• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Quick Conroe stepping question

Z

ziphnor

Weaksauce
Joined
Jan 16, 2005
Messages
65
Hi,

I just go my E6600, its labelled 'L632A413' and the pack date is 10/06/06. Can any useful information be gathered from this(like for example its going to be a bad OC'er, etc)?
 
I bought it from a danish online shop called dataworld.dk, its in Roskilde, which would be local i guess ;)
 
CPU's usually get better as the manufacturing evolves.

Week 24 is one of the earliest chips and it might do just fine.

However, the newer batches have shown much better performance!
 
L624 is ok...you can do like 3ghz easy on low voltages but you will need nearly 1.5v just to do 3.2ghz.
 
I think the chip should be ok... however I like to keep the voltage under 1.45 just to be safe.
 
wow, week 32.... cross your fingers... you might have a great chip.

Thats the latest chip I've seen.

my week 24 only does 3.0Ghz reliably.
 
How are the weeks counted, starting at week one on January 2006? Because then it confuses me that a chip packaged 10/06/06 should be one of the newest ones. At first i thought i might have misread the label, but i double checked and it IS L632. So i guess the week count is based on something else?

Anyway, you can read about how im doing in this thread:
http://www.hardforum.com/showthread.php?t=1117306

Basicly im a bit puzzled by the vCore reported as opposed to what is set in the BIOS, but 3.2Ghz was stable enough for 2 hours of ORTHOS( stopped it myself) while CPU-Z and Asus PC-Probe was reporting ~1.288 as vCore( though it showed ~1.33 when i booted and then dropped under load and was set to 1.3875 in the BIOS ). It also did 3Ghz at 1.25 which is what i get when i set 1.35 in the BIOS....

Finally im getting very different temperature readings from CoreTemp and Asus PC-Probe(CoreTemp being ~10C lower).

Right now im running at stock speed though, until i figure out what the hell is going on...
 
Not sure on the weeks, but I just got a Week 31 that says it was packed Sept. 9.
 
ziphnor said:
How are the weeks counted, starting at week one on January 2006? Because then it confuses me that a chip packaged 10/06/06 should be one of the newest ones. At first i thought i might have misread the label, but i double checked and it IS L632. So i guess the week count is based on something else?
Click to expand...

Say week 32 / 4 = approx 8 months. So it was made in August some time, but its a long process from a large wafer of silicon, diced up to squares, putting it on the package, various testing at various stages, binning, to packaging in Sept/Oct.

Usually the voltage under load drops by about .05 its called vdroop. Look it up, its a very well known issue. Thats why you see that drop.

As for the temps, Core Temp usually shows about 10C+ higher than other windows temperature readings. Like SpeedFan, Asus Probe, etc. Are you sure its showing 10C lower?
 
chrisf6969 said:
Say week 32 / 4 = approx 8 months. So it was made in August some time, but its a long process from a large wafer of silicon, diced up to squares, putting it on the package, various testing at various stages, binning, to packaging in Sept/Oct.
Click to expand...

Now i get it, in denmark we write the date as day/month/year, but i seem to remember that in US/UK its written as month/day/year, that explains alot, as i was confused as to how it could be packed BEFORE being manufactured ;)

Usually the voltage under load drops by about .05 its called vdroop. Look it up, its a very well known issue. Thats why you see that drop.
Click to expand...

Is it also normal that vCore is much lower than set in the BIOS in general? Right now im running 390x9(well CPU-Z says 9x381.2Mhz) ~3.4 - 3.5Ghz(depending on which program you ask :) and the voltage in CPU-Z is 1.432, but was set over 1.5 in the BIOS.

I had tried that high an FSB earlier and failed, apparently because my RAM was trying to run at CL4(its CL5 DDR2-800), because as soon as soon as i set it to auto i could boot and ORTHOS away(21 min so far). The P5B-E bios seems to be missing settings for memory timings, so i cant set them manually (sucks) and for some reason it now has them at 5-6-6-18, even though the SPD says 5-5-5-18 for 400/800Mhz(and its only doing ~390Mz). Stupid bios :)

Guess i should try a lower vcore now that i know it was the RAM holding it back.

As for the temps, Core Temp usually shows about 10C+ higher than other windows temperature readings. Like SpeedFan, Asus Probe, etc. Are you sure its showing 10C lower?
Click to expand...

Im 100% sure, though it varies from 5C to 12C higher in PC-Probe. At my current speed CoreTemp says 59C under load(i think i put too much thermal paste on ;) but Asus PC-Probe is at 63-65C.
 
ziphnor said:
Im 100% sure, though it varies from 5C to 12C higher in PC-Probe. At my current speed CoreTemp says 59C under load(i think i put too much thermal paste on ;) but Asus PC-Probe is at 63-65C.
Click to expand...

