Page 1 of 1
Anyone Folding With One Of These
Posted: Sat Jul 23, 2005 5:13 pm
I want one bad. I always wanted a dually and here is another option.
http://www.newegg.com/Product/Product.a ... 6819103544
If anyone has, or has heard of someone folding with one of these, see if you can find out what their PPD and PPW are.
I noticed that the voltage is 1.35-1.4 unlike the FX cores and more like the venice and san diego cores, so I am guessing that it would be easier to cool.
Still wondering what the wattage draw is. Anyone know
It is out of stock, so it must be selling.
I might have to wait for it to come down in price a bit too
EDIT: I found this thread http://forums.silentpcreview.com/viewtopic.php?t=23583
on a 4600, interesting
, Now I need to find out how it folds. Maybe I'll PM and see if he can "test fold" a few WU's
Posted: Sun Jul 24, 2005 12:37 pm
i got the X2 4600+ a few days ago and it's folding right now .. although still one one core so far.
Posted: Sun Jul 24, 2005 1:15 pm
My first thought is "isn't two separate processors cheaper?" I'm not in the loop as far as dual-core is concerned but... ;p
Posted: Tue Jul 26, 2005 8:54 am
I don't see any reason why it would be just as good as two 4000+ (2.4GHz 1MB). You might see a slight performance hit because the processors have to share the same busses, but I doubt it would hurt much.
Dual 4000+ would probably be cheaper than a 4800+, but unfortunately, there aren't any dual Athlon64 boards available, and there probably won't ever be. Your only option for a K8 dualie would be Opteron, which would surely be more expensive, because the processors themselves cost more (even though there's really no difference, you're paying for "reliability" of a server chip), and you'd need to purchase registered DDR memory to use with it.
Posted: Tue Jul 26, 2005 9:01 am
part of the appeal of dual core processors is because of how you can get a dual processor machine for a much lower price.
normally for a dual processor setup you would have to get two-way capable opterons, plus a dual socket board (which is usually at least $350 for an entry-level one), plus registered memory (which is at least 50% more expensive, sometimes 2x as much).
With a dual core chip, you can use consumer motherboards and RAM to save money.
Posted: Thu Aug 04, 2005 5:42 pm
tomcat wrote:i got the X2 4600+ a few days ago and it's folding right now .. although still one one core so far.
not sure exactly what you're saying - were you expecting it to use both by itself or something, and it's only using one? try running two copies. would be prudent to duplicate the folder first, basically follow whatever directions they give for dual CPU systems. not sure how Windows handles it but Linux used to just show two processes both using 100% CPU if you were doing something like folding on a dual CPU machine.
Posted: Fri Aug 12, 2005 1:59 pm
So has anyone actually done folding on an x2 processor yet and benchmarked the performance increase versus one instance? Like in terms of average points per day for one instance versus two. I know for hyperthreading Intel processors, most folks report about a 15% increase by running two instances. Is the x2 a 40% increase by running two instances instead of one? 80%? 95%?
Posted: Fri Aug 12, 2005 5:53 pm
running an x2 will result in a 100% increase over a single core of the same speed.
unlike hyperthreading which splits threads from a single core, dual core processors are truly two seperate processors. Since FAH is completely processor dependent, putting in the second core doubles performance.
Posted: Fri Aug 12, 2005 6:13 pm
dualcore is slightly faster than two separate CPUs, because some components are shared.
Posted: Fri Aug 12, 2005 7:52 pm
cotdt wrote:dualcore is slightly faster than two separate CPUs, because some components are shared.
hehe and also slightly slower because some components are shared. Depends entirely on workload.
Posted: Sat Aug 13, 2005 10:23 am
I was looking more for someone with real-world test results, not theory. I already know that in theory it's twice as fast, but I've still never seen a multi-threaded benchmark test show more than about a 40% increase. That would imply that adding a second instance of FAH may only result in a 40% increase in PPD, but again, I don't really care about theory, I was hoping someone, somewhere had actually tested it in the real world.
Posted: Sat Aug 13, 2005 3:38 pm
Unlike FAH, other applications are limited by things such as memory bandwith, graphics performance, and disk performance. FAH uses files less than 5MB in size and has no graphics (unless using the GUI version). The only thing that determines performance is CPU speed. Therefore, if you have twice the processors, you have twice the speed.
Posted: Sun Aug 14, 2005 3:43 am
here is some real-world data from my system (X2 4400+)
cpu0: p1149_RIBO_semihelixfrom1140: PPD 253, frame time: 34min 8s
cpu1: p1140_RIBO_FSpeptide_EXT_nospring: PPD 255, frame time: 33min 48s
Posted: Tue Aug 16, 2005 12:25 am
Those are some pretty good #s Markus
Just wondering, if you have a kill-a-watt or similar type meter, have you checked the power consumption?
Does it run as cool as rumored?
Posted: Tue Aug 16, 2005 1:44 pm
i don't have a watt meter ... i don't want to know anyway, my power bill is high enough already
it runs pretty cool, vcore is at 1.280v according to cpu-z. the system (cpu, chipset and graphics card) is water cooled with a single 120mm radiator.
cpu temp: 47C
water temp: 35C
case temp: 30C
this with two cores folding 24/7
not bad i think
Posted: Fri Aug 19, 2005 10:10 pm
Just built a AND64X2 4400+ and just joined folding! Seems to be one processor at about 50% according to task manager in XP. Dave
Posted: Sun Aug 21, 2005 5:55 am
Posted: Sun Aug 21, 2005 7:57 am
WOW, thanks! Nice welcome! Dave