A few days ago we learned of a story that will shake the foundations of the technology industry. After more than fifteen years with Intel, Apple has decided to design its new processors for desktop and laptop computers. However, and although it surprises younger readers, it is not the first time that Cupertino dreams of it; This is the story of how Apple always wanted to have its own microprocessor.
Almost since its inception, Apple has been clear about the advantages of vertical integration and the weakness depend on third parties for the fundamental component of any computer: the microprocessor. Surely many readers remember the abandonment of the PowerPC and the arguments that were used then, but to begin our story we must go back many more years. Specifically, to 1989.
At Archive.org you can consult a leak that contains an extensive technical document of more than 200 pages which describes, in great detail, the architecture of a multicore CPU a decade before we started seeing them on PCs.
Work began in the mid-1980s and continued until the end of the decade, where it was classified under the name “Scorpius Architectural Specification,”. It is likely that this name tells you little, but if you are fans of the history of computing, you will recognize how it ended up being called: The Aquarius Project.
Behind the departure of Steve Jobs From Apple, the company’s R&D department started a project to design and manufacture a multi-core CPU architecture. Apple, at the time led by John Sculley and deeply influenced by Macintosh’s chief development officer, Jean-Louis Gassée, wanted regain technological leadership According to the Macintosh, it was being overtaken by the competition.
Apple’s approach had a high technical level and was supported by huge amounts of money but financial muscle was missing.
“Apple was not a microchip company and lacked the resources to become one. In addition to hiring highly specialized personnel with microprocessor design, buying the necessary equipment to implement the design and manufacturing the final product (or hiring companies like Fujitsu or Hitachi to do it ”). Companies like Intel or Motorola spend millions of dollars every year designing and manufacturing microprocessors. Apple was fine, but it didn’t have billions to spend, ”says Tom Hornby on Low End Mac.
Although the Aquarius project never materialized, if anticipated concepts that we would only see until many years later. In addition to multicore and parallel processing, he anticipated the need for an integrated graphics chip, something that Intel didn’t develop until the 1990s with the Intel i810 chipset.
A couple of years ago, Jean-Louis Gassée recalled that the relative failure of that project marked the present and the future of Apple:
Although the work of the multicore processor did not produce direct results, the Aquarius project is an example of Apple’s constant desire to control the future of its hardware. This longing would be manifested again when Steve Jobs bought Palo Alto Semiconductor to develop Ax microprocessors, the key to the brutal performance of the iPhone and iPad and responsible for much of its success. “
Let’s go back to 1990, with An Apple that knows what it needs but doesn’t have enough money to build it. The solution was an alliance with IBM and Motorola, two titans of the chip, which did much of the “dirty” work for them and promised that they would revolutionize the future of computing.
Although the first PowerPC models surprised critics and the public due to their excellent performance, the first and almost only client was still Apple. The bulk of the computing industry based its progress on x86 and even IBM, a partner in the alliance, never 100% trusted the platform.
Most analysts agree that the biggest problem was volume. Apple never reached a level of sales as great as for PowerPC could compete with the binomial Intel and Microsoft (which, remember, devastated Windows). Instead of succeeding in consumer computing, the PowerPC had a great success as a console processor video game and not in vain, the generation consisting of Nintendo Wii, PlayStation 3 and Xbox 360 used this architecture.
Did you know that the three consoles share architecture?
Much of the reputation of “computers for professionals” of Apple teams is a consequence of the commitment to PowerPC, a truly powerful processor and far from the competition, especially in the G3 and G4 iterations. It is probable that some veteran reader remembers the quality of those equipment and the differences with respect to x86 alternatives. In videos like this, Apple was in charge of remembering them.
PowerPC’s technology zenith came with the G5, the first of 64 bit of the company and the one that Steve Jobs, in one of his famous keynotes, promised that it would reach 3 GHz. Without knowing it, he was marking the beginning of the end of an era.
PowerPC never reached 3GHz due to an architectural problem that was impossible to solve and the transition from Apple computers to G5 ran into an even bigger problem: the processor was designed for workstations and it was not easy to encapsulate it in a laptop without having large energy needs and temperature problems.
With a portable market growing in double digits, expensive processors to manufacture and an increasingly narrow technological gap, Steve Jobs had to call Santa Clara. “Ten years have passed since our transition to PowerPC and we believe that Intel technology will help us create the best personal computers for the next 10 years.” Prophetic.
Steve Jobs did not hesitate to humiliate the G4 to sell the pass to Intel.
Although IBM maintained its POWER division without Apple as a client and there are projects underway, the idea of revolutionizing computing and undoing x86 died the day Steve Jobs presented the first computers with Intel.
In fairness, the partnership between Apple and Intel has been tremendously fruitful and has helped the Cupertino company expand its market share. However, Apple is a whimsical customer, who does not want to know about cadences, roadmaps and, of course, does not want to wait for the competition. It is not the first time nor will it be the last, as you well know at NVIDIA.
With the jump to ARM and its own processors Apple looks for the difference, to have something that the competition cannot have and to achieve the desired vertical integration, the perfect symbiosis between the hardware and software that they have been searching for almost from birth, in the 80s. Will they succeed this time? If something is not missing this time it is money to try.