25 years ago this month Apple introduced what has become known as the x200 series of Power Macintosh and Macintosh Performa computers. These machines have been ostracized by those of us in the Mac community, especially here at Low End Mac for almost as long. Now, more than 2 decades later, new data has come to light that calls both our opinion of these machines and the information on our site about them about them into question. Could it be that the x200 Macs are decent machines after all?
It has been nearly 23 years since Scott Barber published his scathing article about the x200 series titled Performa and Power Mac x200 Issues here on Low End Mac .
Scott made it very clear in his article that he thought all of the Performa and Power Macintosh 5200 – 5300 and the 6200 – 6320 machines suffered from a horrible design that made them much slower than they should have been. His article prompted our publisher, Daniel Knight, to call the “x200″ machines the worst Macs ever built and inspired him to create a list of Macs called “Road Apples”. The x200 Macs are at the very top of that list.
I too believed that the x200s were the worst Macs ever, based mainly on the information I’ve read online.
Recently, while I was browsing through some vintage Mac forums, I came across an article written by Daniel L. Taylor. The article (The Mythical Road Apple), details how he believes that Scott Barber got the x200s all wrong. I’m not an engineer, but Mr. Taylor seems to have a good understanding of how the hardware works. His article convinced me that I needed to take a much closer look at these “Road Apples” and determine for myself if they really are as bad as I’ve always been lead to believe.
I decided that the only way I could learn the actual truth would be to pull my old 6200 out of storage and run some tests. When I first set up my 6200 I reinitialized the Hard Drive and installed a fresh copy of Macintosh System 7.5.5, ClarisWorks 4, Netscape 3, OpenTransport 1.3, Stuffit Expander 5.5, and MacBench 4.0. MacBench was a standard in Mac benchmarking many years ago and provides a comprehensive suite of benchmarking tests.
I’ll be focusing on general benchmarks, testing the CPU, FPU, Graphics, and then dividing the total from those categories to give an overall result. Baseline results are from a stock Power Macintosh 6100/60. Other comparisons include my personal Power Macintosh 950 (a Macintosh Quadra 950 with a 66 MHz PowerPC 601 Upgrade Card), and a stock Power Macintosh 7200/75.
Machine CPU FPU Video Disk Overall
6100/60 100% 100% 100% 100% 100%
Q950/66 105% 91% 122% 101% 104.75%
6200/75 92% 95% 93% 115% 98.5%
7200/75 102% 117% 115% 107% 110.25%
I also ran some tests of my own. They examine Network performance (by using MacBench’s Disk performance test on my File Server via Ethernet), Application launch speed, website loading times using theoldnet.com, and decompression speed of AppleFritter’s new eWorld Application using Stuffit Expander 5.5.
I chose these programs and websites because I felt they represented the use case for most people who would have bought a 6200 when it was new.
Here are the results of my personal real world tests:
Machine Network Disk Speed Test
Machine Load Netscape Load LEM Load Cyber24.com
Q950/66 :18 :19 :13
6200/75 :20 :21 :17
Machine Load ClarisWorks Decompress eWorld Load eWorld
Q950/66 :08 :52 :06
6200/75 :08 :52 :05
To say that I was surprised by these results would have been an understatement. Something in me still expected the results to show the 6200 to be far slower than this. In reality, the 6200 is only slightly slower (only 1.5% overall!) than the 6100. Comm Slot Ethernet networking does appear to take a decent hit (only 3/4 as fast as my Q950/66), but still runs at acceptable speeds. It would appear that it isn’t nearly as crippled as I thought it was – or Scott’s article implied.
Which machine should you consider?
Let’s pretend that it’s July of 1995 right now, and you’re looking to buy a new Mac. The new Performa 6200 was just introduced and was replacing the Performa 611X models. The 6200 has 4 times the storage (and, contrary to popular belief, it was faster than the 61XX’s SCSI drive), twice as fast a CD-ROM drive, unique expansion options such as the Apple TV/Video System (which included a remote control for TV and Audio CD functions), Automatic Power On/Off capabilities, front mounted volume buttons, a front mounted headphone jack, a much smaller footprint, and it cost $300 less than the Performa 611Xs. In my mind the initial 10% performance hit would have been worth it considering the plethora of other features and options that were gained over the Performa 611Xs. I think the same value equation rings true today for those wanting to get into Retro Mac Computing. The x200s offer a very decent value for anyone wanting to use real hardware to explore vintage software.
One could argue that the 61XX series has an edge over the 6200 because you can add an L2 cache to them, but I would argue the 6200 has a bigger edge when you consider that the logic board on the 6200 can be easily replaced with the logic board out of any faster x200.
Unfortunately, it appears that using the board out of 5400/5500/6360/6300/6500 in a 6200 – 6320 isn’t possible. The edge connector and pinouts are the same, but I attempted to use a 5400/120’s board in my 6200 just today in an effort to run some tests, and it wouldn’t even chime when powered on. Putting that 5400 board in my 6360 worked just fine. Despite the fact that I know I used a 6500’s board in a 6200 about 20 years ago, it appears the Power Supply in 6200 – 6320 models isn’t powerful enough to drive 5400/5500/6360/6400/6500 logic boards. Perhaps the 6200 I had all those years ago had its original Power Supply replaced at some point with a 6360’s Power Supply. What is more likely is that the Power Supplies in the 6200 – 6320s are wearing out and no longer have the output they once did. Either way, if you want to use a 5400/5500/6360/6400/6500 logic board in a 6200 – 6320 you’ll also need to install a 6360’s Power Supply. Honestly though you’d be better off just picking up a complete 6360.
With all of that being said, If you have a 5200 – 5300 you shouldn’t have any issues installing the logic board from a 5400/5500/6360/6400/6500 in it. Doing this provides a huge performance increase. Using the board out of a 5500/6500 also adds onboard ATI Rage II graphics, which give it very good graphics performance.
I’ve been shocked by what I’ve discovered about these machines over the last few days. While it is true that the x200s aren’t quite as fast as their 601-based cousins, I no longer believe that they’ve been treated fairly. Their performance was more than adequate for most general purpose tasks back in 1995, and for most mid-90s software today. Even the thing they were supposedly the worst at (loading and interacting with websites) feels fine when using theoldnet.com today.
Based on the new information that I have discovered and the tests that I’ve run myself, I no longer believe that the x200s are anywhere near as bad as their reputation suggests. I now believe that they are definitely worth considering if you want to get into Retro Mac Computing.
On a more lighthearted note, there is also something unrelated to performance that the x200s have that no other Macs have (other than those with a Power Macintosh Upgrade Card): a very unique Startup Chime. I think it sounds much better than the Guitar Strum Startup Chime on the 6100/7100/8100 series, and is different enough from all later Macs to make it stand out in a very good way.
Bottom line: Keeping in mind the market the Power Macintosh and Performa 5200 – 5300 and 6200 – 6320s were made for, they are much better than their reputations suggest. They may not be quite as fast as the more powerful 7xxx, 8xxx, and 9xxx series Macs, but then again they were never meant to be.