But all the HPC people are nervous as Intel doesn't have something that can replace AltiVec at the moment.
This is a question that I was wondering about myself.
I was reading this info from the register, and thinking about the possible advances that Intel can make regarding the line-up. I then had a look at this line:
Quote:
But it's clear from the tone of Gelsinger's comments that the Pentium M architecture is going through something of a complex redesign to incorporate these features, which is possibly why Intel is reserving it for second- rather than first-generation dual-core processors.
It has always been rumoured that Intel have been trying to make their processors more elegant, however this is hampered somewhat by the entrenchment that M$ is committed to due the software dependencies. Is it possible that in addition to their standard chips, Intel may be considering a small tweak to the processor line (stripping out the 'dead wood' if you will) to accomodate Apple and give a boost to the architecture?
Maybe even develop a new VMX unit that is compatible with Altivec?
(I don't know anything about the practical implications of this.)
Yes that was me on a bad day. I finally took a look at forthcoming Intel products and had to agree. While the 970MP would have been competitive enough at the desktop level IBM simply didn't have an answer for laptops and eventually Apple's product mix will likely be 70/30 portable to desktop sales ratio.
I'm not overly concerned with optimation tools. Apple is more serious about Xcode than I ever thought they would be. It is a distant memory of yesteryear when Apple used to charge big money for dev tools.
I'd love to see a new SIMD implementation from Intel that is every bit the equal of Altivec. Who knows maybe it'll happen. Apple will definitely support it because they love offering that extra something that the LCD PC market doesn't want to.
I'd love to see a new SIMD implementation from Intel that is every bit the equal of Altivec. Who knows maybe it'll happen. Apple will definitely support it because they love offering that extra something that the LCD PC market doesn't want to.
And yet, even if they came up with a killer Altivec replacement, it would mean that MS Longhorn and Windows programs have access to the same instruction set--so it no longer gives Apple any competitive advantage. That's the dual edge nature of this move. It guarantees parity, but it also guarantees we'll never have anything more than parity.
Apple does own a lot of the IP to Altivec, though. I wonder if they'll share with Intel.
Shrinking? Where on earth did you get that idea from?
Largest WWDC attendance ever - over 500,000 ADC members - more and more enterprise software appearing on Mac OS X.
Shrinking?
Watch again. He only said it was the largest in a decade. I worked on WordPerfect for the Mac starting in the early 90's, and can remember it being much larger. It's also skewed because Apple backed out of the summer MacWorld, so I know a lot of people who go to WWDC now that would never have attended a decade, or even five years ago. Now, it's the place to go and do business in the Mac world weather you are a developer or not.
Try finding a job as a Mac developer, then you will see what I'm talking about. At least around here, there is almost nothing left.
Yes but PC developers can be pretty lazy. I've seen the SSE and SSE2 get virtually ignored by many. Apple is also likely to incorporate the SIMD into their OS.
Microsoft has generally refused to add any specific SIMD support in Windows because they don't want to link themselves to one processor. Apple can take that risk if the reward is good enough.
Watch again. He only said it was the largest in a decade.
I was there! I know what he said. He said that he don't remember if it has been bigger.
Anyway, the numbers of ADC members are growing, and you're seeing new apps coming to the platform almost every week.
A japanese bank recently chose to go the Apple route, and they convinced TIBCO to port all their apps to Mac OS X which they are currently in the process of doing.
And yet, even if they came up with a killer Altivec replacement, it would mean that MS Longhorn and Windows programs have access to the same instruction set--so it no longer gives Apple any competitive advantage. That's the dual edge nature of this move. ...
Or, since apple is only a prestige custumer to intel -
apple could have convinced intel to design custom chips
with all the trimmings ... just for prestige. Possible?
As for the kernel, there's no reason why that couldn't be improved - either into a true Mach microkernel, or with other improvements. Apple will find that sufficiently important to address on the server side at some point. In any case, I doubt it's a show-stopper for anyone.
I was hoping for low-power, high-speed Cells, and going to 32-bit old-technology Intel chips is definitely a step backward. I don't think I can contemplate buying an Intel Mac, nor can I consider buying an obsolescent PPC Mac either. (You guys do know the difference between "obsolescent" and "obsolete" I hope.) Maybe Steve will recant his "transition complete by 2007" statement and Cell-based Macs will come out and be supported in parallel. What a weird world that would be...
