“Apple says it bought P.A. Semi to get its hands on some patents and a bunch of very bright engineers who are already used to working together. I’m sure that’s true – but they also got an irresistible opportunity to shake the money tree while doing something good for the country,” Paul Murphy blogs for ZDNet.
“P.A. Semi designs PPC chips primarily for advanced military and robotics applications and the one they talk about in public, the PA6T-1682M, has had an unprecedented uptake in the industry – so much so that major defense contractors include Lockheed Martin and Curtis-Wright have locked in ten year supply contracts on it,” Murphy writes. “The chip itself is impressive: it’s a 15 watt, 2+Ghz, dual core, dual Altivec, 64bit PPC “system on a chip” with 2MB of level 2 cache per core, hardware packet management (including cryptology), on board memory busing, and eight concurrent PCI/E channels.”
“Why would Apple use it in a laptop? Because Apple has been facing pressure from U.S. Department of Defense (DOD) customers for more secure (i.e. non x86) gear that’s made in America and because all of P.A. Semi’s big customers are long term experts at selling to the DOD. In other words, what Apple really wanted here was technical expertise, but what they got along with it was specific market expertise and a golden opportunity to sell from three to five hundred thousand American made, PPC based, MacOS X machines to the DOD every year,” Murphy writes.
Paul Murphy is betting that Apple will make PPC laptops again. Read the full article here.
[Thanks to MacDailyNews Reader “Fred Mertz” for the heads up.]
MacDailyNews Take: 3Ghz within a year!
The simple fact of the matter is that PA does not have a fab plant. Dreams of Apple made processors are quite simply, fvcking stupid. PA sent a warning to it’s clients that it would be able to meet future orders and that current products will be end-of-lifed.
You watch how badly the other idiot so-called analysts will twist this. The doom and gloom factory will ramp up to exploit the potential future failure of imaginary hardware based on a fantasy of Apple made chips.
Magic Word: “full”, as in “analysts are full of sh!t”.
The hardware security issue in this article is BS, plain and simple. Murphy could do a quick check with any military clients and find that the Panasonic ToughBook is ubiquitous in the field. It may be heavy, but the thing can take a beating and work in hostile conditions.
The PA Semi purchase isn’t about PPC laptops–that train has left the station. It’s about talent and IP. By getting these great innovators and engineers, Apple can make its own chipsets more energy efficient so that a MB or MBP goes from 5 to 7 hours on a charge. That sort of market differentiator is important in a business where becoming an also-ran would certainly hurt Apple.
For the best take on Apple’s purchase of PA Semi and what it means for Apple, check out this Roughly Drafted article.
I hope Apple does utilize PPC chips again. The PPC version of OS X is way better than the Intel version. Just look at the flakiness of Address Book as an example. It is so buggy on Intel. My 400 MHz PowerMac G4 is still working great, at 8 years old.
LOL That was MDN in 03….this is MDN 08 and their nothing but Apple cheerleaders.
” This is why I like MDN. They are not Apple cheerleaders (all the time) and will call them on mistakes when Apple makes them (yes, even Apple makes mistakes). I agree Jobs should shut his yap. What’s he going to predict next, that Kerry will win the U.S. Presidency?”
Mac OS-X is all fleshed out on PPC, why not leverage their knowledge base and produce some low power devices that just run OS-X and most apps without too much engineering? It could happen? It’s not likely but it could happen.
This is all nonsense.
They need Wintel compatibility for the foreseeable future.
Chips or expertise could find its way into future portable devices that aren’t considered true laptops. However, it is highly doubtful the chip will end up in a core laptop line where the need to run windows is still high for grabbing market share.
As a PowerBook owner, I’m happy to hear this rumor. I’m not sure that I believe this, but I’m happy to hear it, especially with the future of Mac OS X and PPC. I do remember Steve Jobs speech about dumping PPC and going with Intel processors. The main reason was obvious, that Motorola wasn’t delivering the faster chips that Apple needed to stay competitive with the Intel versions of computers.
I do remember Steve Jobs saying that they would go back to PPC if it was advantageous for them to do so. And, this purchase of P.A. Semi is a good way of Apple to keep their hands on their hardware, which they have always preferred doing.
But, frankly, 2 GHz dual processors isn’t that fast a speed in todays computers. Intel will surpass that in a few months. What is more valuable is the knowledgeable employees and the built-in defense contracts that the company has to offer.
From the article:
“We don’t know much about its performance because the actual benchmarks have not been published – reasonable estimates, however, put it at about 50% over the IBM 970Fx at 2Ghz – roughly comparable to Intel’s top end dual cores for simple character pushing, about twice as fast as high end Xeon based systems for unencrypted network and graphics I/O, and about an order of magnitude faster than dual core x86 for floating point.”
Longtime readers of this site may remember, when the Switch was made I said that it wasn’t made for performance reasons, or cost, or chip availability, or Windows compatibility – and I had to deal with quite a bit of fanboism as a result. But facts are facts:
* As PASemi has shown, Apple had more than enough performance and chips available (Freescale’s e600 core G4s would have been Yonah-class transiition chips, until the PASemi CPUs came online).
* Costs with Intel are actually higher – still! All Macintel prices went up about $100, and except for the extremely high margin MacPros and servers, have never come down. Apple used to sell iBooks for under a grand – when was the MacBook ever sold for less than $1100?
* Sales of PPC Macs were ALREADY OUT PACING THE INDUSTRY before the Switch was announced. They slowed as people waited for the new kit, and have since taken off again. If the goal is simply more sales, Intel was/is not needed, Windows was/is not needed (only a minority of Mac buyers use Windows on the Mac, and the numbers are getting smaller all the time). OSX was converting people all on it’s own.
There was only one thing that Intel offered that Apple (Jobs) needed. Integrated hardware DRM on every MoBo they produced. Jobs needed it b/c the movie and TV studios, at the time, would not give him content unless software-only Fairplay was augmented, and Intel’s TPMs were what they had been involved with developing, and thus favored. The ultimatum was made – transition to Intel or no content – and Jobs blinked.
Since then, events have overtaken the original plan and intentions, perhaps on both sides, but certainly for the studios. Video online went viral, Apple’s iTunes Store is the only consistently popular and profitable place to sell video downloads, and thus Jobs was able to drag his feet on the draconian ‘lock-down’ methods he initially signed-off on, while the studios had to suck it up and provide the content anyway. The TPMs, so far, have not been lit up. But make no mistake – they, and Intel’s closely collaborated hardware DRM ‘total solution’, were the real impetus for the Switch.
Now that dropping DRM on audio tracks is a fact of life (and the world hasn’t come to an end fr the record labels), it’s certain that DRM for video will … eventually … follow the same path to oblivion. In which case, why stick with Intel? Especially when PASemi can and does make a better PPC CPU, and Universal Binary coding makes any future transitions – of any single product, or the entire lineup – an academic exercise?
THAT’s why Apple bought their own PPC chip designer. Bank on it.
” width=”19″ height=”19″ alt=”cool smirk” style=”border:0;” />