90nm 3GHz PPC 970FX by Summer

dmdimon sent in linkage to a Forbes story on the upcoming PPC chips and notes "IBM is said to be ready to deliver a new version of its PowerPC processor to Apple by the end of this year in from sizes of 130 nanometers to 90 nanometers... Apple CEO Steve Jobs has already gone on the record saying that the G5 computer will contain PowerPC chips that run at 3 GHz by the summer of 2004. A mid-step between the current systems, which top out with two chips running at 2 GHz, and systems with chips as fast as 2.6 GHz would be a logical move come January..."

Comments

Apple

[KoolDude]

wow... with this Apple will be dying much faster ! ;)

Great for consumers

[pissoncutler]

It's a boon for consumers. Now we have a real choice in architecture (ppc vs. x86) as well as brand (amd, intel, ibm) without sacraficing performance.

Perhaps this will force Intel to to up the ante.

Re:speed

[the Man in Black]

Most of the people encoding audio and video

Macs have been the native platform of artists and designers doing serious image creation/manipulation, video editing, and music composition for a long time. OS X just continues that tradition, but makes it simple for the end-user to also get into how powerful a multimedia machine the Mac is with tools such as iMovie, etc. And of course, on the other end, you've got these two.

So, to answer your question, ramping up speed on the G5 chips is not only good for the whole marketing "Mine is bigger" approach, but there is also real value to Mac users, from casual to hard core.

Re:POWERBOOK EATS YOU.

[American AC in Paris]

I would have to wait until the second generation of that machine b/c I would like to make sure that Apple just isn't unloading G4 parts with a G5 chip.

I agree with the wisdom of letting others find the flaws in a first generation laptop--It's too easy to get burned with a brand new laptop design, pun intended.

That said, Apple puts more effort into laptop design than just about any other manufacturer I can think of. I seriously doubt they'd slap a G5 processor into a G4 design and call it done.

Re:POWERBOOK EATS YOU.

[micsaund]

You missed the part about the G5 running cooler than G4 and also, the switch from .13 to .09 processes should help as well. The net effect may be an overall cooler machine which does not sacrifice MHz.

Re:Sweet

[rupert2000]

If you haven't already started saving, its probably too late.

Unofrtunately...

[Guano_Jim]

The new G5 chips will require 1.21 jiggawatts of power to operate effectively.

Re:speed

[TempusMagus]

Most of the people encoding audio and video and playing games are running x86.

I'll take the bait. You sir are incredibly ill-informed. You confuse the needs of the home-fiddler with the needs of the media professional. I'm sure most audio encoding that takes place in people's homes is PC based, while most audio encoding, recording and mixing done in professional studios is mac based. I'm sure most wedding videos, bar mitzvahs and "baby's first steps" videos are editied on PCs while more independent films, music videos and documentarys are editied, composited and compressed on Macs (Powerbooks even!). You mention that mac people are folks who are more interested in doing normal everyday things. And on that point, it wins hands down in terms of user experience. Speed is important there as well I might add. I own a crap-can of computers running a variety of OSes and when people ask me what I prefer I always tell them "I love them all 'cause they all make me money!". Macs suck for games and they don't run 3DSMAX. Macs arent good choices for low-cost render machines either.

Re:speed

[scrod]

Does anyone even care about the leetness of their speed with Apple stuff?

Yeah, I'm willing to bet that the people who make 80 million dollar movies with them do. Your friends who encode Simpsons episodes into DivX don't count, sorry.

Fobes just reads the rumor sites

[joekra]

As has been seen time and time again, Forbes.com simply reads and summarizes the Mac rumor sites. This is not new information.

See MacRumors.com for Forbes' "sources".

Re:ooo surpise!

[Junks Jerzey]

Macs will run at 3 ghz? WOOHOO! That means AMD and Intel will have 6gig chips!

Realistically speaking, though, performance increases have slowed down in the x86 camp. The jump from 2.8GHz to 3.0GHz came with a much larger increase in power than than the 7% increase in raw clockrate. Ditto for the 3.2GHz P4. Now Intel is apparently having a lot of trouble just getting bumped up to 3.4GHz, a CPU that dissipates over 100 watts. I'm not saying Intel won't break past this barrier--of course they will--but diminishing returns have kicked in hard. A 4GHz P4 is going to dissipate 150W at this rate. How long can it keep up? These are not the kind of CPUs you can easily put in a desktop, let alone a small-form-factor PC or notebook.

IBM is going to have the same troubles with the PPC970, but at least they're ahead of the game. The cleaner design of the PPC line has suddenly become a powerful advantage.

90nm Soft Error Rate

[holland_g]

To quote Intel's own engineers: "As semiconductor technology advancing to 90nm feature sizes, radiation induced soft errors have become a major reliability concern." C.Dai Presentation

I wonder how many software errors will be caused by neutrons hitting the processor and upseting logic gates? I have not seen any test results from Los Alamos for 90nm processors using EIA JESD 57, (1996) JEDEC Standard - Test procedures for the measurement of Single Event effects in Semiconductor Devices from Heavy Ion Irradiation. Unfortunately the Radhard server at NASA is down right now so I can't check the server for the latest test results.

Some people think Failures in Time (FIT) rates will get better at 90nm than 130nm. Some think the opposite. Xilinx and Actel are arguing over it. Caches are epecially vulnerable. In a critical software application, this is unacceptable, and sometimes the cache needs to be disabled altogether.

One method of addressing this is built in checksumming on the cache, and triple redundancy on certain registers like program counter, etc... This does induce a performance hit.

Re:speed

[TomorrowPlusX]

Well, I for one am doing robotics & AI simulation on a mac.

In fact, it wasn't until I *left* x86 (linux) that I got a platform where opengl worked well enough that I could write a proper display layer on top of my system, not to mention that my PB g4 was actually cheaper than the pIII thinkpad it replaced and in my tests was significantly ( e.g. more than 3x) faster.

Now, I don't do any audio, and I don't do any video; but my simulation is pretty f*cking heavy on the cycles -- and it rips. I have no complaints.

People who gripe about mac performance just haven't actually *used* one. And they certainly haven't written any code for one.

Re:speed

[jo_ham]

We're a 100% Apple shop running Final Cut Pro 4 and DVD Studio Pro 1.5 and I assure you we need that power.

We upgraded from a DP 450 G4 box (which was no slouch itself) to a DP 2.0GHz G5 recently and we've more than quadrupled our productivity when it comes to big renders, mpeg2 encodes and multiplexing.

I don't know of anyone in our business using x86 for video editing. None seriously anyway.

I know a couple of shops who use x86 boxes as cheap horsepower in render farms - but ultimately controlled by a Mac at the nose end.

We use our DP G4 as a Quake III server for company LAN matches when it's not encoding mpeg2 on a job. I know Q3A isn't exactly a taxing game on today's graphics cards (none of our client machines even break sweat) but you can't beat it for gameplay.