x264 CPU benchmark - Compare different CPUs encoding the same file

[Soopafresh]

Avisynth 64, DGDecode64, AND x264-64!

http://members.optusnet.com.au/squid_80/

[jagabo]

@jag - thanks for the data set. Can you report the multiplier and FSB, mem timings and settings, and northbridge chipset also?

Added cpuz.htm to previous post.

[graysky]

thanks for the 64bit links, soopa.

Thanks for the data, jag. This is the most extreme example I have seen (about 16 % slower on vista).

[Webster]

Here is the freshly installed Vista business 32 bits run on the same machine (see spec below):


---------- RUN1PASS1.LOG
encoded 1749 frames, 42.29 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 38.93 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 38.66 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 39.37 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 40.20 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 10.28 fps, 1826.37 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 10.26 fps, 1826.37 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 10.59 fps, 1826.37 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 10.56 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 10.56 fps, 1826.37 kb/s

running Windows XP Professional x64 Edition

---------- RUN1PASS1.LOG
encoded 1749 frames, 42.95 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 43.04 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 43.42 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 43.29 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 43.32 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 11.03 fps, 1826.37 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 10.97 fps, 1826.20 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 11.09 fps, 1826.37 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 11.08 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 11.07 fps, 1826.37 kb/s

[graysky]

thanks man!

[redwudz]

Here's a comparison of Vista Home Premium VS XP Pro SP2. I did a fresh install of XP on a second HD. No other programs. The Vista setup has quite a few added programs, but none were running at the time of the test. UAC was operating.

Vista Home Premium
AMD Brisbane 64 X2 4800+ (overclocked to 2.8Ghz )
Multiplier 12.5 Buss speed 224Mhz HT Link 1120 MHz
2GB RAM Corsair XMS DDR2-6400 (800Mhz)
Dual channel 4,4,4,12,23 timings
Gigabyte GA-MA69G-S3H MB
Nbridge ATI RS690/RS690M
Sbridge ATI SB600

---------- RUN1PASS1.LOG
encoded 1749 frames, 72.66 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 72.91 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 72.02 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 72.56 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 73.06 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 17.84 fps, 1826.37 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 17.60 fps, 1826.37 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 17.83 fps, 1826.22 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 17.79 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 17.79 fps, 1826.37 kb/s

XP Pro SP2 32
AMD Brisbane 64 X2 4800+ (overclocked to 2.8Ghz )
Multiplier 12.5 Buss speed 224Mhz HT Link 1120 MHz
2GB RAM Corsair XMS DDR2-6400 (800Mhz)
Dual channel 4,4,4,12,23 timings
Gigabyte GA-MA69G-S3H MB
Nbridge ATI RS690/RS690M
Sbridge ATI SB600


---------- RUN1PASS1.LOG
encoded 1749 frames, 75.63 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 75.78 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 75.53 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 75.73 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 75.58 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 18.13 fps, 1826.38 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 18.14 fps, 1826.37 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 18.12 fps, 1826.38 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 18.15 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 18.10 fps, 1826.38 kb/s

Not really a lot of difference. The CPU will be upgraded to a Athlon 64 Windsor 5600+ with the same settings.

[graysky]

Thanks for the results... 2.2 % slower on Vista. Can you verify that the memory is actually clocked @ 800 MHz? Running @ 12.5 x 224 and running the mem @ 400 (actually 800 since it's DDR2) is a ratio of 3:5 which I dunno if your board supports.

Thanks!

[redwudz]

That (800Mhz) was just part of the memory name. I changed the CPU and overwrote the CPU-Z file, but it should have been 424Mhz, I think. I have it running at 432Mhz with a 232Mhz Buss speed at present. I don't have the drive with XP in it any more. It idles at 34C and when encoding goes up to about 42C so I'm happy with that.

This is the log from the 5600+ Athlon:

Vista Home Premium
AMD Athlon (Windsor) 64 X2 5600+ (overclocked to 3.02Ghz from 2.8Mhz )
Multiplier 13 Buss speed 232Mhz HT Link 1163 MHz
2GB RAM Corsair XMS DDR2-6400 (800Mhz)
Dual channel 4,4,4,12,22 timings Frequency 432Mhz
Gigabyte GA-MA69G-S3H MB
Nbridge ATI RS690/RS690M
Sbridge ATI SB600

---------- RUN1PASS1.LOG
encoded 1749 frames, 80.52 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 81.28 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 81.48 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 81.47 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 81.58 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 19.92 fps, 1826.38 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 19.92 fps, 1826.37 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 19.89 fps, 1826.33 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 19.92 fps, 1826.38 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 19.90 fps, 1826.38 kb/s

[träskmannen]

This is my main computer (AMD X2 5600).

