My 400's overclock poorly & new usage for my BP6

Batch codes, RAM specs, BIOS settings, etc..
Post Reply
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

My 400's overclock poorly & new usage for my BP6

Post by purrkur »

I replaced my 366's with my 400's to see how well they would overclock. I can have them running stable at 80MHz fsb but not any higher than that. I tried to run them at 500MHz (83MHz fsb) and the board boots and runs but it will lock up after awhile.

So much for that experiment. I would like to say that I think this is also my motherboard restricting me and not the CPU's. I have had my 366's running at 550MHz on another BP6 without any issues. My current BP6 has been modified with new caps but no active components have been replaced (which would probably give me better overclockability).

I have also seen some curious behaviour on this board. When overclocking, the board will do fine over a working session but when I shut it down the last thing that happens before it turns off is that the kernel will do a small core dump for some unexplained reason. This for me is a bit of a warning sign.

I love to tinker with my BP6's but the time I have to tinker is pretty much nonexistent these days. What I have been thinking about doing is putting in a couple of 533's on the board, running them at stock speed and giving my faithful BP6 server duty here at home (as my internal server). The server duty I have for it here at home isn't dependent on disk speed like a desktop would be but dual 533's would be good (I got a single 533 Celeron in my home server at the moment). This way I can run and enjoy my BP6 always! Right now the system spends most of it's time in pieces and collecting dust because I am constantly trying to find the time to try out new things. However, the time I get for playing around are few and far between :(
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Dave Rave
G'Day Mate!
Posts: 894
Joined: Thu Sep 26, 2002 4:28 am
Location: Sydney Australia
Contact:

Post by Dave Rave »

it's heat
or, uhm,,, the third thing.

gotta keep 'em cool
and the voltage has to go up so they work.
bp6's 3 x dual @ 533
. . . . 1 x dual @ 466
. . . . 1 sngl @ 400
[( 2 x dual xeon 2.4ghz )]
[( 2 x dual xeon 2.66ghz )]
[( 1 x 2.4C ghz )]
[( 1 x 2.4B ghz )]
[( 1 x dual AMD 1800MP )]
[( 1 x P4 1600 )]
[( 1 x 500 ppga )]
3 x piii 866
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Dave Rave wrote:it's heat
or, uhm,,, the third thing.
:) Yeah. heat I've got under control with massive heatsinks and 80mm fans. I've even done what everybody here says is a necessity, that is a fan on the BX. At idle the heat measures 27C and during desktop usage they go up to maybe 30-31C.
Dave Rave wrote:gotta keep 'em cool
and the voltage has to go up so they work.
I have tried up to 2.1V on the CPU's. it works better but it is not fully stable though. I have also monitored my voltages (I've got a good 350W PS) and they are very much stable.

This is the reason why I think it would be better to keep voltages and speeds at stock ratings, run the board and get a good life from it instead of trying to find some imaginary time to tinker with it some more.
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Dave Rave
G'Day Mate!
Posts: 894
Joined: Thu Sep 26, 2002 4:28 am
Location: Sydney Australia
Contact:

Post by Dave Rave »

if your standard room temperature is only 30-31, lucky you.
mine run at 45-46
if it was cool, they'd be at 78 or 80fsb but I run them std cos it's summer time.
bp6's 3 x dual @ 533
. . . . 1 x dual @ 466
. . . . 1 sngl @ 400
[( 2 x dual xeon 2.4ghz )]
[( 2 x dual xeon 2.66ghz )]
[( 1 x 2.4C ghz )]
[( 1 x 2.4B ghz )]
[( 1 x dual AMD 1800MP )]
[( 1 x P4 1600 )]
[( 1 x 500 ppga )]
3 x piii 866
davd_bob
Confused
Posts: 1043
Joined: Fri Feb 13, 2004 2:30 am
Location: Houston, TX

Post by davd_bob »

Im with ya on setting it up as dual 533s working faithfully and dependably as a server. I hate to think of retiring mine unless I 'mentor someone in OCing...cause the BP6 is the KING in the lessons in that arena.
Jordan
Posts: 64
Joined: Sun Dec 11, 2005 1:15 pm
Location: North Vancouver, BC

Post by Jordan »

purrkur, you put on a different chipset cooler; what is it? With all the USB trouble i'm having I wonder if I should get a better cooler...

Also, with your unexplained core dumps - windows XP will frequently do similar things for me! After lots of trouble-free reboots it will occasionally throw several "windows has recovered from a serious error" messages for no reason. weird...
The collector of the old and unwanted.
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Jordan wrote:purrkur, you put on a different chipset cooler; what is it? With all the USB trouble i'm having I wonder if I should get a better cooler...
Seems many people are having issues with USB on the BP6. I haven't used it at all to be honest with you so I wouldn't know. What I did was to buy a 40x40mm fan for the greenie sitting on the BX. Make sure that you put some good thermal compound between the greenie and the BX! I my tests I would say that the most significant gain you get when improving BX cooling is a thermal compound.
Jordan wrote:Also, with your unexplained core dumps - windows XP will frequently do similar things for me! After lots of trouble-free reboots it will occasionally throw several "windows has recovered from a serious error" messages for no reason. weird...
This one is a bit weird because it doesn't happen during regular runs but only when I shut the system down. It even happens after I get the "shutdown system" message which means that everything is off really, but then the kernel dumps this small trace onto the screen. I haven't really looked into it because stability hasn't suffered. It could be that this is a new kernel feature since I am running a bleeding edge Linux kernel.

