2.4B vs 2.4C

[LastRide]

**My video card was clocked at 409.50/369.00 on all tests performed.**

I ran just a few tests in comparison with both my 2.4B and 2.4C CPU's.Aida32,Sandra buffered and Unbuffered and 3DMark2001SE.This is not a major Thugsrook testing but just a few comparisons.:)

Test Setup Major Components= Asus P4P800,2X256MB OCZ Gold PC3700,2.4B & 2.4C CPU's,ATI Radeon 9800 Pro

**2.4B @ 3.42GHz,2/3/3/6(4) 475 DDR 4:5 190 FSB 2.65 Vdimm**
**2.4C @ 3.42Ghz,2/3/3/6(4) 456 DDR 5:4 285 FSB 2.65 Vdimm**

**Testing done with Performance on Standard and MAM Auto**

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

****2.4B****

---Aida32---

Read= 5156 Write= 2231

--Sandra--

Buffered= 4857/4859
Unbuffered= 3038/3077

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

****2.4C****

--Aida32--

Read= 5451 Write= 2210

--Sandra--

Buffered= 6098/6122
Unbuffered= 3035/3105



****3DMARK2001SE**** 2.4B= 20,233 and 2.4C= 20,512



============================================================================================

UT2K3 1024 X 768 (2.4C @ 3.42GHz 285FSb 5:4 456 DDR 9800 Pro)
Everything on performance on video card or off.9800 Pro clocked to 409.50/369.00

Flyby: 256.1
Botmatch: 95.6


--------------------- Q3Bench RESULTS ---------------------


BENCHMARK SETUP:
----------------
DEMO: DEMO001
CFG FILES RUN: Normal High Maximum
RESOLUTIONS RUN: 1024x768

NUMBER OF RUNS: 3 times.


NORMAL Configuration File Settings:

(1024x768) 387.7, 387.4, 388.5, AVERAGE FPS: 387.87

HIGH Configuration File Settings:

(1024x768) 386.7, 386.9, 387.2, AVERAGE FPS: 386.93

MAX Configuration File Settings:

(1024x768) 377.0, 377.5, 377.1, AVERAGE FPS: 377.20

============================================================================================

2.4C 250 FSB 1:1 2.5/3/3/7 (8) 500DDR 1024 x 768

UT2K3

244.9
88.8

I disabled my sound for this run on Q3 so I don't know as of yet if it reflects the score being higher.UT2K3 is lower on 1:1 250FSB DDR 500.

BENCHMARK SETUP:
----------------
DEMO: DEMO001
CFG FILES RUN: Normal High Maximum
RESOLUTIONS RUN: 1024x768
NUMBER OF RUNS: 3 times.


NORMAL Configuration File Settings:
(1024x768) 403.0, 402.5, 402.8, AVERAGE FPS: 402.77

HIGH Configuration File Settings:
(1024x768) 402.5, 402.3, 402.5, AVERAGE FPS: 402.43

MAX Configuration File Settings:
(1024x768) 389.8, 389.0, 379.4, AVERAGE FPS: 386.07

============================================================================================

Disable sound.2.4C 285FSB 5:4 2/3/3/6 (4).I used a little more aggressive timings since its 5:4.Same as the first run.

BENCHMARK SETUP:
----------------
DEMO: DEMO001
CFG FILES RUN: Normal High Maximum
RESOLUTIONS RUN: 1024x768
NUMBER OF RUNS: 3 times.


NORMAL Configuration File Settings:
(1024x768) 434.8, 433.6, 434.3, AVERAGE FPS: 434.23

HIGH Configuration File Settings:
(1024x768) 433.5, 433.2, 433.5, AVERAGE FPS: 433.40

MAX Configuration File Settings:
(1024x768) 409.6, 414.7, 412.9, AVERAGE FPS: 412.40

UT2K3 1024 x 768

Flyby: 266.8
Botmatch: 96.6

============================================================================================

Here is 2.4C 285 FSB 5:4,2.5/3/3/7 (8) 456DDR

UT2K3

Flyby: 266.2
Botmatch: 96.2

------------------------------------------------------------------------------------
Quake 3 bench

DEMO: DEMO001
CFG FILES RUN: Normal High Maximum
RESOLUTIONS RUN: 1024x768
NUMBER OF RUNS: 3 times.


NORMAL Configuration File Settings:
(1024x768) 435.0, 434.5, 434.8, AVERAGE FPS: 434.77

HIGH Configuration File Settings:
(1024x768) 433.5, 433.6, 433.8, AVERAGE FPS: 433.63

MAX Configuration File Settings:
(1024x768) 414.2, 413.5, 413.8, AVERAGE FPS: 413.83

============================================================================================

3DMARK2001SE 2.4B and 2.4C

1:1 250 FSB 500 DDR 2.5/3/3/7 (8) = 18,999 2nd 19,034 2.4B
5:4 285 FSB 456 DDR 2.5/3/3/7 (8) = 20,413 2.4C

I thought maybe there was a problem with the first run on 1:1 so I ran it again.No difference basically on the second run.I know at 5:4 I am always around that area.

