results added for VIA C3 1.2 GHz "nehemiah" on SIS 630 motherboard.

playing around on my file server:

Just goes to show you, 2 processors are better than one! Also done on the newer pci-e chipset.

Results added for Celeron coppermine 600 on SIS 630 motherboard.

Looks like upgrading from my old OCed Q6600 to i5-6600 (again 6600? aargh 😵 ) in 7zip brings ~2x more performance.

added results for my new Pentium II 350 MHz Toshiba Equium 7100s.

Added results for my Core 2 Extreme QX6800 system (2.93GHz stock, OC'ed to 3.3GHz). WinXP, 4GB ECC RAM, Intel D975XBX2 motherboard.

Added results for an Opteron 1216 overclocked at 2.88 GHz

Thanks for adding your results. 😀

clueless1 wrote:

Thanks for adding your results. 😀

np, i probably won't stop untill i run out of CPUs/PCs xD

Just added an i5-6400. Sadly, it is slightly slower than the i5-4460 even though it's got 100Mhz on the 4460.

Just added a Cyrix 6x86 MII PR333 (75MHz x 3.5) @ 262 MHz.

BSA Starfire wrote:

Just added a Cyrix 6x86 MII PR333 (75MHz x 3.5) @ 262 MHz.

First Cyrix on the list, nice! Looks like it's right on par with the K6-II 233 overall. The Cyrix has higher memory bandwidth, the K6-II slightly more processing power.

Added my other Cyrix MII 333, this one is on a ALi Aladdin 5 motherboard(83 MHz x 3.0) for 250 MHz, it's quite a bit slower than the other machine on the SiS 5591/5595 despite slower bus speed of 75 MHz. SiS chipsets do really like the Cyrix, Linear burst mode is enabled(I've edited my spreadsheet entry now it's enabled, the CMOS battery was flat so BIOS defaults had been loaded, sorted that out now with a new CR2032!), I don't think the ALi has that option.

Added Athlon64 X2 4200+, it was running v9.20 x64 version on XP64.
Does the disk performance matter for this benchmark? This system is running from an SSD, not sure if that should be noted. Seems like it ranks higher than I would expect so I'm wondering if that's why.

First time I've edited one of these things. I checked the revision history and it's highlighting as if I edited a couple cells below the row I added. I didn't touch them so I don't know why it says that but I compared with the previous version to make sure the values didn't change.

shamino wrote:

Added Athlon64 X2 4200+, it was running v9.20 x64 version on XP64.
Does the disk performance matter for this benchmark? This system is running from an SSD, not sure if that should be noted. Seems like it ranks higher than I would expect so I'm wondering if that's why.

First time I've edited one of these things. I checked the revision history and it's highlighting as if I edited a couple cells below the row I added. I didn't touch them so I don't know why it says that but I compared with the previous version to make sure the values didn't change.

Everything looks good shamino. The SSD shouldn't effect the score unless you're in a very low-RAM system (even then, the ssd should let you run the benchmark on a low ram system with comparable results). There's lots I don't understand about what Sheets shows in the revisions. But when I first made this sheet, I set up the cpu usage column as a formula that would auto-fill when you put in the rating and rating per usage. But it wasn't set up very clearly, so people just typed the percentage in. If you check those two below your entry, they are using the formula, whereas the surrounding entries are typed in. That might have something to do with the revision weirdness.

Thanks for the entry 😀

Ran a few tests on my Phenom2 machine. It's triple booting right now so I tried 3 OSes:

Phenom2 X2 555 (2x 3.2GHz) (WinXP64)

Phenom2 OC and unlocked to 3 cores (3x3.6GHz) (WinXP64, 4 execution threads used):

Both of these above scores were posted to the spreadsheet.

