is running pci bus at 39mhz safe??

[jjyiz28]

i managed to raise my fsb from 133 to 154, but since i have no pci lock, it raised my pci to 39mhz. it runs fine while im playing games with it. but will it damage my components?? my audigy and modem is on pci slots.

went from 1.47 to 1.7 on cheapo msi km2m. =)

 

[boyRacer]

id be more worried about the harddrive since it could throw up and corrupt your data... i ran my old system at 160 before... but not too long. but some people have been running a 40MHz PCI for 2 years with not problem.

 

[jjyiz28]

drats, didn't even think about the HD sharing the pci bus. ahh.. ill let it go for now, see if i get probs later on. my data aint "that" important. =)

 

[Matt2]

I ran my old P3 800 and CUSL2 at 40mhz PCI BUS and I fried two different hard drives, a maxtor and an IBM. SO from my experiences I would say no.

 

[T9D]

..

 

[jjyiz28]

higher than 40mhz? i can't imagine, most high OC'ers have pci locks or pci dividers. so no worry.

 

[Sunner]

It might work, it might not work.

I used to run an old TX97(430TX chipset) at a 41.5 MHz PCI bus, worked fine for a few months, then all of a sudden my entire HD got corrupted, I couldn't even finish a format.
Set it back to 37.5 MHz PCI bus and it worked again.
After that, I try to stay away from raising the PCI bus too much.

 

[bacillus]

IMO anything above 38MHz is asking for trouble!

 

[Eug]

Anything above a 15% overclock of the PCI bus is foolish IMO. (15% is 38 MHz.) I think it's smarter to stick with a 0-10% overclock.

 

[bgeh]

Originally posted by: bacillus
IMO anything above 38MHz is asking for trouble!

me thinks the same too

 

[maluckey]

I'm the party crasher. I never, ever, ever, run below 43mhz on the PCI bus, and often go higher. This system has run 24 hours per day for eighteen months, in it's current configuration. I've had the FSB as high as 180mhz on a 1/4 divider, for 24 hours (45mhz PCI). It was the CPU that ran out of steam first!

To say a certain percentage above 33mhz is the limit isn't accurate. To say a certain frequency is the limit, isn't accurate. It's all about heat, cooling, and the quality of your components.

 

[Eug]

Originally posted by: maluckey
I'm the party crasher. I never, ever, ever, run below 43mhz on the PCI bus, and often go higher. This system has run 24 hours per day for eighteen months, in it's current configuration. I've had the FSB as high as 180mhz on a 1/4 divider, for 24 hours (45mhz PCI). It was the CPU that ran out of steam first!

To say a certain percentage above 33mhz is the limit isn't accurate. To say a certain frequency is the limit, isn't accurate. It's all about heat, cooling, and the quality of your components.

No, it's about risk takers and non-risk takers. Consider yourself lucky, and I hope you have a bazillion backups.

 

[Sid59]

I also think you are the exception, not the rule.

 

[Twista]

I have used 149/39 for my overclock with high Mutil. since i only have 2100ddr and it worked fine and my video card got higer fps since it was running faster with an overclock core also.

 

[Sunner]

Originally posted by: maluckey
I'm the party crasher. I never, ever, ever, run below 43mhz on the PCI bus, and often go higher. This system has run 24 hours per day for eighteen months, in it's current configuration. I've had the FSB as high as 180mhz on a 1/4 divider, for 24 hours (45mhz PCI). It was the CPU that ran out of steam first!

To say a certain percentage above 33mhz is the limit isn't accurate. To say a certain frequency is the limit, isn't accurate. It's all about heat, cooling, and the quality of your components.

I remember alot of 3Com cards had a very hard time coping with OC'd buses back when I worked at an OEM.
Many SCSI cards suffered from this too, both Asus and Adaptec.

As for harddrives, the one I used back then was a Fujitsu, and I know it's happened to others with IBM Deskstar's, don't know about Maxtor/WD.
All on Asus mobos by the way.

