Added more CPUs.

- Pentium 4 650 on Dell Dimension 5100 (i945G), Windows 7 - Now highest P4 HT score
- More tests and combinations of old 400 FSB P4 based CPUs using the same QDI Platinix 2 and QDI Superb 4 with Celeron 2.2, Celeron 2.6, Celeron 1.7, P4 1.5/1.7/1.8, P4 1.8A/2.0A
- Pentium III 667, 800 EB and 933 on my Optiplex GX150
- Pentium II 266, 400 on the same Tekram P6B40-A4X
- Pentium III 450 Katmai and Pentium III 650E Coppermine on the P6B40-A4X again

Thanks, Carlos! I like to see how the old timers compare. 😀

Yes, i did kinda contribute a lot, i might plan to test some systems when i have time again and my exams are over.

I got an old Intel D850EMV2 motherboard with RAMBUS memory, so i can bech Pentium 4 systems with PC800 Rambus RDRAM, also try my FSB 533 P4s.

New result : Pentium Extreme Edition 840 @ 4,25GHz - LINK
BUT I screw up adding my result... can someone add Rating for Atom X5 Z8500 ?
Sorry 🙁

New reuslt: Pentium 4 1.5 GHz Willamette (Socket 423 version) with PC800 Rambus memory
Just an old PC i built recently

agent_x007 wrote:

New result : Pentium Extreme Edition 840 @ 4,25GHz - LINK
BUT I screw up adding my result... can someone add Rating for Atom X5 Z8500 ?
Sorry 🙁

I fixed the Atom by recalculating: 1494x3.72 😀

Thanks for all the results, guys!

I resorted and noticed the newest entry seems off. It's a mobile i5-450m, but the reported score is about 1/3 of what it should be (2441 MIPS). Compare that to the i5-430m at 6717. Not sure who put it in, but maybe you can double-check it. Thx.

Added some Athlon XPs to the list:

I've got a couple of Durons (1300 and 1100) that I'll eventually get to as well.

CPU-Z seems confused about the Duron 1300. Calls it a Palomino instead of a Morgan.

I've added a Pentium 4 HT 3.4 GHz Northwood result on 7z920

Intel D875PBZ
P4 HT Northwood 3.4
2GB PC3200 2-3-2-5 dual channel
XP SP3

Added a Pentium 4 Prescott 3.2 Ghz HT on a Gigabyte GA-8IPE1000-G (Rev 4.0) and a Athlon XP 3000+ (FSB 400) on a MSI KT6V-LSR. More Athlons and Athlon XPs comming soon

Added an Athlon 1.2 Ghz Thunderbird and a Duron 700 as well

Added results from multiples AMD K6-II models (actually was a K6-II 500 underclocked to simulate other models) and a hypothetical AMD K6-II 166 with underclock. Pentium MMX and Pentium 1 results comming, 7-zip is running really slow on the Pentium MMX.

Note. Since i installed 512 MB RAM, i exceded the max cacheable RAM area (256 MB) and it might slowdown the system, idk how much slowdown though

Added Pentium MMX 166 and 200, also underclocked to 133 Mhz as well.
Added Pentium 100, 133 and 166
Same testbed as my AMD K6 II Tests

http://prntscr.com/bp9wyh
I don't think nobody will beat this xD

Carlos S. M. wrote:

Added Pentium MMX 166 and 200, also underclocked to 133 Mhz as well. Added Pentium 100, 133 and 166 Same testbed as my AMD K6 II […]
Show full quote

Added Pentium MMX 166 and 200, also underclocked to 133 Mhz as well.
Added Pentium 100, 133 and 166
Same testbed as my AMD K6 II Tests

http://prntscr.com/bp9wyh
I don't think nobody will beat this xD

Thanks! There's some really interesting results in there. 😀 Like
-MMX beating non-MMX by 11-14%
-K6-II beating Pentium MMX by 25%
-Pentium II beating K6-II by 36%
-Pentium III beating K6-II by 50%

Very cool stuff.

clueless1 wrote:

Thanks! There's some really interesting results in there. :) Like -MMX beating non-MMX by 11-14% -K6-II beating Pentium MMX b […]
Show full quote

Carlos S. M. wrote:

Added Pentium MMX 166 and 200, also underclocked to 133 Mhz as well. Added Pentium 100, 133 and 166 Same testbed as my AMD K6 II […]
Show full quote

Added Pentium MMX 166 and 200, also underclocked to 133 Mhz as well.
Added Pentium 100, 133 and 166
Same testbed as my AMD K6 II Tests

http://prntscr.com/bp9wyh
I don't think nobody will beat this xD

Thanks! There's some really interesting results in there. 😀 Like
-MMX beating non-MMX by 11-14%
-K6-II beating Pentium MMX by 25%
-Pentium II beating K6-II by 36%
-Pentium III beating K6-II by 50%

Very cool stuff.

Yes. i still need to try other CPUs i have, also i added results from a Sempron 3100+ (Using a PC with a Foxconn SiS 760GX motherboard) and a Pentium D 935 (using another PC with my ASUS P5PE-VM).

