4 (GRT360ZG), both Sony Xbrites.

Once I get them home I'll do some testing and see what happens.

Cheers

Steve

Steve,

I would love to see a comparison between those two setups, particularly how much of a speed difference there is rendering the same file. Should be interesting to see how well the Dothan does against a P4 with HT. I'll be looking out for your posts.

That's the thing. I'm saying that a processor's ability to multitask has nothing to do with performance, it has to do with responsiveness.

ChaosBlizzard: There is (usually) no performance HIT from running hyperthreading. You forget that even on non-HT processors, they still have to run multiple threads, just not at the same time. Threads are still competing for cache space and registers. Only they get swapped in and out repeatedly.

That said, the entire idea behind hyperthreading is that one single thread is rarely able to fully use all the processor's resources. The intention of HT is that two threads can use the existing resources more fully than one thread. This is where HT's 5 to 10 percent speed boost comes from. However on top of that you are truely running two threads simultaneously, which leads to greater responsiveness; for one thing each thread can have interrupts happening on it, so if one thread is trying to use up the entire CPU, the OS threads on the second virtual CPU can still receive interrupts.

A processors ability to multi-task does have to do with performance. What do you think responsiveness falls under?

Why do you think Sony, IBM, Toshiba and others are developing a 16 core Cell processor? So things can respond faster? You do realize that processor is over 100x's faster than a P4 (Performance). This is simple because it can handle 16 threads at once.

What do you mean still receive interrupts? Regardless of what a processor is doing, if it receives an interrupt, it must stop. This is how it works on any pre-emptive system.

One thread can't use a single core's resources huh? Go out and buy a normal P4, and see if you can't hit 100% utilization. I bet you will be able to, and seeming how that is a single thread I guess that means it can utilize the chips resources quite well.

I run two P3's at 1.0GHz apeace. They don't just respond faster, they ARE faster.

Just because a CPU is at 100% utilization doesn't mean it's resources are being fully utilized. Is the cache empty or full? Is the branch predicter mispredicting? Is the pipeline full? Etc.

Anyhow I was referring more to software interrupts. If one thread is totally monopolizing the CPU, and the OS tries to send a message to an app that a mouse click has occured in it's window, it might not receive it until the OS can get the first thread off the CPU. If there is a slot open for a second thread (HT), it can do it without needing to wait for the first thread to be pre-empted. At least, this is my understanding of it. And my observations have backed by observations of how one runaway thread can totally hose system responsiveness if you let it.