So, tell me, what exactly do you mean by "quality components"?
I suppose Asus, IBM, 3Com, and Adpatec are all out?

Just cause it works for you doesn't mean it will work for everyone, some people are lucky, some aren't, as is almost always the case with overclocking.

 

[Eug]

Originally posted by: Sunner

Originally posted by: maluckey
I'm the party crasher. I never, ever, ever, run below 43mhz on the PCI bus, and often go higher. This system has run 24 hours per day for eighteen months, in it's current configuration. I've had the FSB as high as 180mhz on a 1/4 divider, for 24 hours (45mhz PCI). It was the CPU that ran out of steam first!

To say a certain percentage above 33mhz is the limit isn't accurate. To say a certain frequency is the limit, isn't accurate. It's all about heat, cooling, and the quality of your components.

I remember alot of 3Com cards had a very hard time coping with OC'd buses back when I worked at an OEM.
Many SCSI cards suffered from this too, both Asus and Adaptec.

As for harddrives, the one I used back then was a Fujitsu, and I know it's happened to others with IBM Deskstar's, don't know about Maxtor/WD.
All on Asus mobos by the way.

So, tell me, what exactly do you mean by "quality components"?
I suppose Asus, IBM, 3Com, and Adpatec are all out?

Just cause it works for you doesn't mean it will work for everyone, some people are lucky, some aren't, as is almost always the case with overclocking.

Yep, I had similar problems with Maxtors on Asus mobos.

 

[ElFenix]

Originally posted by: maluckey
I'm the party crasher. I never, ever, ever, run below 43mhz on the PCI bus, and often go higher. This system has run 24 hours per day for eighteen months, in it's current configuration. I've had the FSB as high as 180mhz on a 1/4 divider, for 24 hours (45mhz PCI). It was the CPU that ran out of steam first!

To say a certain percentage above 33mhz is the limit isn't accurate. To say a certain frequency is the limit, isn't accurate. It's all about heat, cooling, and the quality of your components.

hard drive corruption seems to have everything to do with the PCI mhz. overclocking your pci bus certainly doesn't increase the heat, and there doesn't seem to be any correlation with the drive maker, either

 

[Ilmater]

Originally posted by: ElFenix

Originally posted by: maluckey
I'm the party crasher. I never, ever, ever, run below 43mhz on the PCI bus, and often go higher. This system has run 24 hours per day for eighteen months, in it's current configuration. I've had the FSB as high as 180mhz on a 1/4 divider, for 24 hours (45mhz PCI). It was the CPU that ran out of steam first!

To say a certain percentage above 33mhz is the limit isn't accurate. To say a certain frequency is the limit, isn't accurate. It's all about heat, cooling, and the quality of your components.

hard drive corruption seems to have everything to do with the PCI mhz. overclocking your pci bus certainly doesn't increase the heat, and there doesn't seem to be any correlation with the drive maker, either

You're right. This has nothing to do with the quality of your components. No hard drive manufacturer builds a drive that will operate correctly with a 34MHz+ PCI bus. When something goes wrong, don't blame WD, Maxtor, or Seagate, blame yourself. It's luck if you get a hard drive that happens to work well in that situation.

 

[maluckey]

TextThis has nothing to do with the quality of your components. No hard drive manufacturer builds a drive that will operate correctly with a 34MHz+ PCI bus

first let me say that your statement is a crock of Sh**! You know this how? You got your source of information from where? Go to the overclocking section on these very forums and say something like this and they will eat you alive!!! Stick to what you can prove, or state your source. Otherwise it's just opinion, assumptions, and/or B.S.

I've built six systems in two years, and not ONE has failed. I'm the luckiest guy in the world? I think not. I use common sense, imagination, and COOLING, as I stated before.

Raisng the PCI frequency causes heat....period. Raising the voltage causes heat.....period. Cool the drive, or any component, and extend it's life....period. It's not magic, or rocket science. Ask anyone on the overclocker board if they'd OC without good cooling. They won't, because you cook the goods.

