using Seti as a benchmark [Archive]

tking

10-28-2002, 03:07 PM

I was reading maximumPC today and they had an article on the new AMD processors.

In the comparison section they run SETI as one of their benchmarks.

The new 2800+ Athlon did a work unit in 154 minutes! That's two hours and thirty four minutes! Holy cow!

They also ran it on an overclocked (@3.05Ghz) P4 2.8. It crunched seti in 149 minutes (2 hours and 29 minutes)!

I can't wait for our office upgrade next year!

Logan[TeamX]

10-28-2002, 04:04 PM

I have to say it... it takes an O/Ced P4 only 5 minutes less to do what a stock AMD XP 2800+ does for the WU. That's kinda sad, and proof that the GHz numbers don't always mean the winner is the one with the higher one.

I salute the Intel fans for their dedication to their CPU of choice, but something has to be said for getting it done at a lower clock frequency!

RampageIII

10-28-2002, 07:16 PM

And half the price!! :p

smithdw

10-28-2002, 08:22 PM

That's why I went with AMD on my system at home. Lots less $$$ and more performance, especially in the FPU. This one is a 1700+ Athlon XP, the last one was a 900 Athlon Tbird.

tking

10-29-2002, 07:53 AM

The ddr400, xp2800 with a 266mhz fsb was a sweet looking piece of kit. Evidently it has more overclocking potential than previous XPs and the price difference between an equivalent Pentium gives you enough for a nice mobo *and* 3 western digital 8mb cache drives for RAID... and this is still Athlon technology!

I hope the Hammer lives up to the hype.

Scavenger7

10-29-2002, 09:46 AM

I have a p4 2.4 ghz and it takes the same amount of time to do a wu as my amd xp 2200+[1.8ghz]. 3hrs

vibe666

11-01-2002, 07:20 AM

i would say that seti@home makes a good benchmark as it uses all the CPU power all the time.

gibsinep

11-05-2002, 02:16 PM

Yea, it is a great benchmark. Especially to test out your board and your RAM. If you have a friedn with the same CPu but different board/or ram then it is great to see which cmputer is faster.

vibe666

11-06-2002, 08:52 AM

and a good excuse you you to install S@H on a friends PC to boost your numbers eh, Gib ;)

vibe666

11-06-2002, 08:54 AM

BTW, you might have noticed consistently higher numbers by me recently. I thought it was down to making sure all the PC's at work were all turned on first thing rather than being off until people trickled in during the day and switched them on, but I'm pretty sure now that its down to me installing S@H in my local internet café the weekend before last.

Might have to pop in there at the weekend and do a couple more to see if it really works.:D

gibsinep

11-06-2002, 09:45 AM

The cafe lol oh wow

j.m@talk

11-10-2002, 02:48 PM

My seti scores are all over the place :(

Power cut one day, Mum with the Hoover the next :( :( :(

Me doing a complete rebuild yesterday :mad:

Need a seti turbo :)

tking

11-11-2002, 07:39 AM

Originally posted by j.m@talk21.com
Need a seti turbo :)

Here's the seti turbo you're looking for:

http://news.bbc.co.uk/olmedia/1870000/images/_1872410_cool300.jpg

With a new 2.8 in there you should be running at a cool, stable 3.65 ghz. That'll do a wu in about an hour.:x

Scavenger7

11-11-2002, 09:10 AM

VapoChill.. Brrr.. http://jeeptalk.org/crack/smilies/otn/other/loki3.gif

gibsinep

11-11-2002, 09:50 AM

lol,

I wish I could do a WU in 1hour. :(

Fastest I have done a WU is 2hr and 55min.

Xp2100+

nh3com

11-11-2002, 09:50 AM

dont know where guys got your cpu's from but in the uk there is very little between intel and amd high end cpu prices 2600xp £216 intel 2.67 £230. i would go for the intel as it will overclock big time. plus in quake 3 engine games it kicks amd's ****

tking

11-11-2002, 09:53 AM

At full CPU output the core is running at -18 degrees. How's that for cooling? The capability of your cpu is entirely based on internal architecture with this thing, cooling is taken care of absolutely.