---------- RUN1PASS1.LOG
encoded 1749 frames, 69.96 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 69.96 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 69.83 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 69.87 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 70.00 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 17.38 fps, 1826.38 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 17.39 fps, 1826.37 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 17.41 fps, 1826.38 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 17.44 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 17.38 fps, 1826.37 kb/s

I ran the same test in safe mode, just to see how big the difference (if any) would be.

---------- RUN1PASS1.LOG
encoded 1749 frames, 73.79 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 74.43 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 74.43 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 73.98 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 74.08 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 18.34 fps, 1826.37 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 18.37 fps, 1826.37 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 18.39 fps, 1826.38 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 18.34 fps, 1826.26 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 18.31 fps, 1826.37 kb/s

cpu-z.txt

[träskmannen]

2nd computer (AMD X2 4200):

---------- RUN1PASS1.LOG
encoded 1749 frames, 59.11 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 58.80 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 58.80 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 58.89 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 58.89 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 14.33 fps, 1826.37 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 14.28 fps, 1826.38 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 14.34 fps, 1826.37 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 14.33 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 14.28 fps, 1826.37 kb/s

cpuz.txt

[rhegedus]

I'm beginning to think that windows has more of an effect than imagined. My initial results were with a clean XP install - now that I've loaded all the usual stuff like Norton, Nero, Vegas and the like my results are about 5% down.

[stiltman]

That's not a windows thing. That's a bloatware thing.

If you read the specs for each app they all require more RAM. Keep adding RAM after each program is loaded and the system should run the same.

You should see how bad the system slows after running any of the latest major anti-virus programs from Norton, Trend, and McAffe. I couldn't beleave it. Even AVG's latest vesion put a drain on the system.

[soulphalanx]

---------- RUN1PASS1.LOG
encoded 1749 frames, 120.46 fps, 1850.94 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 120.41 fps, 1850.94 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 119.18 fps, 1850.94 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 119.40 fps, 1850.94 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 119.74 fps, 1850.94 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 31.83 fps, 1829.48 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 31.43 fps, 1829.12 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 31.19 fps, 1829.20 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 31.81 fps, 1829.51 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 31.65 fps, 1829.39 kb/s

[graysky]

thanks for the data... I assume by the screeenshots you're running vista? 32 or 64-bit?

[soulphalanx]

sorry, forgot to put that

32 bit Vista Ultimate

[Webster]

I was bored, so I dropped an AMD X2 4600+ (65W Windsor) into the machine below and run the test. Here is the result in the same machine/OS with different CPU:


---------- RUN1PASS1.LOG
encoded 1749 frames, 60.28 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 60.57 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 60.28 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 60.70 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 60.24 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 14.97 fps, 1826.32 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 14.99 fps, 1826.37 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 15.04 fps, 1826.37 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 14.95 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 14.99 fps, 1826.37 kb/s

Here is a blazing sempron running Window Vista home premium

---------- RUN1PASS1.LOG
encoded 1749 frames, 23.92 fps, 1854.10 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 24.15 fps, 1854.10 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 23.88 fps, 1854.10 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 18.71 fps, 1854.10 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 23.85 fps, 1854.10 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 5.64 fps, 1825.89 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 5.73 fps, 1825.89 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 5.66 fps, 1825.89 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 5.62 fps, 1825.89 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 5.65 fps, 1825.89 kb/s

[kush]

My little contribution to the effort....

Vista Ultimate (32-bit)
Intel e2140 @ 367x8/2936 (stock 1.6 - sweet 83.5% [so far] o/c!) - 1.355v BIOS
^^^ FPO/Batch #Q711A696, Pack date: 07/31/07 ^^^^ for anyone that cares
2x1GB Crucial Ballistix PC6400 / 2.1v 4:5 divider
abit IP35 Pro (P35 NB/ICH9R (?) SB)






And not that it means diddly squat, but...



-------------------------


---------- RUN1PASS1.LOG
encoded 1749 frames, 76.30 fps, 1850.89 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 78.85 fps, 1850.89 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 78.82 fps, 1850.89 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 78.81 fps, 1850.89 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 78.83 fps, 1850.89 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 18.35 fps, 1826.37 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 18.43 fps, 1826.38 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 18.29 fps, 1826.37 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 18.14 fps, 1826.37 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 18.44 fps, 1826.37 kb/s

-------------------------

Some notes: I tried NOT using anything while the test was running...But have the usual Vista stuff, and had task manager open (minimized not front window though), abit's uGuru (monitoring/overclocking/etc tool) running in the tray. Also, with TM open, I notied that the x264 program at most showed 94-95% processor time, uGuru, Firefox & TM at MOST 1% each.. So either there's hidden system services/progs that were taking up the remaining, or x264 wasn't quite taking all it could - may have made a difference (albeit TINY).