The following are for curiosity, I did not post them but feel free to add them if you don't think it's just clutter:
Same clocking, this time in WinXP32:

Same clocking, Linux Mint 17.2 x64:

17z -mmt=4 b
2
37-Zip [64] 9.20  Copyright (c) 1999-2010 Igor Pavlov  2010-11-18
4p7zip Version 9.20 (locale=en_US.UTF-8,Utf16=on,HugeFiles=on,3 CPUs)
5
6RAM size:    5966 MB,  # CPU hardware threads:   3
7RAM usage:    850 MB,  # Benchmark threads:      4
8
9Dict        Compressing          |        Decompressing
10      Speed Usage    R/U Rating  |    Speed Usage    R/U Rating
11       KB/s     %   MIPS   MIPS  |     KB/s     %   MIPS   MIPS
12
1322:    9263   270   3341   9011  |   107838   283   3438   9729
1423:    9396   283   3382   9573  |   106917   284   3439   9784
1524:    9527   292   3514  10244  |   106773   288   3433   9906
1625:    9384   289   3710  10714  |   103443   282   3445   9727
17----------------------------------------------------------------
18Avr:          283   3487   9886               285   3439   9786
19Tot:          284   3463   9836

I think the score is supposed to be an average of Compressing MIPS and Decompressing MIPS on line 25 (32MB), so that score is 10220.

So with the same system and CPU clocking, WinXP32 scored 9467, WinXP64 scored 10087, and Mint x64 scored 10220. There is some variance of course, with these being the best of a few runs each. I don't think the difference between linux and XP64 is repeatable/significant, but the slight gap between 32bit and 64bit was definitely there.

One thing about the "Number of CPU threads" setting:
7zip doesn't allow using 3 CPU threads so with my 3-core system I set it to 4. I also tried using 6. I found that using an excessive number of threads increases the score a bit further by a few hundred points (I didn't post pics of those). Since everyone up to now has been using a number of threads that matches the number of CPU cores, to be as consistent as possible with that my triple core score used 4 threads, not 6.

I wish I still had a Pentium Extreme Edition-capable motherboard... I remember upgrading from that ridiculous CPU ("Presler XE", dual-core Pentium 4 with Hyperthreading) to a Pentium Dual-Core (Conroe/Allendale) and almost everything I did was noticeably faster. I had the PEE 965 (my favorite acronym for it,) usually overclocked to 4 GHz - a $999, 130W CPU. I "upgraded" to a Pentium Dual-Core E2180 at 2 GHz, an $84, 65W CPU. (I was actually turning my former gaming rig in to a server, at the same time I dropped the PCIe GPU in favor of onboard graphics; I had gotten a new Core 2 Extreme gaming rig.) "Max benchmark" power on the new config was lower than idle power was on the old - and for almost everything CPU-related, it benchmarked faster.

Sadly, I ended up frying that motherboard, and my only S775 board is Core 2+ only.

luckybob wrote:

playing around on my file server: http://i.imgur.com/0Pe6wStm.jpg […]
Show full quote

playing around on my file server:

Just goes to show you, 2 processors are better than one! Also done on the newer pci-e chipset.

i just noticed now, why used 32 bit 7zip under Windows Server 2008 R2 which is x64? I'm pretty sure 7zip x64 will give a better score on x64 Systems

@carlos S. M.

Interesting. I could have sworn i was using x64. Next time im inside my server, i will have a look to be sure.

Anonymous Freak wrote:

I wish I still had a Pentium Extreme Edition-capable motherboard... I remember upgrading from that ridiculous CPU ("Presler XE", dual-core Pentium 4 with Hyperthreading) to a Pentium Dual-Core (Conroe/Allendale) and almost everything I did was noticeably faster. I had the PEE 965 (my favorite acronym for it,) usually overclocked to 4 GHz - a $999, 130W CPU.

Sadly, I ended up frying that motherboard, and my only S775 board is Core 2+ only.

I already tested and OC'ed both PXE* 965/840 versions 😀
Also : Almost all Core 2 MB's should support "PEE 965" (since it has 1066MHz FSB).

*My fav. acronym for them 😉