Boosting an Old Power Mac with NewerTech's 1.8 GHz G4 Upgrade
The "Mystic" Power Mac G4 was the first dual-processor G4 and shipped in 450 MHz and 500 MHz versions. Each CPU has a 1 MB backside cache running at 1/2 of CPU speed, and that sits on a 100 MHz system bus. We're comparing that to the NewerTech MAXPower upgrade with a single Freescale 7448 1.8 GHz G4 CPU. These CPUs have an onboard 1 MB level 2 cache that runs at full CPU speed - that's four times as fast as the cache in the stock configuration.
The upgrade is theoretically capable of 4x the performance of the stock 450 MHz CPU under the classic Mac OS, which can only use one CPU, and twice as powerful under Mac OS X, which utilizes both processors. We ran an extensive suite of benchmark tests under Mac OS 9.1, 10.2, 10.3, and 10.4 to measure performance.
My primary computer is a "Mirror Drive Door" Power Mac G4 with dual 1 GHz CPUs, 2 GB of RAM, a 400 GB 7200 rpm Deskstar hard drive, and its stock ATI Radeon 9000 Pro video card. Subjectively, the 1.8 GHz upgrade in the Mystic feels every bit as fast for everyday tasks, and our Xbench test results (below) verify that.
The MAXPower G4 upgrades are compatible with Sawtooth, Mystic, Digital Audio, Quicksilver, and Quicksilver 2002 G4 Power Macs. They require a firmware upgrade to be installed before you upgrade your CPU, and this software is conveniently provided on a CD. (You'll also need to use this CD to uninstall the firmware patch to go back to the stock CPU.) The CPU upgrade works with Mac OS 9.2.2 and OS X 10.3.5 and later. I can also report that it works with OS X 10.2.8, which I have installed on one partition of my hard drive. It is available in speeds of 1.7, 1.8, and 2.0 GHz with a single CPU and 1.7 and 1.8 GHz with dual processors. There are also less expensive and slower models that use the 7447A CPU.
Our testbed computer has an 80 GB 7200 rpm Deskstar hard drive, 768 MB of RAM, and it's stock ATI Rage 128 Pro AGP 2x video card. The drive has separate partitions for Mac OS X 10.2 "Panther", 10.3 "Jaguar", and 10.4 "Tiger". Each has been updated to the latest version of that OS, and the only additional software we installed prior to benchmarking are System 9.2.2 for Classic Mode and the benchmark programs we ran.
The computer is booted, and then the benchmark apps are run in order.
Installation of the processor upgrade isn't difficult, and NewerTech even provides a Phillips screwdriver in the box. With my Mystic, I had to unclip the heat sink and remove 3 screws to take out the original CPU card. The NewerTech upgrade is fully integrated; the heat sink isn't separate from the CPU card. The upgrade clips into place, but putting the retaining 3 screws in is difficult if you don't have a magnetized Phillips screwdriver. Without one, it took some doing to get two of the screws in place, since they go in holes in the heat sink.
Once the upgrade is installed, there are two more steps: You need to attach the second cooling fan, which covers one of the screw holes, and then connect the CPU upgrade to your Power Mac's power supply, as the CPU and fans draw more power than the CPU socket provides.
Then came my only disappointment: I hadn't installed the required firmware update before installing the CPU upgrade. Back to the old CPU, insert the firmware CD, hold down the programmer switch, press the power button, release the programmer switch after the tone, and then hold down the C key to boot from the CD. It boots right into the software that lets you change the firmware. Very well done, NewerTech!
Then repeat the installation process. If I'd read the instruction manual first....
Anyhow, the system boots quickly into Mac OS 9.1, 10.2, 10.3, and 10.4. I don't have a 9.2.2 installation that works with Mystic, although I do have 9.2.2 setups that work just fine in Classic Mode. It looks like I'll have to invest in a full-fledge Mac OS 9.2.2 install CD so I can boot the Mystic and MDD into OS 9.2.2.
The dual 450 MHz Mystic isn't a fast computer under OS X, but it's usable. Drop in the 1.8 GHz upgrade, though, and that changes. Safari and Camino load quickly, YouTube videos run smoothly, windows fly open, and it easily feels twice as fast. That puts in on par with my primary computer, a dual 1 GHz G4, which is very impressive for an old computer and a single processor, albeit a very fast processor.
If you have an older AGP Power Mac that you're pretty happy with but lacks horsepower, a "brain transplant" is definitely worth considering. If this were my primary computer, I wouldn't hesitate to drop in a faster CPU card.
