Let's count how much power we have! (another team has over a million ghz)

zevia: E6400 @3.2GHz, E6300 @3.15GHz, P4 3.2GHz, P4 3GHz.


PS: how do we count them? lol

I'm not sure how to count dual cores. I think all dual cores should count twice, because technically, they do 2 things at once with 2 separate processing units. Where it gets confusing for me are the P4's with HT. Cuz they are 1 processing core doing 2 different things at once.

Anyway, here are mine:

Athlon XP @ 2.2GHz, P4 @ 1.7GHz, P4 @ 1.7GHz, Celeron @ 2GHz, P4 @ 2.66 (not 24/7), P3 @ 933MHz, P3@ 866GHz.

Counting Zevia + mine (dual cores count twice) = 30.96 GHz
Counting Zevia + mine (dual cores count once) = 24.61 GHz

I think dual cores should count twice since they technically are 2 cpus and HT should count once

FX-60 @2.6 (I'm trying to get my hands on some better cooling so I can OC )
A64 3700 @2.35 (hasn't been running much lately, I wish I had a better mobo cus this chip can do sooooo much more than 2.35 )
P4 @2ghz

Running Total = 40.51ghz

I forgot to add my P3 900MHz laptop I have. Me stupid.

Running Total = 41.41 GHz

3.40 gig Pentium D here, and soon to add a Sempron 1.5 gig.

I'll let someone else do the maths.

After added 3.40Ghz x2 + 1.5Ghz, my calculator says 49.71 Ghz total.

Thanks ...my typing sucks, of course that's a Sempron

I'd sneak F@H onto my mum's Celly 1.2 gig as well, but she likes to play a game called Arcade Bubbles, which I know from personal experience doesn't play well with F@H

You probably better off using the new windows SMP client on your Pentium D, Arachne. In my experience they use more CPU power but I don't notice any slowdown (unlike VMware), plus you will get to help newer projects and more ppd (points per day).

WOW I want that one too

Cool, I was looking at that thread. I'll look into it later

Edit: downloading now at a snail's pace

LOL @ H3rB3i - me too

heehee, yeah, well most of my machines are rescues from the garbage so...

you could install the text-only console as a service, she'll never even know its there, unless she opens task manager

FX-60 @ 2.90ghz, 3800X2 @ 2.25ghz.

60.01ghz running total.

HT is just one core, but multi-threaded input to ensure the core is utilised 100%, rather than just 90% with 10% waiting for the rest of the system to ctach up.


Alright .. lets see.

I have:
Turion x2 1.8Ghz Lappie = 3.6Ghz
Athlon x2 2.1Ghz Home Beast = 4.2Ghz
Athlon XP2100 -> 1733Mhz Home Server = 1.73Ghz

Athlon x2 2.0Ghz Work Beast = 4Ghz
Intel P4 3.0Ghz HT - Work Server = 3Ghz
Intel P4 2.8Ghz - Cad Server = 2.8Ghz
Intel P4 3.0Ghz HT - Cad PC 1 = 3Ghz
Intel P4 3.0Ghz HT - Cad PC 2 = 3Ghz
Intel P4 2.66Ghz - Work Machine 1 = 2.66Ghz
Intel P4 1.8Ghz - Work Machine 2 = 1.8Ghz
Intel PM 1.73Ghz - Work Machine 3 = 1.73Ghz
Intel C 1.0Ghz - Retired Work Lappie. = 1Ghz

Grand total = 32.52 Ghz.

But keeping in mind that the work PC's are only on 8-10hrs/day (and only 5days a week), and my home beast is only on maybe 4hrs/day, we need scaling factors
So I really end up with ....
14/24*3.6 + 4/24*4.2 + 1.73 + 10/24*5/7*( 4 +3+3+2.66+1.8+1.73) + 3 + 2.8 + 1 = 16.15 Ghz

HT is just one core, but multi-threaded input to ensure the core is utilised 100%, rather than just 90% with 10% waiting for the rest of the system to ctach up.

Similarly, Dual Core has two seperate CPU's, but decreases performance due to delays with the shared IO that aren't present on a single core processor.
A factor of 1.7 should apply for x2 CPU's .. but oh well .. it's getting a bit complex


Alright .. lets see.

