DPAD Members: Please Read. Experiment Time!![edit]answer found?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Kokomo

Member
Jan 6, 2004
178
0
0
Originally posted by: Overkiller
Are your 2.8's @ 3.2? Also, i gamed (intensively) for about 4.5 hrs total yesterday so that equates into some down time...
Good for you. Guess I missed the fact this was a gaming topic.

 

Orange Kid

Elite Member
Oct 9, 1999
4,440
2,216
146
Does:

.0065 * CPU speed (in mhz) = Mpts/min
work for anyone else?

Maybe a different way to determine a way to set a standard, but as the CLI gives info as Mpts, thought this might be a way.

I figured this with only a hours worth of client data, so the validity of this info may be suspect.
Also, this is my first running of the client, so I may be a bit suspect :)


EDIT was to change formula slightly and clarify
 

DanC

Diamond Member
Jun 2, 2000
5,553
0
0
The only thing that's going to be just a little bit tough is for those with remote systems - you can never be ENTIRELY sure that what you just dumped is ALL you dumped from a particular machine.

I'll have to ponder that. -- actually - I went to double-check MuonMonitor - and it does tell me how many particles. The only problem I guess is that I can't cut & paste. I do see that you can create a comma delimited file with it though - and plug into Excel. I'll futz with it. ;)
 

Overkiller

Platinum Member
Feb 22, 2003
2,461
0
0
haha orange kid. i like your post :p

Everyone try Out Orange Kid's Equation!!

For my AMD system there is a variation of under 1000 Mpts (over an entire day) which is normal b/c the pc IS used and the equation is probably a *highest yield*
For my Intel System it is 6000 Mpts off (which makes sense b/c the pc was used extensively for gaming fora few hrs).

I'll use this equation tomorrow to check my results and if this has a low standard deviation then we have found a *rough estimate* for our machines..now we just have to Compare a P4 @ 3.2 ghz vs. a Barton 2500+ @ 2.2 ghz (i'll be able to test this one soon ;) )

Now something just hit me....there are no client optimizations that i know of...

So a Barton 2.2 ghz system should put out 20K Mpts/day *if* it ran @ stock speed. but with the system being overclocked then it should put out a few thousand more mpts/day [just think, the 1700+ @ 2.1 ghz does 20K mpts/day...with the barton having the extra cache & 100mhz advantage i forsee it actually doing 23-24K mpts/day]


While a P4 @ 3.2 puts out ~23-29 Mpts/day...

While i am not sure where i'm heading with this(i'm a tad bit tired :D) i have a feeling that i will need non-overclocked systems to compare with overclocked ones.

So if anyone has non-overclocked systems (p3's, early p4's, Tbirds, non-overclocked Xp's/bartons, newer p4's) please specify if they are running at stock speed and give me the mpts/day...also use the equation to see if it substantiates your results..

For those that have overclcoked machines...just let me know what they *should* be running at and what they *do* run at :)
[i think a comparision of say a 1700+ @ 1.8 ghz to a stock barton 2500+ could be interesting]

-ok , enough rambling...send me your results :) :beer:
 

r0tt3n1

Golden Member
Oct 16, 2001
1,086
0
0
From my Duron 850, for the past 2 days:
file size 37k, 7750Mpts
file size 36k, 7566Mpts
 

nickPOWERZ

Member
Jun 7, 2003
54
0
0
Athlon 2400+ stock, dual 256mb Crucial DDR 266 @ 2.5-?

24hrs = 19463 Mpts or 79KB or 75 results

This is my second run.
 

Orange Kid

Elite Member
Oct 9, 1999
4,440
2,216
146
My equation seems to be comming in at + or - 1000Mpts for a 24hr. run, except on that dual of rybobcat

Must refine it, and more info on that dual would help.

One thread or two running, normal usage?
 

OverVolt

Lifer
Aug 31, 2002
14,278
89
91
Originally posted by: Orange Kid
My equation seems to be comming in at + or - 1000Mpts for a 24hr. run, except on that dual of rybobcat

Must refine it, and more info on that dual would help.

One thread or two running, normal usage?
Or the two CPU's could have been fighitng for memory bandwidth.

 

rybobcat

Member
Sep 12, 2003
63
0
0
Originally posted by: Orange Kid
My equation seems to be comming in at + or - 1000Mpts for a 24hr. run, except on that dual of rybobcat

Must refine it, and more info on that dual would help.

One thread or two running, normal usage?

Its running 2 threads, Normal.
 

Overkiller

Platinum Member
Feb 22, 2003
2,461
0
0
Originally posted by: Orange Kid
My equation seems to be comming in at + or - 1000Mpts for a 24hr. run, except on that dual of rybobcat

Must refine it, and more info on that dual would help.