CoreTemp is the correct one, as it reads the temperature inside the chip. Not sure what to say about PC-Probe, but motherboard temperatures have long been known to vary greatly.
 
ziphnor said:
Is it also normal that vCore is much lower than set in the BIOS in general?
Click to expand...

Not really depends on the motherboard, PSU, and overclock speed.
Usually the BIOS setting is close to the windows reading. (though slightly lower) If you have a weak PSU (supply), crappy motherboard (filtering phases), or super high clock (draw) it could be lower.

A normal example would be:

Set 1.45v in BIOS
reads 1.44v in windows
drops to 1.39v under load in windows (dual prime / orthos)
 
I suppose something is fishy then, because i get ~0.1V difference between BIOS and windows. My PSU (SeaSonic 430W) is supposed to be pretty decent so i dont think thats what causing the problem.

However, when i start in at 3.4-3.5Ghz with a reading of 1.44, the initial reading is much lower in CPU-Z, around 1.25 and then goes up to 1.44 when ORTHOS starts, i guess it must be some automatic voltage adjustment at work or is my computer completely crazy?

Btw, i also managed to boot at 3.6Ghz, which gives and initial vCore of 1.2-1.3 as well but ORTHOS gives an error immediatly.
 
In a forum such as this, is there ever really such a thing as "A quick CPU stepping question"? Or a least one that doesn't spawn a nice lengthy thread in answer? :D /threadjack
 
ziphnor said:
I suppose something is fishy then, because i get ~0.1V difference between BIOS and windows. My PSU (SeaSonic 430W) is supposed to be pretty decent so i dont think thats what causing the problem.

However, when i start in at 3.4-3.5Ghz with a reading of 1.44, the initial reading is much lower in CPU-Z, around 1.25 and then goes up to 1.44 when ORTHOS starts, i guess it must be some automatic voltage adjustment at work or is my computer completely crazy?

Btw, i also managed to boot at 3.6Ghz, which gives and initial vCore of 1.2-1.3 as well but ORTHOS gives an error immediatly.
Click to expand...

That might be a different issue. I've seen where when you set the voltage around 1.45-1.5v+ it gets reported in CPU-Z as 1.2_v

So maybe you're right on the virge of that wrong reporting... and when you're underload your actual voltage drops to under that spot, so it displays the correct reading of 1.4_v underload, but not underload it shows the (incorrect) 1.2_v reading
 
Hmm, i always seem to get the same vcore reported by Asus PC-Probe and CPU-Z(though i didnt check for the very high voltage).

I will check it to make sure. If both PC probe and CPU-Z reports the same then something fishy is going on.
 
ziphnor said:
Hmm, i always seem to get the same vcore reported by Asus PC-Probe and CPU-Z(though i didnt check for the very high voltage).

I will check it to make sure. If both PC probe and CPU-Z reports the same then something fishy is going on.
Click to expand...

Sorry, Its an error with any voltage monitoring program (not just CPU-Z) at the higher end 1.45-1.5v+, for some reason gets read as 1.23-128v usually.

Look at the overclocking database, and you'll see a ton of people with screenshots that say 1.23-1.28v, but then they state their actual voltage was set to 1.5v+- in BIOS, etc.
 
Okay, thats a bit scary. I guess ill just have to sample a few lower voltages to determine the difference between the BIOS and the actual voltage is and then based on this calculate what voltage im actually getting at higher OC's.

Apart from this i will try to disable all powersaving things in the BIOS, at least for a while. I would like to see 3.6Ghz stable ;) However, for day-to-day operation i only want 3.6 if i can get dynamic voltage and multiplier control working, so that when idle the multiplier drops to x6 and the voltage to something suitably low. Otherwise ill rather go for an undervolted 3.2Ghz.
 
I hate this voltage crap. If i boot at 9x400 = 3.6 with auto voltage i can run orthos for a few minutes, so it might not need that much more voltage, but i have no idea how much its actually getting fed because all the programs report 1.25 which is definitely wrong!

I tried manually setting 1.55, and this gave a lower temp and was less stable(orthos failed within seconds) so it must be setting more than this. However, im a bit afraid to go above this setting, even though it seems 1.65 actually corresponds to 1.55. How high can the E6600 go?

Still, im not even sure i want to go that high because booting at 366x9 = 3.3Ghz with auto voltage yields 1.4V in CPU-Z, and then RMClock works, ive currently used that to undervolt to 1.162 ( becomes 1.050 in CPU-Z, seems im just always off by 0.1V) and it seems to run 3.3Ghz at this voltage fine(12 minutes ORTHOS so far)
Nice combination of OC and undervolting and my RAM dont mind running at CL4 1:1 at this speed.

Going to try and see tomorrow if i can climb up to 3.4Ghz and still run at such a low voltage.
 
