Try to do a test where you run a demanding game like Doom 3 and in the background you do video encoding and browsing the web.

Then see what FPS AMD and Intel processors get.

And then overclock the FX55. Smoking!!!

Hey sander,

How high can you get the FX-55 , and what are the benchmarks with the same cooling as the P4?

Can you do that or provide us with your data?

How is it physically possible to browse the web while you play Doom 3?

You can have pages open eating up memory.
You could drain the cpu with flash games and ads in the background I guess, but WHY?

But you can't BROWSE while you're playing the game. The minute you start browsing you stop playing the game. Such a test, regardless of its effect on fps, is pointless.

As for video encoding, you would want the fps from the video encode, not Doom 3.
Why? Anyone who has done encoding while playing games knows you set the vidoe encode to IDLE, and play your game. Why? Because the game is a REAL TIME thing. It easts up CPU in real time, as you play. The game may leave some of the cpu unused. The video encode is not real time (unless it's a live stream). If there is extra room for more work do be done, it takes it. (Unless your hard drive becomes the bottleneck, which is virtually impossible when you do any serious encoding)

Basically, those are pointless tests.

I'm with Brian on this one, some tests are pointless as they do not represent an actual situation. PCmark04 actually is pretty good at giving an accurate presentation of use. The only downside is that they factor memory performance in too much, skewing the scores in favor of the P4 whereas in real life you see no benefit of this faster memory, so it helps to look at the individual scores. For comparison, the 5.2GHz P4 scored 8294 PCmark04 points, but as you can see from the other benchmarks that's not saying much.

Umm to point out an exsample, What I do on testing my systems,...

I run 4 servers of it and run all service's including on access scanning, and then run, Aquamark3, and PCmark03 or 3dmark04 or 3dmark05

This is about as multi tasking as you can get.... But still, I get great results,... I just got 4334 from 3dmark 05, but I just sorted out a error (me doing things my way was a touch wrong) now I clocking 2.54 benching then moveing to 2.6Ghz - BTW im running a 3200+ and a 6600GT and 512MB DDR 333, and a standard IDE drive.. so its not actual blistering performance hardware..

I didn't say play Doom3, I said run a Doom 3 BENCHMARK in a window or full screen alt tabbed, and then do more tasks.

This way you can test how responsive the system is while peforming a doom 3 benchmark or other game benchmark.

toms a while back go tthe FX-51 upto 3.8 GHZ

With similar cooling.
I'm Intreged ::

Multi-task test makes sense

Set up a macro command to do say 5, or 6 office, editing/encoding, & printing tasks @ once while you browsing, & see which one chokes first

OCGW

PEACE

even though i agree with you regarding the POS that is prescot, WHY do you skew your test results by even further crippling the preschot with high latencies (4, 4, 4) even though using OCZ DDR2 can do better ?

What are the latecies on that OCZ you are referring to?

Ya knooow you should have gave the spec when you posted your comment?

Ya' gonna' make us google it? lol

OCGW

PEACE

I didn't say play Doom3, I said run a Doom 3 BENCHMARK in a window or full screen alt tabbed, and then do more tasks

Even more pointless.
Who cares about a benchmark that has NO bearing on an actual gameplay situation?

For a multitasking test try running a BOINC app, setiathome, climateprediction, etc. while playing a game, or while watching a movie or a few video clips and download an iso for a DVD. To really test multitasking set the boinc app to run in normal mode.

Are there differences now?

Yeah I agree it was not fair to run the DDR2 at such crap latencies. The quickest DDR2 afaik is the OCZ PC2-4300 which is rated at 3,2,2,8. I've heard that there's new stuff gonna be available very soon that can run up to 3,3,3,4.

Cheers,

Suman

Not fair to run the DDR at crap latencies? That's a 288MHz front side bus, let's see you pull that off at anything but 4-4-4-11 settings, it just wouldn't work at anything else. And these ARE the OCZ PC4300 DDR2 DIMMs, not some cheap aftermarket brand. But in reality faster timings will not have helped the P4, as it still has almost 40% more memory bandwidth than the AMD.

Yo' Sander how about running that test w/ the Intel's DDR2 set to lower fsb, & low latencies stteings

That would be a worthy experiment, I think

OCGW

PEACE

How about their done and its pointless to do anything from here on out. How about Run minesweeper while running Doom 3. Then when thats done run minesweeper while running half life 2 then run half life 2 and doom 3 together then go backa nd do all those test again with diffrent latencies then go back and do it with diffrent FSB then go back and do it while surfing the web doing the 3-4 test agan that way.... see where this thread is going?

Sorry guys, I'm not running more tests this week, took it all apart again as the lab was filled with phase change coolers, buckets with water and water chillers for the chipset etc. I couldn't even sit at my desk properly.

I gotta ask this..... with the P4 660 stock clock at 3.46 and then pushed to 4ghz and then to 5.2ghz .... and also with the FX55 stock clock at 2.6 and pushed to 3.46ghz or faster........ How long will these processor last as in longevitity.???????? If they were ment to run at the stock speeds with a certain mean time failure so by pushisng them to the limits would'nt shorten that mtf ????? sure is a lot of risks for $700 plus processor chip.

These were run in a controlled environment cooled down to -40 degrees celsius, but only for a short amount of time, enough to run the benchmarks. Obviously this is not something you'd be able to do 24/7.

did I read that correctly??... minus 40 deg????..... dear lord, it takes that much effort to cross the 4GHz barrier? .... guess we aint gonna see any mainstream CPUs clocking at that speed for a while eh.

I must ask though, is there such a big difference in internal technology (I can think of only the HTT and wide deep pipelines.. perhaps the integrated mempry controller as well) or manufacturing process that sets the AMD so much ahead of Intel??.. Both have been around for ages, I'd expect them to know all about micro-design by now... atleast be on par on the knowledge bit.