3.5Ghz@Aircooling (CNPS 7700ALCU)
Asus P5B
2Gb Mushkin DDR2-800
http://users.pandora.be/overclocking/E6400@3.5.JPG
Printable View
3.5Ghz@Aircooling (CNPS 7700ALCU)
Asus P5B
2Gb Mushkin DDR2-800
http://users.pandora.be/overclocking/E6400@3.5.JPG
how come your cpuz reports real v-core in a p5b. mine is bugged.Quote:
Originally Posted by o_neill_dude
i can only read v-core with ai suite.
o cpu-z le bem a voltagem , so deixa de ler a partir de 1.47v +-Quote:
Originally Posted by remorema
cumprimentos
tkx mate :slap:Quote:
Originally Posted by mascaras
What temps are u getting and with what ambient? Is it prime stable?Quote:
Originally Posted by o_neill_dude
My config 24/7
E6400 batch L627A730
cooler : watercool nexxxos xp silver
mobo : Gigabyte DS3 (F3 bios)
Ram : Teamgroup PC 5300 CL3-4-4-8
E6400@3300mhz @ 1.33v || ORTHOS -Blend Test
http://img153.imageshack.us/img153/1250/3300135vyo7.jpg
my E6400 week 27..
air cooling
430*8 @1.36vcore
http://img91.imageshack.us/img91/871/13afu3.jpg
I prefer run Orthos using Blend test (CPU & RAM) or Small FFts ( stress CPU) cause large FFTs test just stress RAM .
regards
Large FFT's doesnt just stress ram, it stresses the CPU also. Its just with the size of the FFT's, it cannot all fit in cache, so system ram is also used extensively.
Small FFT's do ONLY stress CPU.
i believe large fft is more accurate then blend test.. last time during my AMD rig, blend test nvr gave error, but one day i tried large fft, it failed right after 10mins..
Indicating you had a ram problem.Quote:
Originally Posted by irenic
suppose blend test is mix ram and cpu.. rite?
Processor: E6400 @ 3.8Ghz
Motherboard: P5B VANILLA
Memory: GSKILL 2 x 1 GB HZ PC6400 @ 4-4-4-12 950mhz
Cooling: NEXXUS XP SILVER
FSB: 475
Memory Mult: 1:1
vcore: 1.6V
vdimm: 2.2V
pcie: 110
Still looking for 24/7 config. i'll update when i find something stable for 24/7.
http://img235.imageshack.us/img235/4525/3800ok5.jpg
playing with p5b vanilla :D
http://img112.imageshack.us/img112/2662/500wa9.jpg
yep , thats rightQuote:
Originally Posted by irenic
regards
ok,since this dual core processor is so great that i can easily run dual prime and play pro evolution soccer 5, so i ran another 8hrs (it's almost 9 act) of orthos, but this time i used blend test.. :D
http://img153.imageshack.us/img153/4656/14ru5.jpg
Could someone tell me something or inform me a bit more about the Batch L629A755 batch of E6400 conroe's?
Regards
This one is 45 hours dual prime stable with Large FFT's. ABSOLUTELY 100% STABLE!
e6400 Week 25
DS3 F3 Bios
Vcore 1.38 manually in bios
All other voltage default. I have messed with the other voltages before but found them to have little effect.
Everything else per sig.
http://members.dslextreme.com/users/...stable_f3b.JPG
Quote:
Originally Posted by dinos22
So if my week 28 E6400 runs 3.2-3.3Ghz at only 1.32v .. I should keep it?
Im thinking of buying a E6600, but I dont know if it worth it. I dont bench much theese days, so as long as my games run smooth im kinda happy.
Either my P5B or my XMS 5400C4 keeps me from running any higher FSB, so I dont know how far it will go.
With 3.3 @ 1.32v ... 3.5-3.6 on 1.4-1.45v maybe?
so decide - 3.2 or 3.3 GHz. But in both cases my answer is YES!Quote:
Originally Posted by Formann
in games the difference between E6400 and E6600 is almost none. If your E6400 overclocks well, then a 3.2 GHZ E6400 will be even better than an X6800.Quote:
Im thinking of buying a E6600, but I dont know if it worth it. I dont bench much theese days, so as long as my games run smooth im kinda happy.
XMS5400 is not a problem, 'cause it should do DDR900 without problems.Quote:
Either my P5B or my XMS 5400C4 keeps me from running any higher FSB, so I dont know how far it will go.
you need 0.1V more to get about 180 MHz.Quote:
With 3.3 @ 1.32v ... 3.5-3.6 on 1.4-1.45v maybe?
My week 28 e6400 does about the same, cant check its max though because my RAM is FUBAR.
My E6400 week 27 max SPI 1M benchble with vcore Default is 3350mhz
E6400 @ 3350mhz @ Vcore default
http://img171.imageshack.us/img171/3...defaultop5.jpg
regards
That's mine exactly. And it's just not worth the extra heat imo. I'm below 50C and happy. :DQuote:
Originally Posted by Formann
on bad axe
http://img74.imageshack.us/img74/1010/shot0014dw2.jpg
RAM is 2 x OCZ2P8001G , PC6400 4-5-4-15 2.1v
Lapping
I’ve finally got my E6400 running at a reasonable temperature. Using Orthos/dual Prime it had been reaching 70 deg C (T Junction) even at 400 MHz FSB, so I decided to take another look at the heat spreader. As you can see in Photo 1 it was clearly raised at the edges. See also that it has scratched the base of the heat sink where it was clamping at the edges only.