The returns of higer voltages are not worth the extra heat and reduced lifespan of the CPU itself. I've been playing with my OC a little bit last night, and it seems as going past 3.32GHZ (ORTHOS stable) requires way more voltage than my current 1.475v setting. To get to 3.4GHz (only extra 80MHz right.. :rolleyes: ) takes 1.55+ volts... That add's an extra 3~5deg C to my temps. I've noticed that if ORTHOS is gonna fail, it will fail at ~2m15sec first.... The next hurdle is at about 12min. Next one is at ~23min. The more voltage I give my CPU, the farther I can get, but 1.55V is quite "extreme" for my taste... :eek: So I'm happy with my current OC. BTW, the program for monitoring voltages and temps I use is called SpeedFan (goolge it). It's quite accurate, especially with voltages.
Good luck.
 
Mine does:
(CL4 is 4-4-4-12, CL5 is 5-6-6-18 because even though the SPD for 400Mhz is 5-5-5-18)
The first voltage given is the one shown in Speedfan after booting, the vdroop is 0.5V. The BIOS setting is typically +0.05v to +0.1V higher( dont know why).

366x9 ~ 3.29Ghz at 1.4(auto) CL4
375x9 ~ 3.37 at 1.46 (auto) CL5 ( tested 8 hours Orthos)
391x9 ~ 3.52 at 1.5 (manual, auto is 1.53, bios setting is 1.5625) CL5 (tested 1h 13m ORTHOS so far). Load temp is 54-55C according to CoreTemp.

I think an actual vCore of 1.50 is okay (even though its scary to set it to 1.5625 in the BIOS, but i just dont get vcore i set!), so i might just keep the 3.52 OC.

Turns out that i was wrong about RMClock, it can only adjust the multiplier, even when the vcore is set to auto. This is really annoying because it means the idle power consumption will be that much higher.
 
I dont get this BIOS. It keeps settings strange memory timings and no matter what i set manually it wont boot (i even tried all timings at max). To test the RAM i booted at 400x8 = 3.2 with timings set by SPD in the BIOS. This yields the following:

http://img64.imageshack.us/my.php?image=timingswc6.png

That is, CPU-Z says the SPD for 400Mhz is 5-5-5-18 but its set to 5-6-6-18 !?
 
ziphnor said:
I dont get this BIOS. It keeps settings strange memory timings and no matter what i set manually it wont boot (i even tried all timings at max). To test the RAM i booted at 400x8 = 3.2 with timings set by SPD in the BIOS. This yields the following:

http://img64.imageshack.us/my.php?image=timingswc6.png

That is, CPU-Z says the SPD for 400Mhz is 5-5-5-18 but its set to 5-6-6-18 !?
Click to expand...

I see thats what the stick in Slot # 1 says, but did you check the stick of ram in Slot # 3.... use the drop down menu. I've seen it happen MANY times before when a supposed "pair" of ram are programmed differently. So they BIOS uses the slower timings.

Asus BIOS's (ie: the engineers who wrote it) are smart. When you leave a setting to Auto, its set to usually a GOOD STABLE setting. Those settings change with the FSB, to try to keep your system stable. Sometimes all of the settings for ram on Auto are not even available on manual.

OR
Maybe Asus's BIOS is applying "the formula" normally you take your first 3 numbers and they should add up to the last #. So inversely they're taking the 2 most important settings CAS & TRAS & filling in the middle 2 #'s.

IE:
4+4+4=12 (setting 4-4-4-12, etc)
5+5+5=15
5+6+6=18

Since A-Data programmed Cas 5 & the Tras = 18, its filling in the other two numbers as 6's.
 
So, I just got my E6600 and the packaging says L629B359 Pack Date: 10/03/06. Does it mean it is the newer B stepping??
 
chrisf6969 said:
I see thats what the stick in Slot # 1 says, but did you check the stick of ram in Slot # 3.... use the drop down menu. I've seen it happen MANY times before when a supposed "pair" of ram are programmed differently. So they BIOS uses the slower timings.
Click to expand...

It happens to be called Slot #1 and Slot #2 on my PC, but they both list the same.

Maybe Asus's BIOS is applying "the formula" normally you take your first 3 numbers and they should add up to the last #. So inversely they're taking the 2 most important settings CAS & TRAS & filling in the middle 2 #'s.
....
Since A-Data programmed Cas 5 & the Tras = 18, its filling in the other two numbers as 6's.
Click to expand...

This certainly sounds plausible. However it does not explain why i cannot mimic the timings the BIOS normally sets based on the SPD. When i disable 'by SPD' i can manually set about 10 different settings,but no matter how relaxed i set them i cannot boot, and using 5-6-6-18 in this way also fails....
 
You must log in or register to reply here.
Back
Top