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.
Developer Delves Into Reasons Why Apple's M1 Chip is So Fast
- Thread starter MacRumors
- Start date
- Sort by reaction score
another mean spirit towards Intel.
PS the Mac is having a hard time surviving. roughly 7 to 8 percent of all computers are Mac
92 Percent are windows and use Intel.
Where do you see the best prospect for customers, business and money? The 92 Percent. The majority.
And if the M1 Macs don't sell well the Apple OS Share will be less than 7 or 8 percent.
I was one of 15 people who invented x86-64 (we didn’t call it that.). That better?
What’s your CPU design resume?
(PS: powerpc, MIPS, and SPARC are all still used, with new CPU designs released and sold in actual workstations and devices every year).
Intel’s full of mean spirits. They made job applicants pee in a cup (maybe still do. I dunno). Their CEO would walk around the parking lot checking for cars that arrive late. They would make fun of the guy responsible of the FDIV bug, in front of his face, to strangers.
Nasty place.
No. According to you x86 is a poor design. Two dead processor, one almost dead processor, and (according to you) a poor design processor. Not really impressed with that resume.
Not impressed with your resume. Your experience is with dead end, almost dead end, and poorly (according to you) processors. One doesn't need to be the captain of a ship to say the captain of the Titanic didn't do a good job.
When did I say x86-64 is poorly designed or a dead end? You are making things up. I just said that RISC has inherent advantages over CISC. And I didn’t ask if you were impressed by my resume. I asked what the basis is for you to have any opinion as to whether ”RISC is overrated.” What is it about your CPU design experience that leads you to that conclusion?
The PowerPC chip on which he worked was the fastest PPC chip made at the time.
Pretty sure he worked on the fastest SPARC chip.
Do not know on which he worked.
Yup, it is a great resume. Especially when one includes the. AMD x86_64 on which he also worked.
Since we are comparing resumes, which CPUs have you designed? In what is your PhD? How many published papers do you have?
Power is not dead. PowerPC is !== Power. PowerPC (small side RISC) is would not scale down. ARM scales and won the small system. So not sure that's a great arguement you've presented.
The MIPS was a bipolar GaAs design in the mid-1990s. Ran at a gigahertz. Not commercial - made as a proof of concept, funded by DARPA.
Because I didn’t keep working on PowerPCs. Duh.
First, Apple was never a customer of Exponential for various reasons, I said it was the fastest PowerPC chip, not the fastest chip from anyone. Second, Apple moved from PowerPC because the two companies making PowerPC chips (IBM and Motorola) were not focused on the same types of systems that Apple wanted to sell.
Motorola’s primary customers wanted processors for embedded systems, lower performance and inexpensive. IBM was making Big Iron and they did not care about performance per watt which is fine in a data center, but not so good for a laptop.
SPARC suffered from a lack of scale, hence an inability to invest enough to keep it at the cutting edge.
Apple’s Silicon has a guaranteed market many times larger than that for Windows, ensuring that they have the budget to invest to maintain their lead.
Somewhat off tangent from the RISC vs CISC debate but I had a then-new 1.8GHz (or 1.6GHz? Can't remember now) Power Mac G5 and then-top-of-the-line Dell consumer desktop with a Pentium 4 at 3.2GHz in the office and, out-of-curiosity, ran the whatever version of Cinebench was available circa 2004.
My knowledge of CPU architecture was (and still is) rudimentary, so I was genuinely surprised that the lower-clocked G5 was significantly faster than the Pentium 4 in all the benchmarks. I was buzzing, but nobody else in the office knew what Cinebench was back then and just gave slightly condescending smiles to the poor Mac guy in the office talking about his shiny new toy again.
While the Intel Core CPUs closed the gap with the PowerPC 970, I was still surprised that Apple made the switch. It felt like they were abandoning the superior architecture because they couldn't afford to continue its development.
My knowledge of CPU architecture was (and still is) rudimentary, so I was genuinely surprised that the lower-clocked G5 was significantly faster than the Pentium 4 in all the benchmarks. I was buzzing, but nobody else in the office knew what Cinebench was back then and just gave slightly condescending smiles to the poor Mac guy in the office talking about his shiny new toy again.
While the Intel Core CPUs closed the gap with the PowerPC 970, I was still surprised that Apple made the switch. It felt like they were abandoning the superior architecture because they couldn't afford to continue its development.
Your complaints about the other products are all about commercial issues, over which I am pretty sure the CPU architects have no control. Blaming the captain of the Titanic for the substandard rivets the too small rudder and the watertight bulkheads that did not go all the way to the ceiling, seems odd.
Apple wasn't really involved in the powerpc development. The real reason they switched is that IBM had little interest in the soon to be exploding laptop market, and Motorola wasn't competent. Exponential wasn't big enough to supply an entire line of processors, and targeted only the top-of-the-line desktop.
Well, there goes your credibility. It is not a change of name. Power and PowerPC are related but they are not the same.
PowerPC was an Apple, IBM, and Motorola joint venture. If the PowerPC was such a superior processor why was it abandoned? PPC was RISC. RISC versus CISC is the only factor that matters.