Slooow city

[MerePeer]

Anyone else notice that gen zero is taking about twice as long? The new protein is actually smaller (was: 121; is: 119) too. Since the engine {foldtrajlite} does not appear to have changed since July I can only guess that this protein is so much more complex that even the random generation is harder.
:Pokes:

[GHOST]

i only run two computers now. am wondering why i bother running it on my laptop.

my athlon xp 2500 with 512mb ram does three times as much. currently the athlon is doing 100,000 structures per day on first generation according to dfgui.

my pentium 4 2.2 with 256mb is doing 33,000 per day. on first generation.

hmm. cpuz tells me it is a pentium 4. mpga-478. put mpga-478 in newegg seach and i get celeron processors, p4 BASED. suckered again by dell and pentium.

[willy1]

My Dell workstation with a *real* P4 2.0GHz is doing about 62,000 per day on gen 0.

Sounds like you did get suckered. Sorry.

[MerePeer]

Sounds like a real P4 2.0GHz compares to my 2.8Ghz celeron (512m): its 60% through gen 0 after 8 hours; structure/day hovers around 60000.

I've yet to figure out those new Intel D designations.

I did notice Intel was slashing prices. Intel slashes prices by up to 35 percent (http://biz.yahoo.com/ap/040823/intel_chip_prices_2.html)

I also wondered if (not overclocked)
an AMD Sempron 2500+ @1.75GHz for $78 was just as good as
an AMD Athlon XP 2500+ "Barton" @1.83GHz for $90, and if either compared favorably to a
trusty AMD Athlon XP 2400+ "Thoroughbred" @2.00 GHz for $79

I'll have to visit Anandtech for some reviews...
:)

[bwkaz]

But Anandtech won't benchmark DF, either... ;)

My XP2500+ Barton (@1800 MHz) is running about 40000 structures per day according to dfGUI for Linux (which may not use the same algorithm as dfrGUI for Windows -- in fact, this seems likely, since the numbers are pretty low, however the benchmark has only been running for about 20 minutes). My client is currently on gen 16, though, so that's probably the reason. (Plus, it segfaulted seemingly shortly after downloading the update yesterday. I restarted it when I got home, 5 hours later. All in all, it did gen 0 plus 15 other gens in somewhere between 5 hours and 8 hours last night -- I can't see the segfault having happened more than 3 hours after the update.)

[Thor]

@ MerePeer
The Sempron is aimed at the Celerons...

While the BArton has 512 Kb L2 cache, the Sempron has only 256KB like the Thoroughbred.

It also has less "real" Mhz than the Barton and the Thoroughbred so it will probably be slower....

I'm not shure, wether a 2400+ Thoroughbredor a 2500+ Barton will be faster at DF...
I think I remembr, that DF likes real Mhz better than a bigger cache, but I'm not quite shure....

In real life, the Barton will be faster...

Hope that helps...

[GHOST]

i think the semprons are barton chips that they perform a lobotomy on to dumb them down, then put a fancy cover on them. forcing you to buy the fx or athlon64 if you want more power. same old tricks.

[HaloJones]

Sun E-Series server 480MHz after 18 hours has only done 12000 out of the first 30000 :(

[Thor]

IBM x445 with 4x 3Ghz did about 1.25 sets in 24 hours (per cpu)


:cheers:

Thor

who found a big fat afterburner for his XP1800+ :smoking: :moon:

[magnav0x]

I suppose it's slow. I stopped running DF to return to FAD a week or two ago and I seem to still be pulling in 150k a day lol. I wonder where that box is..... :drink: