That's why I've switched to @wicknix 's IceWeasel; fantastic difference, even on a Pismo
Cheers
Hugh
That is for Linux isn't it?
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Problem with overclocking PowerBook G4 Hi-Res
- Thread starter JoyBed
- Start date
- Sort by reaction score
That is for Linux isn't it?
Not this version - [URL="https://forums.macrumors.com/threads/so-this-finally-happened-sort-of.2172031/page-3#post-27366588"]So this finally happened. Sort of... 
[/URL]
Cheers
Hugh
[/URL]Cheers
Hugh
Didn't see that. Downloaded and typing while using it. Though I would be worried about security. TFF is about to get a Chemspill because of a potential threat.
Tell me more please...
Cheers
Hugh
http://tenfourfox.blogspot.com/2019/06/stand-by-for-fpr14-spr1-chemspill.html
It’s a disappointment that Geekbench is incorrectly displaying the clock speed for your device, as you did not run GB2 with a 1667MHz system. As far as I’m aware, this score remains the highest posted stock 17" DLSD PowerBook 5,9 on GB2.
So then I am 46 points ahead of the highest stock dlsd. Now just to wait for the ssd and new ram and it will be higher(i hope at least).
Which is fine, but at some point you’ll need to tweak the OF configuration to correctly reflect the clock speed of your system.
Also, an SSD won’t affect Geekbench’s scoring matrix. It will, however, favourably affect your Xbench scores.
(Incidentally, I’m the owner of said highest-posted stock DLSD 17". I'm fortunate the system was in excellent condition when I found it for, like, $65, and also, that it was produced at the very end of the production run might play a factor in the quality control of the internal parts being fairly good, given that any obstacles to assembly would have been worked out by then.)
Why not try my 'real world' benchmark and see how it compares with other machines?
https://forums.macrumors.com/threads/diy-benchmark.1862039/
That's 17:20 - my old 15" DLSD did 10:07 and my old 1.25Ghz Powerbook did 12:05 - you didn't have anything else running did you? Those temperature gauges will steal some CPU cycles?
Going to do the test on the DLSD I have now.
[doublepost=1561190484][/doublepost]On my current 15" DLSD score is 15:09 - full 5 min longer than previously. If you check the original thread, I think the key is the Quicktime version, 7.7 seems to be far slower than 7.5.5.
I had running the keyboard backlight adjust program which as I now checked takes up to 2-3% of the CPU and the gauges are consuming from the CPU too but I didnt checked that because im not at home right now, so maybe I am stealing somewhere about 5-6% of the CPU and also as you mentioned I have the Quicktime 7.7 and not the 7.5.5 so when I get home I will disable the gauges and the keyboard backlight program, uninstall the qt 7.7 and install 7.5.5 and rerun the tests.
Proceed at your own risk - I've just tried that using Pacifist - it crashed the whole system, now having to do a fresh install!
EDIT: Update on that - my HDD died in the process
Which I then replaced and when I did....the DVD drive stopped working :/
Last edited:
Man that sucks... Anyway I downgraded the QT using Pacifist to 7.5.5 from 7.7 and also disabled the gauges and backlight control program. Here are the new results from Xbench, GeekBench and your style benchmark:Proceed at your own risk - I've just tried that using Pacifist - it crashed the whole system, now having to do a fresh install!
EDIT: Update on that - my HDD died in the process
Which I then replaced and when I did....the DVD drive stopped working :/
If this is to be a standard benchmark metric on par — or even beyond — something like a Geekbench test, it is more appropriate, if not also accurate, to run these tests with the latest/last of everything for a particular OS X build, even if this means it may not be the fastest way to eke an arbitrary score/time. The last/final updates for everything within an OS X build is a threshold which is more likely to crop up with most systems/users.
To undo QuickTime 7.7.0 and to install 7.5.5 for the express purpose of obtaining a “fastest” score is, well, a gaming of the very benchmark which no longer has much to do with the hardware running the benchmark tests.
The rationale was that in my original benchmark thread all my results were done using 7.5.5 - as back then I never felt the need to upgrade beyond it, so any comparative hardware tests are useless using 7.7
Note, on a fresh Leopard install, after running Software Update, Quicktime was still left at 7.3 - which benchmarked the same as 7.5.5, so in this scenario, the "most systems/users" would be using the older version without intervention.
Of course, we now know 7.7 is a real benefit to HTML5 playback however.
So 1 minute faster than stock. But hey, I done the overclock in like 15 minutes in my free time so... I gained something at least. Its still just 7447A and not anything newer to be amazingly better and fast even with overclock.
Yes, it's obviously a beneficial speed hike - I'd be great to see a side by side comparison video with a stock machine...if you ever get another one
I am thinking to get a PowerMac G4 besides the quad G5 I am searching for, and test the 7448 swap on the PowerMac G4. Which brings me to the question, is there a dual PowerMac G4 ? So I can swap both CPUs to 7448?Yes, it's obviously a beneficial speed hike - I'd be great to see a side by side comparison video with a stock machine...if you ever get another one
EDIT: Meanwhile I will continue trying to overclock the PowerBook G4 DLSD to full 2GHz(with 200MHz bus and CPU multiplier intact) and somehow lowering the vCore to not get thermals to skyrocket to something higher than 70-71 degrees celsius.
Most versions of the Power Mac G4 came in dual-processor variants, exceptions being the first two (Yikes! and Sawtooth; but later revisions of the Sawtooth can take dual processors). There were a couple of 7448-based commercial CPU upgrades for Power Macs, and I think one of those was a dual. In any event, you'll almost certainly have to apply one of the proprietary firmware patches that came with these commercial upgrades in order to get a 7448 going on a PMG4. Search around the forum a bit, I know I've seen it discussed here.
Today I opened my DLSD and I set it to 189MHz bus so the cpu clock is 1,89GHz and not 194MHz bus and 1,94GHz cpu. So the results I posted are with the bus on 189MHz and CPU on 1,89GHz. Here are the BOM resistors:
So sorry for wrong info but that speed bump is good for the clock I am running, but I am worried for the thermals if I go higher.
So sorry for wrong info but that speed bump is good for the clock I am running, but I am worried for the thermals if I go higher.
Thanks for the correction, and especially the close-up photo for anyone who might want to try this!
That being said, how about an update? Has your PowerBook been stable? Have you used it much? Is there any specific indication that the speed boost presents a thermal problem?
I am using it on daily basis, mainly for web browsing and sometimes for games(Minecraft and Ghost recon), its stable, no lags, no system or app crashing, whole powerbook is slightly quicker. And the last thing you asked are thermals... Well the stock cooling is handling it pretty well. No shutdowns from overheating, the fans are kicking in more often but nothing horrible.