Motorola to Boost 0.13-micron PowerPCs

Anonymous Cow writes "From The Register: 'Speculation that Motorola may soon cease to be a supplier of processors to Apple may be premature. The chip maker yesterday said it had successfully implemented low-k dielectric materials in its 0.18 micron silicon-on-insulator (SOI) processors, bringing an estimated 20 per cent speed bump to the PowerPC line. Motorola expects to roll out the process on its 0.13 micron chips this month...'"

Comments

don't speed bumps ...

[Potor]

slow you down?

This is what Quark was waiting for

[chia_monkey]

Good googly!
This is what Quark has been waiting for. Now that we can zoom along at these blazingly fast new speeds, Quark will finally release the OS X version and the Mac platform will be saved.

Hurray!

Still Playing Catch-Up

[peatbakke]

Low-k? Welcome to the ballgame. IBM rolled out low-k, SOI, and Cu three years ago ... on 0.13 micron. See here and here. So did Intel.

Re:Why dont they release it on X86?

[Trurl's Machine]

Its just the OS I want, I'll buy my own monitor etc.

It's the "monitor etc." that Apple wants to sell, not "just the OS".

Re:I can't help thinking that this is a bad thing.

[standards]

We know that, for internal development reasons, Apple has a version of Mac OSX that runs on Intel/AMD hardware. (It's been widely discussed in the past, both on Slashdot and elsewhere.)

We do? Sorry, there's a huge difference between an interesting prototype and production quality software. In any case, a popular rumor is still a rumor.

We know that the longer Apple uses the PowerPC platform, the less likely the possibility of it switching to an Intel/AMD platform becomes.

Why? They switched to PPC from 68000 after about 10 years. They could switch regardless of the length of time. You're implying that more software would be available after a longer length of time - implying a growing market.

We know that an Intel/AMD platform Apple Mac would probably cost less than a PowerPC platform Mac currently does, and run faster too.

We do? How do we know this? Just because one chip runs at 1.2 Ghz and the other runs at 2 Ghz? Because the P4 runs at 3+ Ghz? Because of bus speeds?

We know that if they could upgrade their Windows PCs to Apple Macs - say, by installing an Apple upgrade card that contained any necessary Apple ROMs, etc and then installing the new OS - millions of users would be tempted to abandon Windows and convert to the Mac OS.

Really????? Wow, that's a leap. And how much would people pay? I know I'd pay just about $0.

Might as well just have a software licensing key scheme - as Mac Plus ROMs don't go to far these days ;-)

We know that this Intel/AMD platform Apple Mac would get much better support from hardware and software manufacturers.

Really? Just because OS2 ran on Intel didn't help it.

An Apple Mac running the newest hardware would never be significantly disadvantaged performance-wise, and Apple would attract a lot of users who previously considered Macs bad value for money.

Using a particular chipset does not guarentee great performance or value.

We know that this would make Apple a force to be reckoned with once more, make Microsoft very anxious and millions of customers delighted.

I think Apple has already achieved that. Throwing a couple "ROMs" into an Intel box just doesn't fit the big picture.

Not for high end Macs

[adzoox]

The PPC 970 (& later 980) are all but confirmed and shipping according to several sites (non rumor sites too). The most vocal being MacBiduolle. They had a Quicksilver prototype 2 months before it's appearance and have been on the money for most things. There are CONFIRMED machines in Adobe running 970s too.

Apple will most likely use this as an oppurtunity to drop the G3. Finally Apple will have Altivec across the board. You have to take into account that the manufacturing process also reduces heat and well ... size ... making this sound more and more like a processor for an iBook.

I also beleive Apple will use this as an oppurtunity to make everything above 1Ghz this year. We will most also likely see quad G4 Xserves because of this (moto producing better G4s)

The 970 is a great chip. It's benchmarks at the Microprocessor Forum VERY HANDILY beat EVERY processor put up against it - even the AMD 64 bit!