All in all, a I think a fine purchase for $66 (retail boxed). I know, a $66 CPU and $150-170 mobo is a bit of a mismatch, but I only really got it for backup purposes, and a stepping stone to a quad core once I make my mind up - Q6600, or wait for Penryn. It's my first Intel build after all.

Last still tweaking the system, I'm not sure I've hit the so-called "wall" on my e2140 yet, but I'm fairly content with what I've got @ only 1.355v (BIOS, 1.325 stock I think). In order to get it up appreciably more I'll likely have to bump up the voltage significantly. I'm on water, but I already hit low/mid-50c's LOAD, 60c'ish on hot days - my loop needs a thorough cleaning/new fluid, and I have an older version of the Swiftech Apogee WITHOUT the S775 backplate, so the contact pressure may not be optimum. Also, memory can probably tweaked more, maybe hit 4-4-4 timings, or bump speed more if I can get more out of the CPU's FSB.

[graysky]

thanks for the result!

[graysky]

Updated the Intel table. It now contains several Yorkfield ES chips including:

Xeon E5330 (Dual board)
Q9550
Q9350

[rhegedus]

Intel Q6600 overclock settings (3.2GHz 400x8)
Gigabyte P35-DS3P
2Gb DDR2 PC6400

---------- RUN1PASS1.LOG
encoded 1749 frames, 156.55 fps, 1850.94 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 154.61 fps, 1850.94 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 149.04 fps, 1850.94 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 154.18 fps, 1850.94 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 151.68 fps, 1850.94 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 42.63 fps, 1829.49 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 42.16 fps, 1829.25 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 41.27 fps, 1829.16 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 42.00 fps, 1829.51 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 41.43 fps, 1829.08 kb/s

[Alendri]

Dual Intel Xeon E5345 (Clovertown, Rev B3, Stepping 7, Stock settings (2.33ghz))
4x 2GB Corsair PC-5300 (Fully Buffered ECC modules) 5 5 5 15 @ 333mhz
Intel S5000XVN mobo (5000X chipset)
Win Vista 64-bit.


---------- RUN1PASS1.LOG
encoded 1749 frames, 127.91 fps, 1849.61 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 127.71 fps, 1849.61 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 124.26 fps, 1849.61 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 126.19 fps, 1849.61 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 125.82 fps, 1849.61 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 58.34 fps, 1834.86 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 58.04 fps, 1834.96 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 58.91 fps, 1834.85 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 58.90 fps, 1834.85 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 58.63 fps, 1834.22 kb/s

Note that this was done without so much as a reboot after a full days work. I will edit this post with a clean run tomorrow. I noted that for a while the benchmark only used 50% (exactly 50%) CPU time, tomorrow I will look more closely and see if it repeats.
I saw that you had no E5345s in your comparison, so I guess this will complete your list a little more.

Update: I have now updated the numbers with a proper run after reboot and with only minimum of applications and background processes. I have run the benchmark with and without the task manager and various other applications, and note no mentionable difference, and indeed the encoding only uses 95% of CPU time with 8 cores. Also worth mentioning is that the CPUs only works 50% during all Pass 1, this is less computing intensive, but I wonder what would be requierd then to speed it up? Harddrive performance? In that case I might mention that I have two RAID 0 7200rpm drives.

//Alendri

[graysky]

thanks for the data, guys. I'll update the table in the near future. That's an interesting observation on the 1st app, alendri. I dunno why it was exactly 50 %. Scaling issue w/ 8 cores? I'm assuming the 2nd pass was 100 %, right?

[Alendri]

...That's an interesting observation on the 1st app, alendri. I dunno why it was exactly 50 %. Scaling issue w/ 8 cores? I'm assuming the 2nd pass was 100 %, right?

I'll rerun the benchmark again tomorrow with different settings and circumstances to see if I can figure out and reproduce the "problem" with 50% usage. I'll update this post with news when I have it. And indeed the 2nd run was 100%, well actually 95%, 2% for system etc, and 3% idle.

Just a quick thought though, with todays lifestyle more and more geared towards mobility and 'on-the-go', wouldnt it be neat with a notebook compraison. The notebooks are getting very powerfull aswell. And musnt it be the ultimate to be able to mpeg4 encode your video while u're on the vacation? Packing harddrive storage has made many a back sore after all.
Maybe i'm overthinking it, but i wouldnt mind seeing a note-book comparisson table. I could supply data from at least 1 notebook. And i'm sure there are a lot of people out there with notesbooks happy to sacrifice a few moments. Encoding video on the go for putting on the web is another important factor.

//Alendri

[graysky]

Updated the tables with another 45 nm chip: the QX9650 -- both at stock levels and @ overclocked to 4.2 GHz! With it, and the others (Xeon E5330 (Dual board), Q9550, and Q9350) there is now data on 4 different 45 nm chips.

