Separate names with a comma.
Discussion in 'Hardware Rumors' started by OKComputer, Jul 31, 2002.
>"We don't have the plans to have the production capacity they have. Rather we're focusing on the high end of the market...working with the cream of the crop," Davari said.
that's Apple for sure!
I'd love to have a Power4 in my next PowerBook... fingers crossed
By the way, I like the name OKComputer
so did Radiohead, when they released that album last year
err....you'd better check your dates again.
Very cool ---> IBM has turned the new plant into a chip foundry, which will produce a wide range of chips on a contract basis for customers in the communications or consumer electronics industries, among others.
This opens up all sorts of possiblities - especially with the no motorola for Apple rumor.
It doesn't say when the 'quick' move to 90 nm will be though, it will open up producing 130 nm first and double their productivity.
You'd like having 17 minutes of battery life, eh? (The Power4 draws 125+ watts)
Oh man, .13 micron AND 300 mm wafers? That's as good as Intel's fabs and much better than AMD's! This could mean a nice price drop in the G3s at least and whatever other chip Apple's getting IBM to make for them if they are made at that fab.
Keyword is "scaled down"
Scratch that, keywords are "scaled down." Gosh you would think a college student could speak english It's been a long day
I heard somewhere that IBM helped make G4's when motorola couldn't put out when they first came out. If this is true, we could have faster G4's at higher yields. That'd be cool
I'm a university professor--and my experience leads me to the conclusion that the statements that one is a college student and that one can speak English sadly have nothing to do with each other!
The Power4 would be sweet I would love to have one if Apple puts it in a computer. Heck give me the full Power4 I work for Enron
Let's see. Besides the 17 minutes of battery life, add in a very scorched desk and/or lap. Maybe your scaled-down Power4 would be a Power3. It's still more powerful that a G4.
Btw, if you have trouble with English, does that speak highly of Rollins and ENC1101?
Oh, and yes, I would be happy with a Power3
Isn't IBM one of Apple's biggest competitors?
From what I hear, IBM will be providing a PPC processor for the PlayStation 3. Sony had many choices for their new chip, which replaces the MIPS processor. It seems logical they would pick one that gives them enough performance to blow XBox out of the water. In game consoles, nobody cares about clock rate, just performance. So, this tells me that the PPC processor line will eventually have better performance than Pentiums. PPC has the brighter future.
IBM's new plant is not for the Power4 only. IBM stated that they are going after "cream of the crop" firms in high-end electronics. They are in business to make custom chips for such customers. IBM is just what Apple needs. The Pentium roasting technology IBM develops for making the Sony PPC processor can also be applied to Apple's processor. And, IBM is using an SIMD Engine in new processors. Their road map shows it. Regarding AltiVec, which is likely the best SIMD Engine available, Apple must have rights to use it. Apple went with a sole source supplier when they chose the Motorola G4, and no one does that without a good contract. Apple would not lock themselves into AltiVec if they could never get it elsewhere. There may be a price, however.
Just for blue sky thinking, it is interesting that those writing games in the future will be working with a lot with PPC code, both PS3 and GameCube. Also, since AltiVec is so useful, could Sony have been interested in it for the PS3? There may already be a deal with Motorola on this, and AltiVec may become a PPC standard, maybe even purchased by STI. This would be a good thing. Just think if it were commonly used in PS3 game code. Developers would soon learn how to use AltiVec well for performance advantage. Such knowledge would benefit development of OS X applications.
A 2.5GHz Pentium 4 scores 60%+ what a Power4 does on SPEC_int2000. This is the top-of-the-line multi-thousand-$ Power4 with 128MB of cache, competing against a desktop processor that costs a few hundred dollars. What happens when you remove all that expensive cache and multi-core silicon? You have a chip that can't compete with the Pentium 4 anymore, especially considering how fast the Pentium 4 is advancing.
I'm all for Apple using faster chips and staying away from x86, but trend over the past decade has clearly been the increased commodization of high-end hardware. The low end has continually eaten ever further into the high end. What I think is that Apple needs to ask itself: "What does this mean for us? How does this affect our markets?" I could see a move to a better PowerPC chip, such as a new Power4-derivative, but it will take more than that to keep up with Intel. Increased reliance on multiprocessing? A bunch of hardware coprocessors? Whatever Apple decides... a simple Power4 derivative won't be nearly enough.