I have:
Turion x2 1.8Ghz Lappie = 3.6Ghz
Athlon x2 2.1Ghz Home Beast = 4.2Ghz
Athlon XP2100 -> 1733Mhz Home Server = 1.73Ghz

Athlon x2 2.0Ghz Work Beast = 4Ghz
Intel P4 3.0Ghz HT - Work Server = 3Ghz
Intel P4 2.8Ghz - Cad Server = 2.8Ghz
Intel P4 3.0Ghz HT - Cad PC 1 = 3Ghz
Intel P4 3.0Ghz HT - Cad PC 2 = 3Ghz
Intel P4 2.66Ghz - Work Machine 1 = 2.66Ghz
Intel P4 1.8Ghz - Work Machine 2 = 1.8Ghz
Intel PM 1.73Ghz - Work Machine 3 = 1.73Ghz
Intel C 1.0Ghz - Retired Work Lappie. = 1Ghz

Grand total = 32.52 Ghz.

But keeping in mind that the work PC's are only on 8-10hrs/day (and only 5days a week), and my home beast is only on maybe 4hrs/day, we need scaling factors

So I really end up with ....
14/24*3.6 + 4/24*4.2 + 1.73 + 10/24*5/7*( 4 +3+3+2.66+1.8+1.73) + 3 + 2.8 + 1 = 16.15 Ghz

I've got a P4 2.8. Usually on about 7 hours a day.

I SUCK!!

Actually, the dual cores run at basically the same speed as single processors - the only difference is they respond quicker and are able to cope with multiple running process's better.

So the speed is not doubled!

Right, the speed is NOT doubled. However, a dual core has 2 cpus so we are counting them twice. And quads have 4 core so they should be counted 4 times.

Where did anybody say it was?

al: E660 @ 2.4hz icy cold yet with the power of a rhino (or two)

(currently 12hrs 4-5days per week expect lower in near future)


I'm not sure how relevant this is in practice but aren't clock speeds only arbitary values when it comes to CPU power OUTPUT - its how efficiently the proccesor works aswell eg Athlon CPU's are (power) rated by their Intel clockspeed equivalent rather than their actual clockspeed which is much lower.Likewise the new intel Gen of Conroes aka Core2duos are even MORE efficient to the point the E660 and above (without even overclocking) generally either equal or beat the performances of all other(previously top) Athlon + Intel CPUs, their clocks are even lower than the Athlons but more importantly there efficiency is leaps and bounds to even a similar top of the market Fx-6X Athlon clock. Not only is the clock low for the power output but so is the relative price (this was at the end of last year at least).

Dual core cpus have the potential to do much more work v. single don't they ? - and its the (at least potential) power we are interested in for DC -after all its the Conroe's superlative multitasking ability (and economical price ) that were the main attractions for me....

Pentium4 HT's don't do two things at once. They have two cache pipelines that allows it to load up two separate tasks into the pipelines. When 1 of the processes has to wait, bam, it starts executing the process in the second pipeline. Then back again.
It just means that the CPU has practically 100% utilisation, rather than 95% utilisation

Dual core CPU's are almost twice the speed of a similarly clocked single core CPU -> in the same series.

AMD provides a *rating* for them though, rather than providing the clock speed. An athlon64 3800 is clocked much higher than the athlon64 X2 3800. AMD expects the X2 3800 to perform similar or slightly better to a single core Athlon64 3800 but the X2 CPU has a much lower GHZ clock speed, and should therefore chew up less power

Intel doesn't suffer from the whole rating issue. They just run on straight Ghz. If an intel dual core is clocked at the same speed as a single core (in the same series), it should perform almost twice as fast.

Of course, going from the pentium4's to the Core's then to the Core2's .. you can't compare clock speeds. The Core2's are 10% for the same clock speed as the Core(1)'s which are also 10-20% faster than the dual core pentium4's.

To top it off, you can get a variety of cache sizes for different cpu's also.

Scorp is right in that you can't compare clock speeds of CPU's ... they are all completely different. The Pentium3's were left in the dust when the lower clocked athlons first arrived, and the relationship between intel & AMD stayed that way up until recently when the Core(1) CPU's were released, when intel finally released a clear winner in the duel.

The only real way to compare CPU's is the end result in a bunch of benchmarks.