The big question is the value of the upgrade: Is it worth $200 or $700 - or somewhere in between - to upgrade your older G4 Power Mac? The MAXPower G4/7448 1.8 GHz Processor Upgrade sells for $325, more than this computer cost me (used). It comes down a question of how much it's going to improve your productivity.
NewerTech has a wide-ranging line of G4 upgrades using single and dual 7447A (512 MB onboard cache) and 7448 (1 MB cache) CPUs (prices are from Other World Computing, which owns NewerTech):
- 1.6 GHz 7447A, $230
- 1.7 GHz 7448, $320
- 1.8 GHz 7448, $325
- 2.0 GHz 7448, $400
- 1.6 GHz dual 7447A, $400 (Xbench 1.1.3 = 175.0)
- 1.7 GHz dual 7447A, $425 (Xbench 1.1.3 = 203.8)
- 1.7 GHz dual 7448, $550
- 1.8 GHz dual 7447A, $550 (Xbench 1.1.3 = 216.6)
- 1.8 GHz dual 7448, $699
As fast as this single 1.8 GHz CPU is, I have to wonder how much more power a dual 1.6 GHz upgrade would provide. Too bad NewerTech doesn't make a dual 7448 upgrade for my Mirror Drive Door.
Classic Mac OS Benchmarks
This is one old benchmark program, and it's reference score is the 8 MHz Mac Classic from 1990. The disk cache is 8 MB, and virtual memory is enabled. The display was set to 1024 x 768 with thousands of colors. Here are benchmark results with the stock 450 MHz dual processors (the Classic Mac OS only uses one) and the 1.8 GHz upgrade under Mac OS 9.1:
450 MHz 1.8 GHz Overall 217.3 721.3 CPU 219.0 876.0 Graphics 59.8 105.5 Disk 48.9 84.3 Math 1019.3 3223.6
Overall the 1.8 GHz G4 is 3.3x as fast, and the CPU test comes in at precisely 4x as fast with Speedometer 3 and Mac OS 9.1.
For Speedometer 4, the reference machine is the Quadra 605 from 1993. The disk cache is 8 MB and virtual memory is enabled. The display was set to 1024 x 768 with thousands of colors. Here are benchmark scores for the stock CPUs and the 1.8 GHz upgrade with Mac OS 9.1 (Speedometer will not run the graphics benchmark):
450 MHz 1.8 GHz CPU 32.8 109.8 Disk 4.57 4.67 Math 1206.2 4678.4
Under Speedometer 4, the 1.8 GHz G4 is over 3.3x as fast as the 450 MHz CPU, a huge leap in performance, but not quite the 4x we'd expect.
For MacBench 5 to run graphics tests, we had to set the display to 1280 x 1024. The disk cache is 8 MB and virtual memory is enabled. The reference machine is a 300 MHz G3, which scores 100. Tests were run using Mac OS 9.1:
450 MHz 1.8 GHz Processor 1368 5347 Floating Point 1692 4420 Disk 2829 3678 Pub. Disk 2449 2957
The CPU score is 3.9x as high with the upgrade, nearly the 4x we would intuitively expect from a 4x as fast clock speed.
Classic Mode in OS X
Apple does something incredible with Classic Mode: In many ways, Mac OS 9.2.2 runs better inside Mac OS X than it does natively. Why? Because OS X handles all of the disk access, networking, and graphics, leaving OS 9 to handle the rest. Best of all, in a dual processor system, Classic is free to take over one CPU while leaving the other one to handle I/O in OS X.
Running Classic Mode in Panther, here are the Speedometer 3 scores:
450 MHz 1.8 GHz CPU 219.0 876 Graphics 85.3 143.2 Disk 379.3 1517 Math 1008.9 3093.5
It's interesting that the CPU scores for Mac OS 9.1 running natively is the same as Mac OS 9.2.2 in Classic Mode, and the Math scores are very close. Because the hard drive and graphics are handled by OS X, their performance is actually higher in Classic mode than when booting into OS 9 itself. The disk score is much, much higher due to OS X caching data from the hard drive.
With Classic Mode in Panther, here are Speedometer 4 results:
450 MHz 1.8 GHz CPU 33.9 108.5 Disk 4.68 4.95 Math 1244.2 4624.9
As above, the CPU and Math scores are almost identical whether booting into Mac OS 9 or using Classic Mode with a single processor, but CPU and Math scores are a bit higher with dual processors. The Disk benchmark is a bit higher in Classic mode, as all I/O is mediated by OS X.