One thread or two running, normal usage?

Is this dependent or independent on AMD vs. Intel ?
-B/c while there are no software optimizations for either architecture one has AMD's higher efficiency/MHZ compared to Intels wicked high MHZ...
 

OverVolt

Lifer
Aug 31, 2002
14,278
89
91
Originally posted by: Overkiller
Originally posted by: Orange Kid
My equation seems to be comming in at + or - 1000Mpts for a 24hr. run, except on that dual of rybobcat

Must refine it, and more info on that dual would help.

One thread or two running, normal usage?

Is this dependent or independent on AMD vs. Intel ?
-B/c while there are no software optimizations for either architecture one has AMD's higher efficiency/MHZ compared to Intels wicked high MHZ...
i would say it depends if its an intel/amd my 1.7Ghz Athlon is almost on par with my 3.0C :Q

 

Overkiller

Platinum Member
Feb 22, 2003
2,461
0
0
I really think that the client really scales linearly with cpu MHZ, be it intel or AMD.
Take the highest AMD system on this board and compare it to the highest INtel system ( my or kokomo's 3.2's [single systems not dual]) and the Intel is about ~7-10K Mpts above an AMD system at its highest speed..this is directly related to the mhz difference between a 2.2 ghz athlon and a 3.2 ghz Pentium...

Would u guys agree that this is what is coming forth?

As always...i thank you guys for going along w/ this...Keep it up.. ;)
 

BlackMountainCow

Diamond Member
May 28, 2003
5,759
0
0
Going to give u my numbers for XP 2000+ Pal and XP 2500+ @ 3200+ Barton by 22.00 o'clock GMT +1 (German time)
 

Overkiller

Platinum Member
Feb 22, 2003
2,461
0
0
[edit] dregged up this old thread...will get to putting these on my machines @ home shortly..can you guys do this as well and then doing this comparision will be MUCH MUCH MUCH easier :D :) ;)

linky

Be sure to read to the end..he has an updated link for the benchmaking client...i can't wait to get home..tehehehe...this guide is gonna be cake now :p
 

Kokomo

Member
Jan 6, 2004
178
0
0
Hope these results are helpful.

350 kpts/sec - Dual Xeon 2.4/Supermicro P4DC6/2x256 PC1066
230 kpts/sec - P4P 2.8/GigaByte 8IK1100/2x256 Corsair XMS LL
250 kpts/sec - P4C 2.8/Intel D875PBZLK/2x256 Kingston Value
250 kpts/sec - P4C 2.8/Intel D875PBZLK/2x256 Mushkin Level 1

Kpts/sec were averaged over several hours. Poor DPAD performance of P4P (Prescott) is disappointing, but it runs with the Xeon on SETI (20 WUs a day). All settings are stock.

Variations in my posted daily Mpts are the result of my experiments with memory, mobos, etc., and sticking an AMD 2400 in the mix from time to time.

Good job, Overkiller. Looking forward the the guide. :beer::)
 

BlackMountainCow

Diamond Member
May 28, 2003
5,759
0
0
24h test:

AMD Athlon XP 2000+ (all at stock) with 512 MB 266FSB RAM

- 171,00 kpts/sec
- 56 results + 2 quarantined runs --> 66 results
- best yield: 9,866
- Mpts: ca. 14800


AMD Athlon XP 2500+ @ 3200+ (just 200 FSB) with 256MB 400FSB RAM

- 252,51 kpts/sec
- 105 results + 1 quarantined run --> 115
- best yield: 9,863
- Mpts: ca. 22300


Hope this helps u a bit Overkiller. Gonna give u some stats for a PIII 600 tomorrow. And I think it's intereting that my 3200+ is even with Kokomo's 2.8 P4. So on the one hand we got 2.2 GHz versus 2.8 GHz. BUt then we got a 3200+ rating (I know it's not really equal) versus a "2800" rating.
 

Orange Kid

Elite Member
Oct 9, 1999
4,440
2,216
146
last 24 hrs gave me 16321.4 Mpts, forgot about the file size thinggy though:eek:

This is on a

Barton 2500 underclocked @1.7 ghz -- Yes I need a new MOBO:eek:

Never enough hours in a day :(

Will try running the new proggy ASAP

 

Stormgiant

Senior member
Oct 25, 1999
829
0
0
I'm averaging about 315kpts...
with the new memory that should arrive next week, i could get well near 350...

P4 2600@3500 with 1GBDDR400
 

BlackMountainCow

Diamond Member
May 28, 2003
5,759
0
0
24h:

Pentium III 600
256 MB RAM
Results: 12 (one recheck run with 7)
MPts: 3644


Hey, I know it's not much but in a week that's 25.000 points :D