A64 X2 4800+ @ 3 GHz & PCMark04: 8548

[urbanfox]

that will be interesting to see.
I never heard of those.

but considering it's between the PSU and the wall socket, wouldn't it be measuring the entire PC's load, not the CPU specifically??

Yep, so the readings wouldn't be able to pinpoint just the CPU, and that's way too high for an AMD DC anyway.

[bachus_anonym]

that will be interesting to see.
I never heard of those.

but considering it's between the PSU and the wall socket, wouldn't it be measuring the entire PC's load, not the CPU specifically??

You can get those in US also (it's called Kill-A-Watt Meter). I have one of those hooked up, too.
I kind of thought that Sampsa measured CPU's power consumption (400W is way too high). This can be measured. I think folks @ LostCircuits explained it in their article about Venice. They took current measurements from 4-pin 12v connector.

[jjcom]

Congrats, nice seeing how high these things are clocking on low voltages

jjcom

[AcEmAsTr]

i was expecting slightly faster than 28seconds at 3ghz from these cpu's

[Kunaak]

thats a cool gadget.

I found it here on google.

http://www.ccrane.com/kill-a-watt-accessory.aspx

[[XC] moddolicous]

wow, nice results. Download Rodzilla's D2OL personnel 2x installer, and crucnh with it. We need more volts!

[M.Beier]

Hmm, how the heck did you get that CPU ? - And the 840 as well... ??

- They sure aint forsale in scandinavia :\

-12k

[vapb400]

i was expecting slightly faster than 28seconds at 3ghz from these cpu's

superpi isn't multithreaded, so there wouldnt be much gain over a a64@3ghz

[Kunaak]

if you want a Multi Threaded PI program download Systool.
then check the box that says "run 2 instances" or something like that.

works great if you wanna test hyperthreading, or dual cores, or dual CPU's.

systool can be found in this thread.
one of the best programs I seen for awhile.

http://www.xtremesystems.org/forums/...ad.php?t=61549

[Bennah]

That little simple gadget is good, measuring watts. I suppose sampsa would be a custom job though. The fins are good at the power side of the pc's etc, take the custom 13**w psu they have

btw nice background Kunaak

[STEvil]

You can get those in US also (it's called Kill-A-Watt Meter). I have one of those hooked up, too.
I kind of thought that Sampsa measured CPU's power consumption (400W is way too high). This can be measured. I think folks @ LostCircuits explained it in their article about Venice. They took current measurements from 4-pin 12v connector.

that leaves out the +12v supply on the ATX/BTX header.

[bachus_anonym]

that leaves out the +12v supply on the ATX/BTX header.

good read... Lost circuits reviewing 4800+, including power consumption

[Kanavit]

why doesn't 28secs sound that good to me, in 1mb super pi for the A64 4800+ X2??

plus Fugger still holds PCmark recored with his Intel system i believe. You know what the A64 FX-55 still beats out the A64 X2 dual core in some games if my memory serves. I am very hesitant to jump on this dual core bandwagon yet. Let us see some 3dmark03, and '05 scores for more convincing.

[jjcom]

Those programs are single threaded so it will preform like a 3ghz A64. It will only show it's strength with multi-threaded apps

jjcom

[Kunaak]

why doesn't 28secs sound that good to me, in 1mb super pi for the A64 4800+ X2??

plus Fugger still holds PCmark recored with his Intel system i believe.

1. Super Pi isn't Multithreaded.
few things are...

Ceinebench, Sisoftsandra and Systool CPU Bench are the only benchmarks I can think of thats multithreaded.
single threaded things should show little improvement, and will produce results similar to a single core CPU clocked at the same speed.

2. theres a hell of a difference between a ully functional overclocking board, and a cascade that does -120C, and a board that can't adjust it's Vcore, and a plain old prometeia.

for being single compressor cooling, and stock Vcore, it looks like this thing can easily take the PCM04 record with a cascade or LN2

[bachus_anonym]

As for more multi-threaded benchmarks...

PCMark04 runs 3 sets of 2 tests, one thread for each:

File Compression & File Encrypting
File Decompression & Image Processing
Virus Checking & Grammar Check

So, this benchmark also shows pretty big gain on systems with HyperThreading, Dual CPU or DualCore. In case of A64 DualCore, it's about 30% better score compared to single CPU @ same clocks.

Sampsa showed that to us as well as many reviewers.

That's why P4 HT setups kill A64 in PCMark04.

[Kanavit]

thx Kunaak, here is a better representation how much faster the A64 X2 is in PCmark04. it's true that the X2 will only be faster in applications that support mult-threading environment. looks like i'm still going FX-55 atm.

wow



http://www.xbitlabs.com/articles/cpu...on64-x2_9.html

[Kunaak]

your comparing a 2.4 ghz CPU, against a 2.6 ghz CPU on a single threaded benchmark.
of course the FX55 is gonna be faster at it, its clocked higher on a single threaded benchmark.

[JohnyKnoxville]

I cant wait!

[(sin)morpheus]

Wow, very nice results. Now you've got me considering a dual core, curses.

[Sampsa]

that will be interesting to see.
I never heard of those.

but considering it's between the PSU and the wall socket, wouldn't it be measuring the entire PC's load, not the CPU specifically??

Yes since it's between the wall socket and power supply, it's measuring the entire PC. But since the setup is same.. I can compare how much more A64 X2 4800+ takes power than FX-55 and if using just single core compared to two cores.

i was expecting slightly faster than 28seconds at 3ghz from these cpu's

Spi doesn't support HT/DC and I was using crappy memory settings because of A8N-SLI DLX only allows 3V to memory (CAS 2.5-3-3-8).

- They sure aint forsale in scandinavia :\

That's true

[Walrusbonzo]

Yes since it's between the wall socket and power supply, it's measuring the entire PC. But since the setup is same..

Remember PSUs aren't 100% efficient. I'd imagine a 550w Antec is about 75%, so although it was drawing 440w, the actual PC components are probably only drawing about 330w. The other 110w is the heat given off by the PSU.

Damn fine results with the CPU though

[Sampsa]

Remember PSUs aren't 100% efficient. I'd imagine a 550w Antec is about 75%, so although it was drawing 440w, the actual PC components are probably only drawing about 330w. The other 110w is the heat given off by the PSU.

Damn fine results with the CPU though

Yes but I should still be able to compare power consumption between A64 X2 4800+ and FX-55 if I only change the CPU in the system, right?

[ettis]

Good to have an AMD46 X2 4800+ @ FX57. That's TWO FX57 in one computer!!!!

[Sampsa]

Switched to DFI LP nF4 SLI-DR and just ran spi1m @ 3,1 GHz (27,5 sec) and PCMark04 @ 3040 MHz score is 8987

Looks like I need better cooling than Vapochill LS and better PSU than Antec 550W if I want to go higher.