Possibly comming soon (Yes, i own all of these CPUs xD):
(ASUS P5PE-VM testbed)
Celeron 420 (Slowest Core 2 based Celeron)
Celeron D 331
Celeron D 336
Celeron D 346
Celeron D 352
Celeron D 355 (Strongest Prescott based Celeron D)
Core 2 Duo E4500
Core 2 Duo E6300 (maybe if the mobo does FSB 1066)
Core 2 Duo E6400 (maybe if the mobo does FSB 1066)
Pentium E2140 (Slowest Core 2 based Pentium)
Pentium E2140
Pentium 4 516
Pentium 4 520/520J
Pentium 4 530/530J
Pentium 4 540/540J
Pentium 4 550/550J
Pentium 4 630
Pentium 4 640
Pentium 4 650
Pentium 4 631
Pentium D 820
Pentium D 830
Pentium D 930
Pentium D 950

(Foxconn 760GX testbed)
Sempron 2600+ (Rev E6)
Sempron 2800+ (Rev E6)
Athlon 64 3200+ (S754, Newcastle, 2.2 Ghz)

(Socket 478 i865PE testbed, ethier my ASUS P4P800-E, Abit AI7 or my GA-8IPE1000-G)
Pentium 4 2.0 Ghz (FSB 400)
Pentium 4 2.4 Ghz (FSB 533)
Pentium 4 2.53 Ghz
Pentium 4 2.6 Ghz (FSB 400)
Pentium 4 2.66 Ghz
Pentium 4 2.8 GHz (FSB 533)
Pentium 4 2.8A Ghz
Pentium 4 HT 2.6C Ghz
Pentium 4 HT 2.8C Ghz
Pentium 4 HT 2.8E Ghz
Pentium 4 HT 3.0C Ghz
Pentium 4 HT 3.0E Ghz
Pentium 4 HT 3.2C Ghz

(Socket 478 i845PE testbed, QDI Platinix 2E/333)
Pentium 4 2.0 Ghz (FSB 400)
Pentium 4 2.4 Ghz (FSB 533)
Pentium 4 2.53 Ghz
Pentium 4 2.66 Ghz
Pentium 4 2.6 Ghz (FSB 400)
Pentium 4 2.8 GHz (FSB 533)

(Socket 478 i845 SDRAM testbed, QDI Platinix 2A)
Pentium 4 2.6 Ghz (FSB 400)

(Socket 478 SiS 645 testbed, QDI Superb4)
Pentium 4 2.6 Ghz (FSB 400)

There are some interesting results you might see like Smithfield vs Presler (PD 830 vs PD 930) and Northwood vs Prescott vs Prescott-2M vs Celdar Mill (P4 3.0C vs 530/530J/3.0E vs 630 vs 631). Also 128 KB vs 256 KB L2 on the Semprons

I know this is gonna take some time due to the amount of CPUs and i do more stuff in real life besides computers

Also i ordered a Pentium 4 670 (3.8 Ghz) which is the highest clocked P4 Intel ever made along with the P4 570/570J/571 which i'm gonna build a PC off it and bench it. (Is gonna take some times since i ordered from China via eBay)

Well, i decided to give a try to my i5 again with the right 7-Zip version and different OCs (by increments of 200 mhz excpet 4.7 which is my current OC and 5.1 ghz which is my max stable) and stock clocks

http://prntscr.com/bpb4d6

at 4 Ghz or less i loss agaist the i5 4570S when i need 4.7 Ghz to barely surpass a 4.4 Ghz i5 4670K, but this 5 year old Sandy Bridge i5 especially when overclocked can still hold agaist newer CPUs

Is it normal, that on my I5 650 dual core w/hyperthreading I get better results on this test when using 8 threads on the benchmark?

CPU overclocked to 3.75 ghz

Underutilised CPU = higher score when it's forced to do more in the same ammount of time.
Also, memory performance sucks on i3 5xx/6xx CPU's, because they use 32nm CPU core's with 45nm L3 + IMC. They are connected to eachother via QPI which can't handle IMC request fast enough (at standard frequency).

Try overclocking QPI bus, and see if it helps with your score.

agent_x007 wrote:

Underutilised CPU = higher score when it's forced to do more in the same ammount of time.
Also, memory performance sucks on i3 5xx/6xx CPU's, because they use 32nm CPU core's with 45nm L3 + IMC. They are connected to eachother via QPI which can't handle IMC request fast enough (at standard frequency).

Try overclocking QPI bus, and see if it helps with your score.

Yeah, the QPI is overclocked from the standard 133Mhz to 150Mhz, giving a QPI speed of 3600Mhz. On my Intel motherboard that is pretty much the max overclock I can get out from it, the mobo being Intel DH55TC. In Aida 64 memory tests the rise in QPI and memory clocks gives around 1,5-5 GB/s rise compared to stock clocks, but thanks to bad overclock options on the board, increases latency by around 3ns