Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
apHytHiaTe said:Another vote for AMD. If you really want a great folder and want space efficiency, loosen your belt, belly up, and plop down for a dual-core. If you want the production of 4 computers, get a dual-dualcore opteron rig.
In my rough guess, an X2 3800+ probably puts out about 230-300ppd. Very, Very nice.
Or for something really cheap, low end a64 (oc'd stably) or a sempron (again oc it) won't fail you.
If you're just going by what you have around... borg the intels and amd's![]()

QFT. 21 hours per WU!Helipotte said:4Ghz P4 plus QMD core = folding bliss!![]()
Tigerbiten said:All the protiens are bench-marked at 110 points per day by Stanford on an 2.8 Ghz Intel P4.
That works out ay 1.6 Points per Hour per Ghz.
So running with SSE boost you should get that.
Running without SSE boost you will run at half that speed.
How much faster you run depends on a number of things.
All AMD's fold Tinkers at twice this speed.
All big packet Gromac's fold at twice this speed.
SSE2 & SSE3 can give you a boost in speed if both your cpu and the protien can use them.
Plus some protien get benched wrong and just fold faster than they should.....![]()
So the range is anything from 1.6 PpHpG all the way upto 7+ PpHpG.
Plus it maybe cheaper to run two mobile CPU's as opposed to one high end one and still do more Points per Day.
Luck.............
![]()

I just computed my PPD on my 3700 (single core) over the last 120 days - 216 PPD. X2 would effectively double that and the 3800 might run at a higher clock speed than the 2.2GHZ of the 3700...(too lazy to look it up now)..apHytHiaTe said:In my rough guess, an X2 3800+ probably puts out about 230-300ppd. Very, Very nice.

Normal.magnusvir said:
Not necessarily. You have to have bigunits enabled in order for the -advmethods flag to work, but you can do big WUs without using -advmethods. On a dual-core, I would suggest only setting one core to big units, leaving the other on normal WUs. Big WUs tend to swamp the memory bus, and two of them will create a pretty bad bottleneck.ICE_9 said:Is that using the -advmethods flag or not?