Apple shouldn't move to x86 as suggested in the redundant Apple naysayers. (hey you "apple is dead people": how about looking in my journal?) I rather like the RISC processor anf the PPC - there is MUCH less code overhead and easier "addon" capability (cache, media functionality, i/o) - Motorola has been the hold up in it's development and needed someone like IBM to step in and lend a hand, they have done so.

Weird Anagram

[jabbadabbadoo]

The following suggests that Motorola is participating in some dark cryptology conspiracy. (all lower case to avoid /.'s lameness filter :-)

motorola processor
otorola processor - m
torola processor - mo
trola processor - moo
tola processor - moor
tola procssor - moore
tola procsor - moores
tola pocsor - moores r
tola pocor - moores rs
tol pocor - moores rsa
tol ocor - moores rsa p
tol oco - moores rsa pr
tol co - moores rsa pro
ol co - moores rsa prot
l co - moores rsa proto
l o - moores rsa protoc
l - moores rsa protoco
moores rsa protocol

Weird indeed... especially when condiering this one (search for RSA in the document)

They can't support all that extra hardware

[caveat]

We know that if they could upgrade their Windows PCs to Apple Macs - say, by installing an Apple upgrade card that contained any necessary Apple ROMs, etc and then installing the new OS - millions of users would be tempted to abandon Windows and convert to the Mac OS. (Obviously, whether allowing non-Apple customers to convert their machines in this way is something that Apple may or may not want to put into practice, for competitive reasons. Remember, one of the first things that Steve Jobs did on his return to Apple was kill off the authorised Apple clones businesses.)

Or, as has been pointed out many times before, Apple doesn't want the toruble of supporting god-knows-what hardware is going to be in the masses' PCs. One of, if not the major, reasons Apple is able to make the OS play so nice most of the time is their control of the underlying hardware - sure, you can get most any peripheral you want (as long as it comes with a Mac driver), but the basic computer is always consistent.
I suppose Apple could just tweak the G4 mobos and replace the processors with P4s, or replace the internals completely, but I doubt that's where the costs of the machine lie, plus the homebrew crowd would scream bloody murder. It makes me shiver to think what OS X would be like if it had to support every piece of x86 hardware under the sun...

A better move than going IA-32...

[Kjella]

We know that the longer Apple uses the PowerPC platform, the less likely the possibility of it switching to an Intel/AMD platform becomes.

What would they go to? IA-32 is a poor choice when processors are starting to move to 64 bit with X86-64 and Itanium. Going Athlon-64 would be ahead of time, it's not even out yet (and if Intel managed to get their 64 bit solution pushed through, X86-64 would fade into a niche) and if they went IA32, by the time they're done it'd be time to change again.

We know that an Intel/AMD platform Apple Mac would probably cost less than a PowerPC platform Mac currently does, and run faster too.

Code designed for X86: Yes.
Code designed for PPC, compiled for X86: Maybe
Code compiled for PPC, emulated on X86: Hell no

Nevermind that Apple has, and always will have a high mark-up to cover the costs of developing their software.

We know that if they could upgrade their Windows PCs to Apple Macs - say, by installing an Apple upgrade card that contained any necessary Apple ROMs, etc and then installing the new OS - millions of users would be tempted to abandon Windows and convert to the Mac OS.

Apple's business plan is to be a cathedral where PCs are the bazaar. They wish to deliver a _solution_, where they control the hardware and the software, that will "just work". They do not want to get into the driver and compatibility problems of PCs, because then they would lose their greatest advantage. And there's a price tag involved, of course. Which is also why getting dinner served (the solution) is more expensive than buying the ingredients and cooking it up yourself (hardware+drivers+OS+applications+utilities).

To me, who likes to mix and match and create my own "dish", that is probably not of that great a value. But it is of value to some, and those are Apple's customers. And it sounds like a viable business plan to me. The rest will say it's too expensive, of course.