As for your error messages, such errors could be because you are overclocking and your system is living on the edge, or because of heat. If you got a radical OC going on then decrease that OC a bit and see if that makes the problem go away. Adding more cooling features might not be enough, although it is hard to say. Time, trial & error are your friends :)
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

davd_bob wrote:Im with ya on setting it up as dual 533s working faithfully and dependably as a server.
Thanks for your input. I am actually warming up to the idea of using the machine as a server and liking it! I am seeing some possible usage that I haven't had the chance to think of before because my servers have been lacking on CPU power and specifically memory.
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Jordan
Posts: 64
Joined: Sun Dec 11, 2005 1:15 pm
Location: North Vancouver, BC

Post by Jordan »

Well funny you should mention changing the overclock...
I tried lowering the core voltages down from 2.2 volts to 2.1 and BOOM USB is working a bit more reliably. Usually playing Quake III for more than a while will kill all inputs but it didn't this time! I know it's only one test in a situation that warrants many, but it used to always crash at 2.1 volts before so YAY!

If i still have problems then i'll probably slow down the overclock and see if that fixes it. I'm amazed with the whole burn in thing. This machine keeps changing with time!
The collector of the old and unwanted.
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

Interesting. Now we got at least one reason why USB won't work reliably on the BP6!

As for the burn-in thing, I don't really buy that theory. When you overclock a CPU, what happens is that you get a higher risk of transistor "leakage" inside the CPU. What some people refer to as "burn-in" won't harden the transistors inside the CPU. You are however, increasing pressure on the individual transistors making them more sensitive to leakage, and most probably even reducing their lifetime with higher wear and tear.

For those who want to argue with me on this point consider this: If burn-in was scientifically correct and a proven method for "hardening" CPU's, wouldn't you think that CPU makers would be running their CPU's beyond their specification for a while before shipping them to customers? Surely it would make a better product?

Also, I have never seen anything on "burn-in" come from reliable sources, only overclockers who so valiantly would like to believe that they are actually doing their system a whole lot of good by overclocking their system :)
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
Jordan
Posts: 64
Joined: Sun Dec 11, 2005 1:15 pm
Location: North Vancouver, BC

Post by Jordan »

I forgot to mention this: I have actually gone overboard and replaced the greenie entirely with a massive cut-down Slot-1 Celeron cooler, lovingly modified to fit onto the board with the little pin holes for the greenie. that new thing is stuck onto the chip with some arctic silver 5, so it really surprises me i still have problems... I wonder if a retail cooler would be better of if the BX be lapped in some way maybe?

Also, scary thing about that transistor leakage. Oh well, by the time my CPUs start bleeding internally i'll have retired my board so i'm not that concerned. Thanks for the warning, though! I had completely forgotten about that. Judging from what you said, maybe my CPUs needed to loosen up!? :wink:
The collector of the old and unwanted.
purrkur
Linux Guru
Posts: 687
Joined: Fri Dec 12, 2003 5:57 pm
Location: Sweden
Contact:

Post by purrkur »

I have removed my old home-brew intranet server and replaced it with my BP6 running 2x533MHz procs at "turbo speed" (544MHz) and 640MB memory. The box it is sitting is is pretty sizeable which means I won't have any issues adding new drives in the future when needed and it is being fed by a 350W power supply. Only cards in it is the VGA card (an old Nvidia, fanless something or another which is not being used since the machine is headless) and a NIC. I am not running the largest coolers I have but smaller jobs with 60x60mm fans running on them. The fans are throttled down and now running at 2900rpm which is more then enough to keep the cpu idle speed at around 30C. The operating system is Debian stable with a 2.6.8 SMP kernel. The machine will now be running 24/7.

I guess it is time to update my sig one last time...
2x533MHz@544MHz, 2.0V
640MB PC100 memory
Realtek RTL-8139 NIC
Maxtor 6Y080L0 80GB hdd
Debian Linux stable with 2.4.8 kernel
g0fvt
Posts: 110
Joined: Mon Jul 29, 2002 3:32 pm
Location: Hertfordshire UK

Post by g0fvt »

My original BP6 which uses a pair of 466s seems to overclock better with old age, once I changed the capacitors the machine reached a level of stability way above what it did before. Some RF devices are known to improve a little with use, I suspect that whatever processes are changing the structure within the CPUs are ultimately destructive however I have never had an overclocked CPU die yet, one of my elderly BP6s is still running overclocked 24/7,
the machine has no video card or sound card, took them out after installing OpenBsd where it runs as my home file server and webserver...
Post Reply