I run 2 Maxtors, at over 43mhz PCI bus, since their purchase. No issues. Quality in memory, video cards, and others gets you a success, or a failure. If your stuff fails, it's not up to snuff for for overclocking, and either lower the speed, or find something that does overclock. It's that simple. Some brands are better than others.

I bought the Maxtors at two different stores, weeks apart. Not luck that they work, but cooling. I have 2, 80 mm fans on them at all times. Like I said. Quality and cooling are the ticket to lasting overclocking.

I keep bringing up overclocking, because the original post was about overclocking, so in that vein I stay.

Also remember that I didn't state that it won't damage your components, because, given time, it will. Most people will have upgraded to newer components by then, and the point would be moot. Who still uses a three year old video card???? If so, UT2k3 must suck for you! Who still uses their VX97 board?

 

[Sunner]

As I said, I used to work at an OEM, I've What Ilmater tried to say(I assume, Im sure he'll correct me if Im wrong) is that no manufacturer will guarrantee you that the component will work out of spec, that's why it's called out of spec.

If my AXP 2100+ won't do 3 GHz, is that due to bad components too?
Or if my harddrive craps out at 50 MHz PCI, is that due to bad cooling?

By the way, it's nice that you have those 80 mm fans on your HD's, but they don't get any hotter no matter how much you OC the PCI bus.
Heat production won't go up much at all for that matter, unless you have some wierd south bridge.

And yes, you are lucky, maybe not the luckiest guy in the world, but lucky.

 

[maluckey]

In this, we agree. There is no manufacturer that will warranty anything that they cannot account for, and Oc'ing is certainly one of those.

As far as overclocking, there are new components that for whatever reason, will not run, and old components that will not die. It is generally not possible to predict these failures, without research, and a bit of trial and error. What you can predict, is whether or not a certain brand tends to do better than another in these extreme environments. A few names do come to mind, and are not exclusive: Abit, Asus, Maxtor, Western Digital, Kingston, Crucial, Lian-Li, Alpha, Gainward, and many others. These brand are normally a notch above, and in most cases handle tha abuse very well.

To state that raising the operating frequency of an electronic component won't add heat, whether directly or not, is ........ not entirely accurate. When the bus throughput is raised, you ask all devices in the path do do operations per clock cycle much faster than they normally are asked to. They often loaf around at 75 percent of their maximum load, and then you ask them to operate at 90 percent, and expect no heat increase? I don't think so. I wish it were so, but even with a CPU, more mhz = more heat. That's why on some boards, you can change the CPU Fast Command Decode setting to accomodate this very phenomenon.

 

[Sunner]

Raising the PCI bus will of course generate more heat, but not by any significant amount, as compared to overclocking the CPU, or GFX card for example.
As for harddrives, they are a different animal compared to CPU's or GFX cards, in that there's not really a whole lot you can do do improve their tolerance, putting a high speed Delta fan on a HD won't ensure it won't get screwed up by too high PCI speeds.
Another difference is that if your GFX card won't handle the higher frequency, you can set it back a notch, no harm done, with a HD, you risk losing all your data.

 

[maluckey]

I mostly agree, and I hear what you are saying, in that hard drive data corruption can be a real bummer.

What I'll add, to this already jumbled mix, is that what many call hard drive corruption, is often caused by RAM Tras settings, which, can be controlled independent of the PCI bus on some modern mobos. It's often mistaken for HD data corruption, when it's registry corruption. High quality RAM, for OC'ers can often fix this problem. This is why many people have great success with raising the PCI bus.

Other issues that can play a big role, are the signal clarity, and the path length, to and from the PCI bus. This is part of the mobo design itself. High-end audio manufacturers write volumes on this subject, and it applies to computers as much, or more, once you are outside the consumer "specs".

So to end my babbling, I still feel that a successfull OC depends on the quality of components, cooling, and common sense of the OC'er.