wallstreet overclock round two

14 posts / 0 new
Last post
westieg3's picture
Offline
Last seen: 11 years 6 months ago
Joined: Apr 4 2004 - 18:03
Posts: 465
wallstreet overclock round two

this has been discussed on here before due to me, i am aware. however, i know that the wallstreet processor can safely be taken to 292 mhz with the 83 mhz bus (next model speed up from my 250mhz model) and then all the way up to 300 mhz with the 66 mhz bus. assuming i could safely take the bux to 66 mhz without throwing anything off, could it be POSSIBLE for the computer to be clocked up to 400 mhz? i only ask because the lombard is 400 mhz with a 66 mhz bus, and it has the powerpc 750 exactly like the chip in the wallstreet. i want to test this, but all i need is the proper heatsink for the computer. if such an overclock were possible then, what heatsink? pismo one won't fit i don't think, but a heat pipe model seems best.

coius's picture
Offline
Last seen: 5 years 8 months ago
Joined: Aug 25 2004 - 13:56
Posts: 1975
problem is the L2 cache

with that speed, the L2 cache will be too taxed, and I doubt it will start. The L2 cache will only overclock so far, and I think that is pushing it. I know that on my B&W G3 (with a huge heatsink) refuses to boot, or freezes when I push the L2 cache to 1:2 when running my machine at 450MHz. It just doesn't seem to like it.

westieg3's picture
Offline
Last seen: 11 years 6 months ago
Joined: Apr 4 2004 - 18:03
Posts: 465
that was my fear. for testing

that was my fear. for testing, i have a heatsink sitting around that should be plenty enoughn for heat not to be a problem. so 292 to start (obviously safe), then i guess ill try a different bus speed and varios cpu speeds from there. hmm...what about installing dip switches in place of the resistors at some point? a lot less to break when switching the clock speed.

alk
alk's picture
Offline
Last seen: 2 years 5 months ago
Joined: Dec 20 2003 - 10:38
Posts: 373
I think you're missing something

Why do you think that 292 is "obviously safe?" Your CPU is rated to 250 MHz, is it not? That means that 292 is not obviously safe and may, infact, be harmful. And even if your CPU can get to 292 without overheating, that doesn't mean that the L2 will be tolerant of the increase in speed/heat.

Just because the Lombard uses a "PowerPC 750" of the same model number, that doesn't mean that your particular example (rated at 250 MHz) will clock to anywhere near 400 MHz. In fact, your particular example isn't guaranteed to run at any speed other than 250 MHz.

That said, give it a try. Don't be too upset if it doesn't work (or worse, fries). Wallstreet CPUs are pretty cheap.

Peace,
Drew

westieg3's picture
Offline
Last seen: 11 years 6 months ago
Joined: Apr 4 2004 - 18:03
Posts: 465
trust me, i know it can go to

trust me, i know it can go to 292. that would be the same speed in the next model up, and i would think that apple just clocked it differently. at least in testing, i should be able to keep it cool enough.

westieg3's picture
Offline
Last seen: 11 years 6 months ago
Joined: Apr 4 2004 - 18:03
Posts: 465
trust me, i know it can go to

trust me, i know it can go to 292. that would be the same speed in the next model up, and i would think that apple just clocked it differently. at least in testing, i should be able to keep it cool enough.

alk
alk's picture
Offline
Last seen: 2 years 5 months ago
Joined: Dec 20 2003 - 10:38
Posts: 373
I doubt Apple just clocked it

I doubt Apple just clocked it differently. It wouldn't make sense in terms of component price to buy a 300 MHz part and clock it to 250 MHz. (Although Apple has done that in th past on PowerBooks like the 1400, it's pretty rare.) The part in your 250 MHz Wallstreet is almost certainly ONLY RATED TO 250 MHz. Of course, you can prove me wrong by cracking it open and examining the actual part. Many of them will have the clock speed silk screened onto the ceramics or etched into the metalic heat conductor over the die.

Achieving 292 MHz may be possible, but I wouldn't count on much more than that. That's a 20 % boost, and most overclocking fails between 20 and 30 %. Of course, you aren't guaranteed to get even that 20 %, but something like 80 % (or more) can be overclocked by 20 % just due to the manufacturer's QA when setting the qualified speed for a batch of CPUs.

Peace,
Drew

coius's picture
Offline
Last seen: 5 years 8 months ago
Joined: Aug 25 2004 - 13:56
Posts: 1975
Re: problem is the L2 cache

coius wrote:
with that speed, the L2 cache will be too taxed, and I doubt it will start. The L2 cache will only overclock so far, and I think that is pushing it. I know that on my B&W G3 (with a huge heatsink) refuses to boot, or freezes when I push the L2 cache to 1:2 when running my machine at 450MHz. It just doesn't seem to like it.