Mac OS X Benchmarks
Let 1000 Windows Bloom
Under Jaguar (Mac OS X 10.2.8), the Power Mac G4/450 dual ran Let 1000 Windows Bloom in 56.4 seconds. Using Panther (10.3.9), it ran the benchmark in 41.1 sec. with the stock processors and just 25.4 sec. with the 1.8 GHz upgrade, a 62% improvement. Under Tiger (10.4.11), the base system took 32.2 sec. while the upgraded machine needed only 11.9 sec. to complete the test, an improvement of 170%!
Power Fractal doesn't run under Jaguar. Under Panther, this test takes 7.0 sec. on the base computer and just 1.8 sec. with the CPU upgrade. Under Tiger, it reports 7.9 sec. with dual 450 MHz CPUs and 2.1 sec. with the 1.8 GHz upgrade. I suspect Power Fractal is only using one CPU.
Xbench 1.3 doesn't run under Jaguar, so we ran it under Panther and Tiger. Xbench is calibrated to a 2.0 GHz dual Power Mac G5, which would score 100.
450 MHz Dual 1.8 GHz 1 GHz Dual MDD 10.3.9 10.4.11 10.3.9 10.4.11 10.3.9 10.4.11 Overall 22.8 25.7 51.2 50.6 42.4 50.4 CPU 29.5 28.9 94.6 75.1 66.0 54.3 Threads 34.9 39.4 80.4 77.0 61.7 71.3 Memory 20.7 23.2 26.8 25.4 38.7 38.1 Quartz 27.2 25.1 65.9 53.5 58.0 55.2 OpenGL 34.0 38.6 82.9 86.0 55.6 60.7 User Int. 9.1 12.0 33.3 44.6 26.3 57.9 Drive 51.2 49.4 52.0 50.0 31.9 35.3
The overall score for the MAXPower 1.8 GHz upgrade is roughly double that of the dual 450 MHz processors that came in the computer (and half the power of a dual 2.0 GHz G5), but that's far from the whole story. The CPU score is 2.5-3 times as high, the Threads rating roughly double, Quartz and OpenGL are over twice as fast, and User Interface is nearly 4x as fast. Drive scores are barely changed at all.
It's interesting that the CPU, Quartz, and Drive scores are higher under Panther than Tiger, yet Tiger wins in OpenGL and User Interface. Threads and Memory are higher with dual processors under Tiger vs. Panther but lower with the single CPU.
By way of comparison, my dual 1 GHz "Mirror Drive Door" Power Mac G4 benchmarks 42.4 overall under Panther and 50.4 under Tiger. The CPU test score is much higher for the 1.8 GHz upgrade, but Memory is much faster on the Mirror Drive Door model with its 167 MHz system bus (vs. 100 MHz in the Mystic).
Geekbench only runs on Tiger. Here are our results under Mac OS X 10.4.11:
450 MHz 1.8 GHz Overall 439 916 +108% Integer 529 1256 +137% Float. 529 1025 +94% Memory 266 508 +91% Stream 165 162 -2%
Overall, you can expect twice the performance with the 1.8 GHz upgrade that the stock dual 450 MHz configuration provides.
The Power Mac G4 line was the last family of PowerPC Macs that can readily take a CPU upgrade, and while the 350 MHz and faster G4 processors that shipped with them offered a lot of power in their day (the era of the Classic Mac OS), Apple never offered G4 Power Macs with 1.6 GHz to 2.0 GHz CPUs like you can buy today.
If you have a Power Mac G4 (Sawtooth through Quicksilver 2002) that you're happy with but need more horsepower, the MAXPower upgrades - ranging from $230 for a single 1.6 GHz CPU to $699 for a pair of 1.8 GHz G4s - are worth consideration.
Only you can put a dollar value on increased performance and productivity. One huge plus to upgrading your current setup is that you don't have to worry about memory, drives, add-in cards, etc. not being compatible, because you're not migrating to a newer Mac.
Manufacturers and distributors: Interested in having your product reviewed? Please read our review policy.
- Mac of the Day: Quadra 840av, (1993.07.29. The fastest Quadra ever included a digital signal processor for capturing video.)
- Support Low End Mac
Low End Mac Reader Specials
Cult of Mac
Shrine of Apple
The Mac Observer
Accelerate Your Mac
The Vintage Mac Museum
Mac Driver Museum
System 6 Heaven
System 7 Today
the pickle's Low-End Mac FAQ