![]() |
|
|
#23 | |
|
Einyen
Dec 2003
Denmark
2·1,579 Posts |
Quote:
ftp://ftp.mersenne.org/gimps/p95v291.win64.zip I'm still using 28.9, so I did not realize at first that the dll files comes with 29.x versions. Last fiddled with by ATH on 2017-04-29 at 06:25 |
|
|
|
|
|
|
#24 | |
|
Sep 2003
5·11·47 Posts |
Quote:
It's good that you're also doing one, though, because the data seemed a bit puzzling (to me at least), so it will be good to have a second benchmark and see if the results are consistent and reproducible. In practice, as mentioned, the c4.large is much more relevant. I think it's worth running a separate benchmark for that too. Of course a c4.large shares a physical server with other Amazon customers, so 8 seconds might not be long enough to even out short-term effects. |
|
|
|
|
|
|
#25 |
|
Feb 2016
UK
43510 Posts |
Maybe the coffee hasn't kicked in yet, but I got results overnight and I can't see where to e-mail results?
For now, they're attached to this post. 3 results included: E5-2683 v3 ES at 2.3 GHz all cores active, 2133 quad channel, single rank ram. HT was turned off. Ryzen 1700 stock (3.2 GHz all cores active), 2666 dual channel, single rank ram. SMT on but not tested. i5-6600k @ 4.2 GHz, 3000 dual channel, 2 rank ram. No HT. More will follow as I switch from existing jobs. |
|
|
|
|
|
#26 |
|
Einyen
Dec 2003
Denmark
61268 Posts |
Amazon EC2 c4x8large
Xeon E5-2666 v3 2.90GHz resultsc4x8large.txt Haswell-E i7-5960X slightly overclocked to 3.5 Ghz. Quad channel DDR4 at 2666 Mhz: results5960X.txt Last fiddled with by ATH on 2017-04-29 at 13:25 |
|
|
|
|
|
#27 | |
|
Sep 2003
5·11·47 Posts |
Quote:
|
|
|
|
|
|
|
#28 | |
|
Einyen
Dec 2003
Denmark
2·1,579 Posts |
Quote:
0 (for Throughput) 16 32768 Y (every FFT size) N (1 core) Y (all cores) N (core combinations) N (hyperthreading) Y (1 worker) Y (maximum workers) N (worker combinations) 8 (seconds) Y (accept answers) Last fiddled with by ATH on 2017-04-29 at 16:03 |
|
|
|
|
|
|
#29 | |
|
Sep 2003
5×11×47 Posts |
Quote:
I also used the special executable George linked in this thread. You can get it on Linux using wget on the URL. Did you by any chance use the regular executable instead, or version 28? Code:
wget -O mprime "https://www.dropbox.com/s/9n3cfkumuqykbbp/mprime?dl=0" chmod +x mprime Last fiddled with by GP2 on 2017-04-29 at 17:05 |
|
|
|
|
|
|
#30 |
|
P90 years forever!
Aug 2002
Yeehaw, FL
11101011011012 Posts |
Thanks all!!! I've got a lot of data to digest over the next few days.
|
|
|
|
|
|
#31 | |
|
Einyen
Dec 2003
Denmark
2·1,579 Posts |
Quote:
|
|
|
|
|
|
|
#32 |
|
"Victor de Hollander"
Aug 2011
the Netherlands
23·3·72 Posts |
|
|
|
|
|
|
#33 |
|
"/X\(‘-‘)/X\"
Jan 2013
1011011100102 Posts |
Yes, it's special for Amazon. Intel was willing to spec it with high clocks because Amazon could guarantee appropriate cooling.
|
|
|
|
![]() |
| Thread Tools | |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| 29.2 benchmark help #2 (Ryzen only) | Prime95 | Software | 10 | 2017-05-08 13:24 |
| Benchmark Variances | Fred | Software | 5 | 2016-04-01 18:15 |
| LLR benchmark thread | Oddball | Riesel Prime Search | 5 | 2010-08-02 00:11 |
| Does anyone have i7 920? for Benchmark? | cipher | Twin Prime Search | 2 | 2009-04-14 20:16 |
| Benchmark Weirdness | R.D. Silverman | Hardware | 2 | 2007-07-25 12:16 |