x264 HD BENCHMARK 3.0 RESULTS
Please do not compare it with older versions of the benchmark!
Please copy/paste everything below the line to to report your data
Results for x264.exe r1342
encoded 1442 frames, 113.37 fps, 3900.68 kb/s
encoded 1442 frames, 100.85 fps, 3900.68 kb/s
encoded 1442 frames, 115.35 fps, 3900.68 kb/s
encoded 1442 frames, 115.49 fps, 3900.97 kb/s
encoded 1442 frames, 30.98 fps, 3971.45 kb/s
encoded 1442 frames, 31.07 fps, 3972.09 kb/s
encoded 1442 frames, 31.26 fps, 3971.17 kb/s
encoded 1442 frames, 31.37 fps, 3971.52 kb/s
Name Intel Processor
Specification Intel(R) Core(TM) i5-3570K CPU @ 3.40GHz
Core Speed 3397.1 MHz
Northbridge Intel ID0150 rev. 09
Southbridge Intel ID1E44 rev. 04
Memory Size 8080 MBytes
Windows Version (Build 9200)
Number of processors 1
Number of threads 4
Number of threads 4 (max 16)
L2 cache 4 x 256 KBytes, 8-way set associative, 64-byte line size
Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, EM64T
Temperature 0 33°C (91°F) [0x21] (SYSTIN)
Temperature 1 46°C (113°F) [0x5B] (CPUTIN)
Temperature 2 36°C (95°F) [0x47] (AUXTIN)
31fps is really disappointing actually, because even my laptop managed to get 26fps (well.... before it started overheating and thermal scaling kicked in). I mean this shit is 4.2Ghz, the laptop is 2GHz. I guess hyper-threading on the laptop helped out a lot.
Well, can't really be helped, since I'm not going to buy an i7 3770 (= pay $150 more for hyper-threading), and the i7 hexacore CPUs are out of my price range. Bulldozer is actually by far the best price for value for multi-threaded stuff like encoding, but its single core performance is like worse than Phenom II, and I need my single core performance to emulate PS2 and shit.
Anyway, CPU sort of runs hot (up to 80c) on the prime95 blend stress test. Sort of my fault, since I totally derped up on the cooler choice - I mean I chose a CPU cooler that intakes air from the side, when I have a solid piece of glass on that side with no vent or fan whatsoever (yeah wasn't really thinking back then), so it's pretty much recycling the hot air from inside the case, and a noise reducing case with bad airflow doesn't help the situation. Doesn't really matter though, since the most intensive thing I'm ever going to do with it is x264 on 100% CPU, and that only ever hits 70c. 100% CPU on 7-zip only hits 60c. And contrary to popular misconception, games are a joke at burning the CPU, and should never really even go beyond 50c, because of low CPU utilization. Note the tests are done on a hot summer day, temps should be better in winter.
Using new x264. r2245, not the newest stuff out there, but still significantly faster than the old one.
RE: [Updated!] EP Computer Benchmarking PowerBoard
(04/03/2013 04:46 PM)ZiNgA BuRgA Wrote: Interestingly, your chip at 4.2GHz (31.17fps avg) seems to perform noticeably worse than mine at 4GHz (36.30fps avg) - around 15% slower.
That's with the old x264. With the new one I do about 43fps, still can't really be compared since the x264 I'm using in this case is probably newer than yours. I'm using JEEB's patched builds with fade-compensate http://x264.fushizen.eu/.
EDIT: Or maybe I posted the results for non-overclocked 3.4GHz instead. I'll check when I get back home.
(04/03/2013 04:46 PM)ZiNgA BuRgA Wrote: I heard that BD had better single core performance over P2, but worse multi core.
Nah, I'm pretty sure I remember it's worse on single and better on multi. Well, bulldozer has 4 modules = more or less 8 cores, PII has 6 or less cores. If bulldozer does better on single thread, it would logically follow that it also does better on multiple threads since it has more cores. Even if it was at or slightly better than PII level, that's still not really good enough, I want like a lot more than PII level.
(This post was last modified: 04/03/2013 08:21 PM by Assassinator.)