Originally Posted by Rodzilla
Think you've become accustomed to my temporary negative exterior Rod.
As far as accuracy... I'm with you. It's the actual engine running actual work... can't get much more accurate than that.
No... just dont reed dat good sometime... lol.
yeah... pooped, popped.. what's the difference
Come up here and say that!!yeah... pooped, popped.. what's the difference
don't worry.. i'm cominCome up here and say that!!
Ok... I ran another test on my Sempron rig running @ 1.9GHz and 2-2-2-5 on the mem.
@ 2-2-2-5 it gets 222sec
@ 2.5-3-3-6 it gets 223sec...
So in short I don't think mem timings mean shat to D2OL.
Last edited by Rodzilla; 03-16-2005 at 01:38 PM.
The only thing I'd recommend to people with farms is that they setup their systems with decent timings anyways. You never know when D2OL will close shop and it sucks having to reconfigure all your rigs when you change projects.
But as always, this is just my 2¢ so take it for what it's worth.
LOL, you love it and you know it!If D2OL found the cures for all it's targets and a new project came, you'd retweak for the new one. It's part of the sickness.
DeXter::2xXeon2.4@3200, GF6800, PC-DL, 1gb Centon, IWT+panaflo air, Win7
Quad::Q6600(g0)@3400(1.30V), GTX275, 965P-DS3, 4gb HyperX, Asus SilentSquare (air), Win7_64U
NuQuad::Q9400@3300(1.24V), 8800GT, G31M-ES2L, 2gb Patriot, stock air (bigger Q6600 one), WinXP64
Crunch::E4300@2400, onboard, GF7100PVT-M3, 2gb Patriot CL5, stock air, Vista64U
what chipset does the board you're using have?
VIA KT600
139s with the rig on my sig, 235x11 ( 2.58 ghz ) and mem @ 2.5-3-3-7 1T . I set up my rig this way since it's crunching 24/7 and NEVER shut it down ...
2x Harpers 5492, 8GB DDR2-800 FBDIMM, SuperMicro X7DWA-N, Raptors in Raid0, 1TB Storage, Dual Watercooling loop, U2UFO, Enermax 1000W
Some nice LianLi case, i7 965 Extreme, FoxConn BloodRage, 3x2GB OCZ Blade 2000, Zotac GTX260, Enermax 700W
if mem timings don't make much of a difference can someone try with 2gb ram?
i7 980x | corsair dominator 6x4 Gb | TJ07
SLI GTX 580x2 | GT430 | rampage 3 | corsair AX1200
LSI 9265 | OCZ vertex 128x4 raid0 | corsair c300 | hitachi 3tbx3 |
blocks: cpu hk 3.0 | gpu aquacomputer 580x2 | mb ek re3 fc | ram bitspower galaxy
pump: laing d5 x2 | res: koolance 452x2
rad: TC 120.3 120.2 | scythe GTx5
NASs: qnap 879 8x3tb | thecus N5200 2tbx5
Upped my FSB and ran it again. Just saw that you cant really see anything with my attachment. I got 145.
Last edited by chunkylover77; 03-17-2005 at 03:37 PM.
RIP Dad. I will miss you always.
Q6600,DFI X48 T2R,280 GTX,4gig ReaperX,XiFi Prelude,LL TYR PC X-2000,Westinghouse 37,Klipsch 5.1 Ultra
Little test I did...
This is on a XP-M 2600+ @ ~2.5Ghz with 1GB of Ram... timings on both tests were 2.5-3-3-11
218 x 11.5
118 x 21
159 seconds with only one instance on my HT capable 4.42EE (Gallatin)
just ran it on another box... its a AXP @ 2ghz with 256mb of BH5 @ 2-2-2-11... got 209 sec out of it..
Asus P5K<*>OCZ 450w<*>E6600@4.0ghz@9x445@1.450v<*>TeamXtreem DDR2-800 cl3 2x1gb<*>Sapphire 3870 512mb DDR4
Asus P5E X48 DeLuxe<*>OCZ 700w<*>E8600 E0@4.6ghz@1.350v wc<*>Corsair Titanium DDR2-800 4x2gb<*>Sapphire 2x3870 2x512mb DDR4 CF
145... Nice!
More experimenting with the worthlessness of FSB to D2OL..
100 x 22
220 x 10
XP-M 2600+ @ 2.6 (200x13)
What was your mem speed in each of those test?????????
Coupla questions:
a) Sticky?
b) Anyone confirmed how this behaves in a multi-threaded/processor invironment? I.E. there a way to run 2 or 4 instances at once and combine the result? (or I suppose assuming 100% efficiency you could divide by 4, but let's stick to the real world here
c) Does the bench account for the random behavior inherent in D2OL task assignment i.e. how some tasks/targets take longer than others?
Just wondering . . .
P4 3.06 @ 3.2 = 423s
Celeron 1Ghz = 524s
PIII 550Mhz = 922s (better than nothing I guess)
Maybe its time to get an A64 to replace the P4... Hmmm.
The reason it might run better is because your not stresing the me at that high of an FSB anymore. Nice results btw.
I would be interested in knowing if memory bandwith helps in D2OL. So can someone do some bench single channel VS dual channel?
Posting Permissions
Bookmarks