Photo 1
http://members.iinet.net.au/~gbyers/cpulap.jpg
After a bit of wet and dry at 360, 600, 800 and 1200 grit I reached a point where the centre of the IHS was bared to copper so the heat sink would make close contact directly above the core. I didn’t want to keep sanding until all the nickel was gone without testing the effect of this first.
With everything reassembled and dual instances of Prime running for over two hours (400 FSB) I now had a CPU temperature of 62 deg C, a benefit of 8 degrees C! – and that’s before the 200 hour bed in of the AS5 thermal grease. (Ambient was 23.5). I didn’t believe it before, but you have to try it if you think your temperatures are limiting you.
North Bridge Strap
See this thread
http://www.bleedinedge.com/forum/showthread.php?t=23803
I’ve done some benchmarking with Driver Heaven Photoshop Benchmark v2 and SiSoft Sandra 2007 to test out this issue of the chipset strap setting. It can be clearly seen that 400 FSB is superior to 401 FSB and almost as fast in day to day tasks as using 425 FSB. However there is the option of setting BIOS to 400 for booting and then using ClockGen to raise the FSB to 425 from within Windows at start-up. This option produces superior benchmarks but requires a substantial FSB and NB voltage boost. See table below.
FSB--------------------------Driver Heaven--—SiSoft Sandra Mem BW
Boot 400 FSB------------------105.0 s------------7272 MB/s
Boot 401 FSB------------------116.0 s------------6700 MB/s
Boot 425 FSB------------------101.1 s------------7200 MB/s
Boot 400 ClockGen 425 ------98.5 s-------------7318 MB/s
Based on the results here it is apparent that this board produces great results at 400 FSB with minimal voltage increases over default required. The voltages required at each of these settings are as follows:
FSB---------------------------CPU V BIOS--Mem V--FSB Term V--NB V
Boot 400 FSB------------------1.3875------1.90--------Auto-------Auto
Boot 401 FSB------------------Ditto
Boot 425 FSB------------------1.5125------2.00-------1.400------1.45
Boot 400 ClockGen 425 ------1.5125------2.15-------1.450------1.55
Actual CPU Vload is 0.06V less than BIOS setting in all cases.
The maximum clock at all the pink (one less than red) voltages in BIOS (CPU 1.55V, FSB 1.4V, NB 1.45V)? That’s 450MHz FSB or 3.6 GHz, but it needs all fans running at maximum which I don’t want, and it fails Orhtos after a time. 438 FSB / 3.5 GHz is rock solid (see Photo 2), but requires too much voltage and fan noise
Photo 2
http://members.iinet.net.au/~gbyers/438%20FSB.jpg
Orthos (dual Prime) and ClockGen
I did have a suspicion that using ClockGen would allow some to cheat the benchmarkings. However, Orhtos reports the boot FSB frequency even after a change is made to the FSB with the PLL of ClockGen. Ha, so you can’t cheat on Orthos.
Memory Settings
Using MemSet 3.0 is the easiest method of optimising your memory timings. From within Windows you can change most timings and apply them without rebooting. Obviously changing some settings will freeze the session. Once you have tested the settings you can either get MemSet to apply the settings for each session, or you can just apply the settings in BIOS. You can see the settings I ran with in the screen shot.
Other BIOS Settings
I’ve found that leaving CIE enabled does little to harm performance. This setting enables the Intel Speed Step function which allows the CPU to reduce from an 8x multiplier to a 6x multiplier when idling. Normally Speed Step would also reduce the CPU voltage, but this is disabled when using other than Auto in the BIOS for CPU voltage. As a result I’m not sure whether the savings are as great, but at least the CPU clock speed is reduced when the PC is sitting idling.
Quiet Computer
The best speed to run for quiet computing is 400 MHz FSB or 3.2 GHz CPU clock. Front and rear fans are at 1000 rpm, CPU fan is at 1700 rpm, GPU fan (Zalman) is inaudible on 5V. The two hard drives are in Scythe Quiet Drive enclosures and supported on foam rubber feet.
Photo 3
http://members.iinet.net.au/~gbyers/PCinside.jpg
Components:
Asus P5B Deluxe WiFi motherboard
Core 2 Duo E6400 CPU
Zalman CNPS 9500 LED heat and Arctic Silver AS5
Corsair 2x1GB XMS2 5400C4 DDR2 RAM
Leadtek 7600GT graphics card
Zalman V700 Cu graphics chip cooler
Creative X-Fi Platinum sound card
2x Seagate 320GB 7200.10 HDD in Scythe Quiet Drive HDD enclosures
Liteon 165P6 optical drive
2x 120mm Antec Tricool fans
Antec TruePower 2.0 430W PSU
Antec SLK3000B case