24 wu's a day on one machine... cool:r

(edit) ... oh, and it's totally quiet:cool:

gibsinep

11-11-2002, 10:03 AM

24WU a day on one machine is crazy. :x

Scavenger7

11-11-2002, 10:24 AM

I guess if your dedicated enough the $460 and up price tag for the case alone should not be an issue. :x

Just seems like alot of money for seti! [famous last words from a future owner] :p

j.m@talk

11-11-2002, 05:15 PM

In the words of the "Great Bovon" Dang' Tootin'

gibsinep

11-11-2002, 05:17 PM

lol I agree:eek:

j.m@talk

11-11-2002, 05:29 PM

Sinep & his Lol's (Dude Should be circumnavigated or what ever):eek:

gibsinep

11-11-2002, 06:34 PM

lol, laughing? whats wrong with that?

Are you saying you want me circumsized?:confused:

vibe666

11-12-2002, 07:13 AM

I think he might be talking about just going around you (circumnavigating) but I could be wrong :eek:

Scavenger7

11-12-2002, 09:55 AM

Bah Humbug http://home.iprimus.com.au/morpheus/smilies/****.gif

vs

LOL http://www.suburbia.com.au/~morpheus/smilies/nutkick.gif

Logan[TeamX]

11-12-2002, 10:26 AM

You guys are insane... not that I'm considering building a few XP 2400+ machines in the next few months just for distributed computing and game server uses :rolleyes:

Oh who am I kidding, I'll build a 2800+ system in the spring, give the O/Ced 1800+ to the g/f, and sink the rest of the cash into the truck and our vacation :D :r

j.m@talk

11-12-2002, 02:08 PM

Originally posted by gibsinep
lol, laughing? whats wrong with that?

Are you saying you want me circumsized?:confused:

Might as well castrate @ same time :eek:

gibsinep

11-12-2002, 02:16 PM

Pervert :rolleyes:

j/k :p

j.m@talk

11-12-2002, 02:27 PM

Originally posted by gibsinep
Pervert :rolleyes:

j/k :p

With Tigers :)

eFrisky

11-12-2002, 03:26 PM

Correct me if I'm wrong, or if I missed the point...

You'll need the standard "benchmark workunit" to compare different systems.
You have to run exactly the same workunit, obviously, to compare systems :)

gibsinep

11-12-2002, 04:12 PM

This is true efriskey, but you can get a 'general' idea even though Wu are different.

AceAirman

11-13-2002, 11:27 PM

Yeah eFrisky, if you run like 15 WU's on two computers you have a pretty good idea how they compare in times processing WU's. Like Gib said you get a general idea. And the more you do on each of the two systems the better that general idea gets.

j.m@talk

11-14-2002, 04:38 AM

I think the term "Benchmark" is throwing a spanner in the works!
It's a tad precise,

Perhaps Using seti as a General System Performance Indicator
might be a more digestable :p

"GSPI" :confused:

Bit more garbage for you guys to ponder ;) :p

gibsinep

11-14-2002, 05:30 PM

Like the "General" says....

I agree :p

hoodedrat

11-20-2002, 10:53 AM

how do you tell how fast you're doing a wu ?

j.m@talk

11-20-2002, 12:43 PM

You could use the timer :p

gibsinep

11-20-2002, 04:25 PM

hoodedrat- most of us use setihide, which keeps a log of how long each WU takes.

j.m@talk

11-20-2002, 04:31 PM

Originally posted by gibsinep
hoodedrat- most of us use setihide, which keeps a log of how long each WU takes.

As per exhibit "A"

gibsinep

11-20-2002, 04:55 PM

j.m@talk21.com- so who is Sarah :p :eek:

j.m@talk

11-20-2002, 05:06 PM

Could tell you but then I'd have to kill you :D

So we'll cut out the un-necassary conversation, Say Bye Bye Gibby Baby :mad:

gibsinep

11-20-2002, 06:01 PM

lol bye??

smithdw

11-20-2002, 06:07 PM

Memo to self - When posting a screen capture to these guys, only post the window, not the entire screen....

Can't get anything by these guys. :D