I wish to make a correction. I meant to say that when I push the L2 Cache to 2:3, it gives me problems. not 1:2

westieg3's picture
Offline
Last seen: 11 years 6 months ago
Joined: Apr 4 2004 - 18:03
Posts: 465
now my patience just ran thin

now my patience just ran thin. screw the bloody upgrade!

alk
alk's picture
Offline
Last seen: 2 years 5 months ago
Joined: Dec 20 2003 - 10:38
Posts: 373
Huh?

"westieg3" wrote:
now my patience just ran thin. screw the bloody upgrade!

If you're waiting for somebody to tell you "Good boy, have a cookie," well, keep waiting. You've been talking about this since July, so why don't you just do it already and stop asking questions and getting upset when you don't get the responses you want to hear?

Seriously, no one is saying not to do it, we're just saying you're not likely to get the speed you want!

But hey, like I've said, Wallstreet parts are cheap. Give it a try. Worst case is that you're out a CPU module.

Peace,
Drew

westieg3's picture
Offline
Last seen: 11 years 6 months ago
Joined: Apr 4 2004 - 18:03
Posts: 465
ugh nevermind. i'm just aggri

ugh nevermind. i'm just aggrivated all my ideas that seem possible always have some loophole to them that makes them ridiculous. im sure youll throw out my concept of figuring out the manufacturer of the drive and switching the parts in that too? ill reserve my overactive imagination to my automotive designing.

and i have plenty of cookies.

Jon
Jon's picture
Offline
Last seen: 8 years 5 months ago
Joined: Dec 20 2003 - 10:38
Posts: 2804
There is sometimes no issue s

There is sometimes no issue swapping drive controller boards between identical drives.

As for your over clock idea, you seemed to "want" to ignore the fact that the CPU in our WallStreet is only rated at the speed it runs at. Just because there is a faster model of the machine made, it has no bearing on the fact that your module might or might not run at the same speed. The module in your machine is there, and at that speed, for a reason. It's very rare to buy faster parts and down-clock them. The profit margin is better to just sell them at their rated speeds.

Having ideas and asking questions is great. Ignoring or getting frustrated for answers to those questions because they aren't want you want to hear tells me that you probably didn't need to ask the question anyway.

To put it another context, can you run your car engine at 6000 RPM? Sure. Can you run it at 8000 RPM? Sure. Will it last if you let it run at 8000 RPM for any reasonable length of time? It's highly dependent on the specifics of the particular engine in question. Do race cars run at RPMs higher than that? Yes, but they go through engines like crazy, compared to a mass market car, because they just simply don't last above a certain guaranteed RPM rate. Same thing with your CPU. It's physics, sometimes things don't work. You gotta just deal with it. Wink

alk
alk's picture
Offline
Last seen: 2 years 5 months ago
Joined: Dec 20 2003 - 10:38
Posts: 373
Hmm.

I'm not normally the kind to be super supportive, so this is the only time I'll say this.

Clearly you're excited about the prospects of overclocking your PowerBook, but you want some assurances that the work you do will be worthwhile. We can't give you those assurances. Whatever reason you have for wanting them cannot be addressed by us if you don't tell us what they are. If you're nervous about the microsoldering involved, then that's perfectly justifiable: it's tough to do, but not impossible. If you're nervous about frying a CPU module, I've already indicated that it's no big deal if you do - you can find some cheap! Etc.

Here's my advice. DO THE HACK. Just do it. Do it do it do it do it. I'm dead serious. I don't think you'll get the speed you want, but you'll never know until you do it. I'm also serious about Wallstreet parts being cheap. They are. They're dirt cheap. Hell, buy yourself a spare CPU right now and OC that one instead so you're not messing up your stock config!

If you don't have the perseverence to follow through when people (only two on this thread) don't think you'll be able to do what you want, then life will be very hard for you. Try not to get discouraged or upset, but deal with the feedback in a positive manner. So what if I don't think it will work? Do it anyway or change your approach to find a different solution that works for you, but don't just give up.

Is there any particular advice you are looking for other than a feel for whether or not it will work?

Good luck,
Drew

westieg3's picture
Offline
Last seen: 11 years 6 months ago
Joined: Apr 4 2004 - 18:03
Posts: 465
what you assumined i'd clock

what you assumined i'd clock it up to 400 immediately? no. that i wouldn't do. 292 seems relatively safe, and its possible i could push it a bit more if i could cool it really well. such cooling means bye-bye modem though (who cares when you have a cisco wireless card on hand). i'll try 292 as soon as i find a desoldering braid. at least that will be a low enough speed to not fry, but quick enough that i will notice it.

Log in or register to post comments