You Can Upload Photos Of Your Results Guys!
First, see subject-line/title (might as well put up the proofs etc., because this forums has that neat upload image feature):
I never ran this before, but those are my results!
* What I gathered upon cursory inspection, is it took my system 36 seconds to pull that test off...
APK
P.S.=>
Azn Tr14dZ said:
Wow, my Pentium D 820 is getting owned by (almost) every processor here!
AMD cpu's show the SAME basic type of results vs. Intel cpu's on SETI@Home's built-in benchmark...
Maybe not the BEST comparison, but still, you could compare my results on my Intel rig & AMD rig here:
http://setiathome.berkeley.edu/hosts_user.php?userid=8294882&show_all=1&sort=rpc_time
APK AMD 4800+ SETI benchmark results
Measured floating point speed 4226.67 million ops/sec
Measured integer speed 13743.99 million ops/sec
APK Intel P4 3.2ghz SETI benchmark results
Measured floating point speed 2087.77 million ops/sec
Measured integer speed 2315.65 million ops/sec
& see how superior AMD cpu's are in the floating-point operations area - this even helps games & is certainly evidenced by AMD's performance in THAT arena as of late/past few years now.
* IMPORTANT TO NOTE - Oddly though, even though the data is more fpop centric & amd rules there? Intel CPU's do better on SETI imo - the client wares are optimized better for Intel!
Azn Tr14dZ said:
oh, it was on. I'ma run it again. But I guess SuperPI is a better for single cores. So basically, I'm getting the same scores as a 2.8GHz Pentium 4 then.
It's a single-threaded app... Systems like yours & mine (assuming yours is dual core (or anyone w/ a dual core/smp cpu)) won't show what they can REALLY do here testing w/ this app, I agree.
Still - imo, it's a good test of the ALU's Floating Point power (@ least on 1 of the dual cores possible on these types of CPU's)!
See, & DO correct me here if I am wrong "hardware-centric" folks (Because, I am not 110% sure of AMD cpu internals architecture here):
I assume there are dual Fpop Units & dual ALU (arithmetic logic unit(s)), & this would only be using 1 of 2 possible in these beasts (assuming that they are indeed FULL cpu's in the dual core architecture of your choice Intel vs AMD, & that there is not a single "shared" unit between cores for this function, which imo, a Hyper-Threaded CPU would use/do for instance)! apk
. Anyways...I got about 35 seconds for a 1m digit run. See specs to the left. I'll have to run it again, though I doubt SATA and NTFS will make any difference lol.
.