============================================================================================

********2.4B @ 3.42GHz 266 (1:1) 380 DDR 2/3/3/6 (4)*******

Quake 3

DEMO: DEMO001
CFG FILES RUN: Normal High Maximum
RESOLUTIONS RUN: 1024x768
NUMBER OF RUNS: 3 times.


NORMAL Configuration File Settings:
(1024x768) 420.2, 420.4, 421.0, AVERAGE FPS: 420.53

HIGH Configuration File Settings:
(1024x768) 419.8, 419.4, 419.8, AVERAGE FPS: 419.67

MAX Configuration File Settings:
(1024x768) 403.6, 403.1, 403.6, AVERAGE FPS: 403.43


UT2K3

Flyby: 262.0
Botmatch: 94.7

3DMARK2001SE 19,960

============================================================================================

2.4B @ 3.42GHz 2/3/3/6 (4) 475 DDR 4:5 (333)

DEMO: DEMO001
CFG FILES RUN: Normal High Maximum
RESOLUTIONS RUN: 1024x768
NUMBER OF RUNS: 3 times.


NORMAL Configuration File Settings:
(1024x768) 435.3, 434.9, 434.9, AVERAGE FPS: 435.03

HIGH Configuration File Settings:
(1024x768) 434.3, 433.9, 434.3, AVERAGE FPS: 434.17

MAX Configuration File Settings:
(1024x768) 413.8, 413.0, 413.3, AVERAGE FPS: 413.37
------------------------------------------------------------------------------------

UT2K3

Flyby: 266.7
Botmatch: 96.7
------------------------------------------------------------------------------------

3DMARK2001SE 20,352

[ThugsRook]

oooo nice :-)

[LastRide]

Thats all I got to testing.I couldn't test 1:1 cause there is no 1:1 with a 2.4B on a P4P800.

[ThugsRook]

you sure? :scratch:

i had 1:1 on that board @ 190fsb :wack:

[Duvie]

What does it mean??? Give us your theory!!!!:wave:

[LastRide]

I was running 190 and all i had was 266,333 and auto,no 400 which is 1:1 that shows up with the 2.4C.Unless auto was 1:1 and I didn't realize it.I can recheck for that later on.

Duvie as for theory,I don't really know.If you consider the 279 point difference in 3DMark2001SE,the Aida32 Read difference and the BS Sandra buffered scores worth it than thats cool.Both feel about the same speed.Its up to the individual to consider if its worth the upgrade.I can see how someone would be happy with a good "B" CPU.

**I get major artifacting with 3DMark03 on CPU test 1 and 2.The computer will crash to desktop,BSOD or reboot.I have no idea whats going on there.Using the new Cat 3.5's.**


Originally posted by ThugsRook
you sure? :scratch:

i had 1:1 on that board @ 190fsb :wack:

[Duvie]

Originally posted by LastRide
I was running 190 and all i had was 266,333 and auto,no 400 which is 1:1 that shows up with the 2.4C.Unless auto was 1:1 and I didn't realize it.I thought it was by speed.I can recheck for that later on.

Duvie as for theory,I don't really know.If you consider the 279 point difference in 3DMark2001SE,the Aida32 Read difference and the BS Sandra buffered scores worth it than thats cool.Both feel about the same speed.Its up to the individual to consider if its worth the upgrade.I can see how someone would be happy with a good "B" CPU.

**I get major artifacting with 3DMark03 on CPU test 1 and 2.The computer will crash to desktop,BSOD or reboot.I have no idea whats going on there.Using the new Cat 3.5's.**


Likely the Cat 3.5's...i would go back to the more tried and true 3.3's....That is where I am at....


I think you guys are convincing me a good B chip and the right DCDDR mobo and I am right there minus the HT benefitd in the heavy multitasking world...

Good info!!!

[LastRide]

A good "B" coupled with an IC7 or other and your on your way.:)

[Duvie]

That is what I am thinking...I thought maybe IS7 but PastorJay may be skewing me from that in the ocing area....

[oldfart]

Originally posted by LastRide
Thats all I got to testing.I couldn't test 1:1 cause there is no 1:1 with a 2.4B on a P4P800.
Since that is a "B" CPU, wouldn't 266 be 1:1 and 333 be 4:5? What does CPUz read as mem speed?

Nice comparison. I've wanted to see a "B to C" comparison @ same CPU speed. The difference is pretty small in 3Dmark (1.4%). As usual, synthetic mem benches dont mean a thing. :rolleyes:

Interesting to note how the 5:4 ratio DDR456 is a bit faster than the 4:5 ratio DDR475 setup. Maybe due to the higher FSB or the HT on the "C" chip?

Can you run some more benches? UT2K3, Q3, SS or SS2 for game benches would be cool. MP3 and Video encoding would be interesting to see. How about archiving? Winzip or WinRar compression??

[ThugsRook]

266 = 1:1
333 = 4:5

:wave:

[LastRide]

Thanks Thugs.Oldfart,I can run some of those other benches.As far as Encoding goes that stuff is over my head.Duvie showed me a little bit on it.

