• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

Supercharge your Mac! - Binning/Observations on certain chips

alectrona6400

Well-known member
As you may know (and as a small Vintage Apple server I'm in knows), I'm usually known for overclocking a 2005 eMac to 1.92GHz. However, that's not the only thing I've done, as I've switched up PLLs several times.

Before I start, if you do want to overclock, you might want to check your system just to see how much of a boost you can get.
Usually it's easily told by when your system was made; After the first 2 digits/characters in your serial number, there should be 3 numbers. The first of the 3 indicates the year in a single digit, and the last 2 indicate the week.
Example:
G8522 - Week 22, 2005
XB011 - Week 11, 2000
QT213 - Week 13, 2002

First off, early production and earlier steppings:
Usually your luck will be lower. That's no surprise, considering it took a while before most of the popular overclockers in the PC market got to fame. These are some observations I have made with early production units:
-2004 eMac, week 12 2004: PowerPC 7447 from the looks of it. Not an A, nor a B. This has its vcore set to 1.350V from what I have seen and can overclock up to 1.42GHz without a vcore increase. Not sure if you can even go higher on the vcore, as 1.42 seems to be the safe limit here. 1.5GHz started causing instabilities.
-2002 PowerMac G4 QuickSilver, dual 1GHz 7455 (not A, nor B. original stepping) - I've had a rough history with this, I'll admit. I do have 2 cards, a week 6 2002 and a week 2 2002. The former I, err, broke. Either it was the ESD from the various times I tested it, or the heat energy going through components it shouldn't have gone through. That one wasn't stable at 1.2 if I recall.
I got another dual 1GHz card, which is the latter. That got to 1.2GHz, but immediately kernel panicked. 1.13 chimed, yet there was no display (some clocks don't work!) but 1.07 was stable no problem. Not too much of an increase, but ah well.
The 7455s were each rated at 1000MHz (1GHz)
-Mid 2004 12" iBook G4, purple Motorola 7447A rated 1064MHz - Not particularly sure if it's earlier production, but it is an earlier stepping. Machine was made around week 17 2004, and only got up to 1.2GHz. 1.33GHz worked for several minutes until the system locked up. 1.266GHz chimed, but no display and the system locked up.

Chip clock ratings:
-2002 iMac G4 (week 6 and week 13 2002): The earlier unit is a 700MHz (now 800) with the motherboard P/N 820-1257-06 if I recall correctly. This booted at 900, but the system became unstable pretty quickly. 800MHz works no problem. Not sure about 850, as it wasn't listed on the site I used to find the PLL settings (thundernerd IIRC)
The latter, an 800MHz iMac, got to 900 happily and is stable. This could be related to early production and steppings, but the 7451 chips never really got touched again as the 7455 was its replacement (I do have a couple 7455As on the way, and one of my first BGA projects may involve swapping the chip out on the 700MHz board!) and as such, the factor I'm boiling down to is the chip's rated clock. 99% of the time these parts are rated at the computer's given clock speed (800MHz is 800, 700MHz is 700, 1.42GHz is 1420, yadayadayada) and these ratings mean better bins. Your headroom will be better with these chips, despite the maximum clock limits specified in the documentation of these chips.
-2000 PowerMac G4 (Gigabit Ethernet) - I've never had a card that was rated at 500MHz, but there have been 550MHz successes on 500MHz cards from what I've seen. The 450s can go to 500Mhz no problem, but not 550. This isn't vcore related. These use the original PowerPC 7400 chips, which were never particularly known for being good overclockers.

Late production units:
-2004 PowerBook G4 1.5/128MB VRAM (week 4 2005) - This has the PowerPC 7447B chip, which was a huge stepping improvement over the original 7447A and 7447 chips. This one got to 1.67GHz stable no problem, not even a vcore increase in sight. Not sure about 1.33GHz models, but it MAY work on those too. This was only a 117MHz bump on a chip that was already rated for 1.5GHz. The earlier 2004 PowerBooks might have been pushed to their limits, unfortunately.
-eMac (USB 2.0, week 23 2005) - this too uses the 7447B. This got to 1.67GHz from 1.25GHz without a vcore increase, and has been very stable. About a 417MHz bump!

Better bins in general:
-1999 iBook 300 > 433MHz (week 40 1999) - Not sure why this was possible, but the later 750L chips have much better headroom as opposed to the older 750 chips. Most people usually got to 400MHz, but I was lucky.
-800MHz single 7450 processor card from week 7 2002 - this got to 933MHz, but I haven't tested it enough to where I could call it stable.

Hypothetical successes:
-PowerBook G4 DLSD (week 4 2006) - One of the last units made. This might be a good bin, but I'm not too sure. You can overclock the bus on this to run as high as 200MHz, but I might just want to get a 7448 swapped in for that. These PowerBooks weren't exactly known for running cool. And yes, it DOES have a 7447B. Let this "7448 in a PowerBook" hoax go to rest, IT NEVER HAPPENED.
-Week 4 2002 iMac G3 running at 750MHz - Someone has apparently tried this to no success, but I'm thinking of trying it on a later G3 iMac board I have. Then of course, these weren't internally well designed either so I'm letting it stay at 700MHz from the original 600.

It's 8:00AM, and that's all I can think of at the top of my head. I might have written this like a stooge, but ah well.
Hopefully this helps. If you're willing to take on the task of OCing, let me know first and I'd be glad to help given that information about the computer overclock is available online.
 
Top