One thing that I found striking about these new chips is that they are only marginally faster than their 65 nm counterparts when encoding x264 (about 5-6 % faster with all other factors being equal or close to equal). Have a look at the general trends table for the Kentsfield vs. Yorkfield comparison at the official host.

[redwudz]

Just installed a AMD Phenom 2.2Ghz in my Gigabyte GA-MA69G-S3H MB. Overclocked to 2535Mhz and seems stable at present. But I haven't tried a 'real world' encode of a full length video to Divx, so I may have to back it down.


---------- RUN1PASS1.LOG
encoded 1749 frames, 141.07 fps, 1850.94 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 143.86 fps, 1850.94 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 143.34 fps, 1850.94 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 142.73 fps, 1850.94 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 142.88 fps, 1850.94 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 35.80 fps, 1829.32 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 35.77 fps, 1829.56 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 35.80 fps, 1829.17 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 35.64 fps, 1829.24 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 35.78 fps, 1829.48 kb/s

CPU-Z 142 is having a problem reading the RAM settings, but they should be 4-4-4-12-22-2T, 2GB of Corsair XMS DDR2-6400 in Ganged mode.

This is a screenshot of the CPU page. I'm running 'Cool and Quiet', so the CPU speed is shown at idle with a multiplier of X5.5. It runs normally at X11.



Here's the data dump from CPU-Z: cpuz.txt

[Jim44]

Nice results, redwudz. I'm thinking it's time to build a new computer (~4 years old, single core AMD ) and have been looking at the Phenom. More food for thought. Thanks.

Jim

[jkleslie]

This is the same system as my E6600@3.2GHz, but I have moved to a Q6600@3.2GHz. XPSP2 32-bit, 2GB DDR2800@stock 5-5-5-18-2T. Some better memory will also be added in the next week or two so that, hopefully, I can move up to 1066MHz or better.


---------- RUN1PASS1.LOG
encoded 1749 frames, 160.83 fps, 1850.94 kb/s

---------- RUN2PASS1.LOG
encoded 1749 frames, 160.36 fps, 1850.94 kb/s

---------- RUN3PASS1.LOG
encoded 1749 frames, 161.29 fps, 1850.94 kb/s

---------- RUN4PASS1.LOG
encoded 1749 frames, 155.47 fps, 1850.94 kb/s

---------- RUN5PASS1.LOG
encoded 1749 frames, 158.77 fps, 1850.94 kb/s

---------- RUN1PASS2.LOG
encoded 1749 frames, 44.05 fps, 1829.36 kb/s

---------- RUN2PASS2.LOG
encoded 1749 frames, 44.09 fps, 1829.50 kb/s

---------- RUN3PASS2.LOG
encoded 1749 frames, 44.03 fps, 1829.20 kb/s

---------- RUN4PASS2.LOG
encoded 1749 frames, 43.67 fps, 1829.43 kb/s

---------- RUN5PASS2.LOG
encoded 1749 frames, 43.88 fps, 1829.25 kb/s



Registers dump

cpuz.txt

[graysky]

Just installed a AMD Phenom 2.2Ghz in my Gigabyte GA-MA69G-S3H MB. Overclocked to 2535Mhz and seems stable at present. But I haven't tried a 'real world' encode of a full length video to Divx, so I may have to back it down.

CPU-Z 142 is having a problem reading the RAM settings, but they should be 4-4-4-12-22-2T, 2GB of Corsair XMS DDR2-6400 in Ganged mode.

This is a screenshot of the CPU page. I'm running 'Cool and Quiet', so the CPU speed is shown at idle with a multiplier of X5.5. It runs normally at X11.



Here's the data dump from CPU-Z: cpuz.txt

Wow dude, you posted this back in Dec and some how I missed it all these months (sorry about that). One question, when you said CPUZ was having problems reading your memory, it was 4-4-4-12 @ ? MHz? What is ganged mode? Have you upgraded to the new CPU-Z yet?

Also, since it's been so long, have you boosted your o/c at all? Care to post new results? How about stock setting results?

Thanks again dude!

p.s. running vista x32 or x64?

[redwudz]

Running X32 Vista Home Premium. The latest CPU-Z properly reads the memory. These are two screenshots done while encoding and show the correct speed.





I'm still running the same OC. The 9500 runs at 2.2Ghz stock settings. The OC takes it up to 2.5Ghz. Runs stable and fairly cool and no crashes with encoding, so I'll probably leave it there.

There are two memory modes in BIOS, Ganged and Unganged. Ganged is the normal mode where RAM is used by all four cores equally. Unganged allows each core to control it's RAM independently. I could never see much difference between the modes, so I settled on Ganged. Unganged may work better for multitasking while Ganged is better for encoding or a program that uses all four cores for the same purpose.

The EPP mode adjusts the RAM timings and voltages automatically. Only works with some RAM types. It allows the RAM to run with tighter timings at 400Mhz.

[graysky]

Cool, man, thanks.