Kjella

Too little, way too late

[TheSwirlingMaelstrom]

Sitting on or near my desk are a 800MHz Athlon (running a Linux 2.4.x kernel), an 800MHz G4 Titanium (MacOSX 10.2.x), and a 1.8GHz P4 laptop (Linux 2.4.x). The Titanium was bought for me by my employer, since many of the people here use them, and I do application and hardware support, as well as Astrophysical research.

I have benchmarked my applications on these three platforms (and the best benchmarks are, of course, your own applications, aren't they?). The G4 is slower, by about 20%, than the 800MHz Athlon. Arguably, if my applications were made 'Altivec-aware' they'd run significantly faster on the G4, but if I were to use SSE2 extensions on the Athlon or P4, they'd run faster on those platforms, too.

Although I kinda like MacOSX (and abhorred MacOS9), and think Apple wins top marks for esthetics, their hardware is way too slow for a 20% improvement in processor speed to give them the boost they need.

The best move for Apple will probably be to go with the new IBM chips.

My 0.02CDN.

Re:Beyond 1 GHz..?

[tbmaddux]

Honestly, I don't know what I'd do with a dual 2GHz G4 at the moment...

Well, let's see. You don't have to have one, of course; as the hypothetical 2GHz G4 dualies ramp up the low end machines can get cheaper, smaller, and faster. And you obliquely mention scientific analysis; I analyze weeks worth of timeseries data on a G4 1GHz/DP machine. Right now the code takes 15 minutes and halving that time would be a significant boost. There are some sensitivity tests to model parameters I that I could make more thorough simply by adding points. I could do more in other overnight simulations. And I'm sure gamers and people who grind Photoshop jobs or render all day will appreciate the speed-up.

Finally, there may be something you or I haven't thought of yet. Apple is doing a good job of finding new things for us to do with our faster processors (iDVD, for example, uses a lot of resources) while other software/OS companies have not done such a good job of finding a "killer app" for having computers faster than they were in 1999. Not that iDVD is the killer app -- I think it's still out there waiting to be found.

Still Too Little, Too Late Anyway

[mgbastard]

This is actually bad news. The MPC7457 still doesn't make full use of the bandwidth available in the DDR400 RAM the Macs are currently using. The MPC7470 does, but we're still not getting that chip - for whatever reason - I assume its a manufacturing & design issue. It's been a very long delay.

Motorola looks pretty amateurish with this feeble boost. This is a manufacturing tweak that intel and IBM have made months ago in their primary foundries. The MPC7457 likely isn't going to get used in any serious Macintoshes - perhaps it will go into the iBook and iMacs eventually.

So perhaps Motorola has given up on the MPC7470, and conceded that market to IBM's 970 and 980 chips. Let's hope so; I would like to buy a new workstation pretty soon. ;-)

Motorola is done

[Drakon]

Motorola has been driving themselves very very deep into the embedded market. They're making the Motorola PowerPC into a chip for cars, phones, handhelds, refridgerators, etcetera, ad nausium

IBM is moving in the other direction, which is frankly the direction that apple NEEDS to go if they want to compete and keep this architecture.

They're planning on keeping this architecture.

Motorola sells lots of PowerPCs

[Erich]

And apple is not the big customer.

The big customer is everyone who's buying PowerQuicc's and putting them in embedded spaces. PowerQuicc's with RapidIO connections, PowerQuicc's four-on-a-board, lots and lots of PowerPC chips going in lots and lots of embedded spaces.

I was recently at the Global Signal Processing Expo and it was amazing how many people were doing tasks involving heavy signal processing -- where you would expect DSPs and FPGAs -- on PowerPC chips. The interesting thing was that raw number-crunching power wasn't always the most important thing -- many times it is bandwith (what kind of interconnect you have to your processor makes a huge difference when you are trying to process gigabytes of information a second). Sometimes it is programmability that is the reason (use of familiar tools is a big plus). Sometimes you just want to use the same chip to do your signal processing as your network I/O.

Companies like Sky Computers are selling more PowerPCs than companies like Apple Computers.