[oldfart]

Thanks!

I like to see/run benches on REAL applications that people actually use on a PC. I think way too much effort is spent on things like Sandra/Aida and even 3DMark to a lesser degree.

MP3 encoding is pretty easy. If you want to try that out, me or someone else here can get you started on that pretty easily.

[Duvie]

Hey oldfart...Have you notice that a few more ppl are using Pinnacle Studio 8.5 in some mpeg2 encoding test???

I agree lame mp3 encoder or besweet is the ones I use....

[oldfart]

Originally posted by Duvie
Hey oldfart...Have you notice that a few more ppl are using Pinnacle Studio 8.5 in some mpeg2 encoding test???

I agree lame mp3 encoder or besweet is the ones I use....
I just wish Studio 8 had a timer so you didn't have to time it manually. TMPGEnc shows elapsed time, which makes it easier for benching. LAME is a great encoder. Never tried besweet myself.

I'm not the biggest THG fan by any means, but I like the suite of programs they use for benching. Besides OpenGL, D3D games, they do video, audio encoding, WinRAR benching, and office apps. This gives a better idea of real world system performance.

People have to get off this SiSoft/Aida kick. They have little to do with real world system performance. Why dont we bench the programs that we actually USE??:scratch:

[oldfart]

Gonna have the 2.4B benches up soon? Nice UT2K3 score BTW. Has anyone broken 100 on botmatch?

[LastRide]

Maybe I'll try 1:1 first befor I put the 2.4B back in.My cable internet and TV was down over 8 hours today.Just got back online.

[anvil]

A little off topic, but your results imply 5:4 is faster than 1:1 for maximizing a 2.4c, ... right?

I'm still failing to see where 1:1 rules, but Thugs, Prom and others say it does?!? What am I missing here?

Seems also to imply that OCZ Gold is probably overkill for 2.4c since max performance is at 456ddr.

Wise ones, ... please straighten out a novice's misinterpretation.


anvil

[oldfart]

I'm still failing to see where 1:1 rules
Thats because it doesn't.

1:1 does not rule for 2.4C and 2.6C setups. You severely limit your CPU and FSB speed running 1:1. My testing, zroc's, numerous web reviews, and also Lastride's confirm this. the "must run 1:1" is simply not true.

All you get from running 1:1 on a 2.4/2.6C (maybe 2.8C is some cases) is the privilege of spending a lot of $$ on ram, running 1:1 and having a slower system.

[oldfart]

Nice work.

Could you glue all the scores together in one chart?

[oldfart]

2.4B @ 3.42 4:5 DDR475 and 2.4C @ 3.42 5:4 DDR456 are basically identical! Interesting info.

Why did you run 2.4B DDR475 with 2-3-3-6 timings and 2.C DDR456 @ 2.5-3-3-7 on the 3DMark test?

[LastRide]

Oldfart,its on my first post near the bottom of the first testing.It has both for 3DMark2001SE at 2/3/3/6 (4).I should have kept things in order.

******3DMARK2001SE****** 2.4B= 20,233 and 2.4C= 20,512

[oldfart]

Oops, Ok :blush:

Thats why you need to glue all this together when its done.

[LastRide]

I know its hard to follow.I did a little bit of a sloppy job:blush:.I am not known for running benches to well,hehe.I should have run stuff in order and keep them together,oh well.Maybe I will be a pro like Thugs one day..LOL.I put all the benches on my first post together.The ----- is to seperate the benches and the ======== is where the new benches start.You will need to read and see what CPU,timings,ratios and whatever I am testing cause its all over the place.The first is both 2.4B and 2.4C.Then its 2.4C for a ways and near the bottom is the 2.4B benches.If someone wants to show me MP3 encoding or Winzip and Winrar compression I'll test them out.Duvie showed me DVD encoding but it was a pain.Had to install a whack of stuff and gator with it.Plus sometimes it worked and sometimes it didn't so I deleted all that crap it installed in my system and be done with it.

[oldfart]

Winzip/WinRar is pretty easy. You just need to time how long it takes to compress some files. There is a guy at AT (I forget who) who used the UT (original) CD. Copy the entire contents of the UT CD to your HD. Compress it all into one file and time how long it takes. WinZip or WinRar should be set for maximum compression. It doesn't have to be the UT CD. anything will work, but make it fairly big, and hard to compress.

MP3 is not too hard either. Download CDEX (http://www.cdex.n3.net/) CDEX is a MP3 tool based on the LAME encoder. It can covert directly from CD audio to MP3, which is a great feature. Doing that is not a good bench test though. You can be limited by the DAE rate of your CD drive. To bench, do it this way. First, convert the CD audio files to wave and store them on your HD using Convert/extract CD track(s) to Wave files. Do a whole audio CD. Next, convert to MP3 with Convert/wave file(s) to compressed audio files. Time it and see how long it takes.

[LastRide]

Thanks for the info.I'll give it a try later.:)