Recently I had the displeasure of reading an atrociously inaccurate and biased article concerning the new Power Mac G5 and its comparison to the new AMD Opteron machines. You can read that here if you want. One of the points the author made was that Jobs suggested that Apple was responsible for too much of the design of the IBM PowerPC 970. In other words, IBM did all the design work on the 970 because they know chips and Apple doesn't. Hence, Jobs was lying!

This poor chap both misses the bigger picture and doesn't know Apple's own history -- much less its current and secret engineering capabilities. In the process of writing this author back I recalled a period in Apple's history where the company actually set out to design its own advanced RISC processor with four cores (or four chips) on a single piece of silicon. This ill-fated project was called Aquarius.

Aquarius

Outside the core group of fifty engineers and the management teams responsible for the project, not much is known about the Aquarius project outside the company. In fact, it's not really clear if even Jobs himself is aware of this project's history.

Aquarius came about precisely out of the same circumstances that drove the IBM PowerPC 970 and future 980. In essence, even back in the late mid 80's Apple felt that Motorola was taking too long to come up with its next-generation processor, the 68040. And Intel was catching up!

So Sculley bought the Aquarius team a $15 million Cray supercomputer to design the processor on and the Aquarius team moved forward. Fortunately, Sculley eventually realized that the project was over Apple's head and eventually killed the project.

The First Jaguar Project

Let me back up a second. In the awful article cited above, the author was under the impression that Apple can't design microprocessors and it has no experience in doing so. Both of those assumptions are for the most part false. Apple does have experience with processor design. It just doesn't have enough experience to compete with the likes of Intel, Sun, Motorola and IBM. In fact, it was a processor design expert at Apple that helped convince Sculley to kill the Aquarius project to begin with. His name was Hugh Martin.

Back in the late 80's Apple was searching for a way to propel themselves past the Wintel juggernaut that was quickly catching up. They realized -- and wisely so -- that the Mac's two primary advantages came down to software OS design and processing power. And by "processing power" we mean more than just the main CPU, but other chips as well. So Hugh Martin, the man Sculley hired as a microprocessor expert, launched the first project with the code name Jaguar. Now for those of you who own the monograph on Apple's industrial design group, you know what Jaguar looked like.

The Jaguar project was an effort by Apple to leap ahead and stay ahead of Intel, which was catching up. Jaguar was headed by Martin and was a new RISC-based Apple computer that would not be compatible with the Mac OS because it was RISC based. It's interesting to ponder why Apple reused this name for the first real performing version of Mac OS X. But likely it simply was a big cat thing and a coincidence. However there is one other important connection, as we shall see in a moment.

The First Jaguar OS

Apple in about 1990 was a company concerned with its future. They wanted to ensure their market advantages and market share. Martin knew that RISC was the future and adopted a new advanced chip Motorola had been working on called the 88110. This was a true RISC chip predating the PowerPC. The 88110 had some potential steam behind it as a viable chip in the market place. According to an article online by Simson L. Garfinkel, Ford Motor Company had committed a billion dollar purchase of 88k based microcontrollers based on the fact that Apple was behind the chip. When Apple moved to PowerPC under the new alliance, so did Ford. This in fact affected NeXT who was developing a next-generation NeXT workstation based on the Motorola 88110. NeXT too would eventually move to PowerPC.

So the new Jaguar RISC 88110 machines ran a new UNIX OS from Carnegie-Mellon University called Mach. This -- I believe -- is when Avi's work at Carnegie-Mellon first touched Apple, officially. Jaguar, as it turned out, failed to pass the higher hurdles in Sculley's management world, not for pure technical reasons but for an inner rivalry between the traditional Mac engineering team and the new Jaguar engineering team. Unfortunately for Apple, Sculley was not equipped to straighten this inner feuding out.

It's very interesting to ponder that Apple back in 1990 was on the tip of embarking on a revolutionary new platform architecture called Jaguar which consisted of an all new RISC-based platform with highend multimedia capabilities (Apple even developed a camera like the iSight for video-conferencing with the new Jaguar system). And that the project ran a UNIX based Mach operating system which would eventually find its way into the Jaguar we use today. --- ANTHONY FRAUSTO, Ed.

For feedback on this article write: [email protected]

Afterthoughts

The Aquarius project is a fascinating event in Apple's history. The idea to create a quad-core microprocessor sounds advanced even today! However, this history supports the idea that Apple might one day take advantage of a dual-core processor design like the future POWER5 by IBM. It should also be pointed out that Hugh Martin, the microprocessor expert behind the Jaguar project in the late 80's was also a supporter that Apple and Sun should merge in order to provide Apple customers a fuller spectrum of computing: from secretarial computers up to highend workstations for engineers.

Aquarius is a code name Mac users should watch with care, if it ever pops up again.