ENC1101 is Freshman Comp. I in Florida.
I feel sure that Sony looked at the processor market very carefully when they decided against a MIPS processor for the PS/3. Likely, there is no way MIPS is going to keep up with Intel, and Sony knew it. Sony could have chosen Intel to supply the processor. I don't think Intel would turn Sony down, but would be happy to get their business. No, I believe Sony went with the PPC from IBM because it can give them the best performance. It is a consumer chip, and yes, it is eating into the very highest end of processor performance. IBM knows well how good Intel is. I'm sure they are busy on a Power4 replacement as well as Sony's processor. I believe IBM will sell something like it to Apple too.
Well, that was a stupid question; I should've known! From what I've heard IBM is using a new strategy to do their business. Instead of focusing on selling as many computers as possible to consumers and businesses, they are instead focusing on making superior computer parts and selling as many of them to other computer makers. That's a good strategy, IMO.
Shrek, I wasn't addressing your post, but addressing the main topic, which is IBM on the move ... It just appeared after your posting. Regarding your question, IBM is a competitor like every other PC maker is Apple's competitor. That never stopped Apple from using IBM processors before. And Apple is less of a competitor to IBM than Dell, HP and Sony. Yet IBM is making a PPC for Sony, so the rumor goes anyway. True, IBM and Apple both make computers. IBM's PPC computers and Macs don't sell in the same markets now, from what I see.
Also, with the objectives IBM stated for this plant, they must realize that potential customer may see IBM as a competitior. IBM knows they must assure folks about this issue, if they expect to get customers. And I think they expect customers.
The game console market is night and day different from the desktop computer market. CPUs in game consoles need to 1) be cheap to manufacture, and 2) offer good FP performance. This is why most game console CPUs have onboard SIMD units and absolutely suck at integer speed, and why the highest-clocked console CPU today runs at 700MHz (I think? The Xbox is 700MHz right?). Game consoles don't typically need tremendous CPU performance - they need tremendous graphics performance. More importantly they need efficient graphics performance at a rock-bottom price, and that means specialized processors that are designed to do one thing and do it well. A game console CPU would make a horrible general-purpose CPU.
Also, a game console has to stick with its CPU for years at the same speed. The product life cycle of a console CPU is spent reducing the bulk per-unit price, not increasing performance. During the PSone's lifetime, for example, the hardware evolved into something completely differnet, as increased miniaturization and better manufacturing processes allowed the consolidation of all the hardware of the original PSX into the tiny little chassis of the PSone. This happens with all game systems.
CPU performance isn't all that matters in a game console. Price is just as important, as is ease of development. The PS2 is said to be unpleasant to program for - easier development times mean more incentive for game programmers to port their games on over to Sony's system.
Do you have any evidence of this? SPEC scores etc.?
No, I simply assumed that they would want a CPU that can out-perform the Intel processor in the XBox. Obviously you know something about this market, and I do not. Also, it shows that rules from one field don't necessarily apply to another.
So the PS/3 processor may not be that swift? Too bad.
I have no idea how well the PS3's CPU will perform - it's just that I bet that it will make a fantastic game console CPU but would make a fairly crappy desktop computer CPU.
If the aim is to beat the CPU used in the Xbox, that really shouldn't be too difficult. Although I haven't seen any benchmark results, I would imagine the Celeron in the Xbox is actually inferior to its competition as a console CPU, even though it is clocked much higher, simply because it is not specialized to its task. Same reason that to achieve the same quality of graphics as on a console, you have to spend at least 3x more money on a PC...
I suppose Apple could actually use this to its advantage. By embracing the tremendous vector speed of a console chip (or a chip founded on the same principles) it would be possible to deliver a chip that delivers mediocre integer performance, but outstanding multimedia performance - much like the G4 and AltiVec do today. If Apple can't compete with x86 on x86's terms, perhaps it would be a good idea to slant the playing field, instead of totally levelling it (by selling x86 Macs) as some suggest. I wonder if Apple has thought about specialized co-processors...