I don't think I can contemplate buying an Intel Mac, nor can I consider buying an obsolescent PPC Mac either.
So what are you going to buy? It sounds like you have given up on computers all together (unless you would rather buy an Intel-windows box rather than an Intel-mac box, which would be strange...)
So what are you going to buy? It sounds like you have given up on computers all together (unless you would rather buy an Intel-windows box rather than an Intel-mac box, which would be strange...)
Hey, there's always Linux on a AMD machine.
Personally, I'll consider an Apple/Intel once I get a chance to see them. Until I have that chance, I'll hold with the Macs I've got.
I was there! I know what he said. He said that he don't remember if it has been bigger.
The exact quote, "It may be the largest ever, but I know its the largest in the last decade."
Jobs was at Next during the early 90's, so he wouldn't have been there at the peak. WWDC seemed bigger and more focused to me then. No big Keynote speeches back then, though.
No thanks! Tried it and hated it (both Red Hat and Mandrake). I don't think that many mac users would be happy with Linux (but many Linux users like the mac) because Linux is a lot more involved and non-intuitive.
No thanks! Tried it and hated it (both Red Hat and Mandrake). I don't think that many mac users would be happy with Linux (but many Linux users like the mac) because Linux is a lot more involved and non-intuitive.
Hence the smiley. I've been working on getting some multimedia training working on Linux. It has promise, but mostly it just drives me completely insane. Everything almost works, at least if you know how, but intuitive it is not.
As for the kernel, there's no reason why that couldn't be improved - either into a true Mach microkernel, or with other improvements. Apple will find that sufficiently important to address on the server side at some point. In any case, I doubt it's a show-stopper for anyone.
I was hoping for low-power, high-speed Cells, and going to 32-bit old-technology Intel chips is definitely a step backward. I don't think I can contemplate buying an Intel Mac, nor can I consider buying an obsolescent PPC Mac either. (You guys do know the difference between "obsolescent" and "obsolete" I hope.) Maybe Steve will recant his "transition complete by 2007" statement and Cell-based Macs will come out and be supported in parallel. What a weird world that would be...
Where did Apple or Intel say they were going to be 32-bit? The only thing we know is that the development box is running a 3.6ghz Pentium 4. We don't even know if its the 560, 560J, or 660.
This move to Intel seems strange to most Mac users (BASED ON CURRENT TECHNOLOGY). Hopefully the two companies will develop a new chip together (knowing Apple, they will). Like: "the new Apple PowerMac with an Intel X-TRON 6.2 MHz inside...." maybe.
One thing I do agree with: It's going to be hard for them to sell PowerPC Macs until the Intel switch. I don't wanna buy a new Mac thats going to be useless in 2 years (am I missing something there? that just seems to be common sense).
One thing is obvious - Apple needs to be more specific about this move to Intel. A lot of people are freaking out.
Apple-"Come over to Mac OS Leopard and get all these new whizzbang features!!" .
Switcher- "Great and I have a P4 3.8Ghz computer! How much is Leopard I want to install it today"
Apple- "uh sorry...Leopard won't run on your computer but we have a shiny new P4 4.2Ghz for you!"
Switcher- "but...they are the same platform. Why should I buy a new computer to run your OS on the same hardware?"
Apple- "Hey I gotta eat man"
Switcher- "Wow a $400 premium you must be eating filet mignon"
This stinks folks.
Yeah, and so today it's,
Apple: Come buy a Powerbook!
Switcher: But they're only at 1.7 GHz. Isn't that kinda slow?
Apple: Naw, that's a myth. MHz don't mean anything.
Switcher: <tries the powerbook out a bit> Well it sure seems slow. Much slower than my Sony Vaio. I think I'll stick to the Vaio.
Apple: No, wait! The Powerbook has the Velocity Engine! You must understand the Velocity Engine, it is the key to the Powerbooks great speed!
Switcher: <walks away>
No more of this. Apple is never behind Wintels again. They're never ahead of Wintels, but in the history of PPC, Apple has been behind about 95% of the time, so it seems like a good trade off to just switch to x86.
Switcher: But they're only at 1.7 GHz. Isn't that kinda slow?
Apple: Naw, that's a myth. MHz don't mean anything.
Switcher: <tries the powerbook out a bit> Well it sure seems slow. Much slower than my Sony Vaio. I think I'll stick to the Vaio.
Apple: No, wait! The Powerbook has the Velocity Engine! You must understand the Velocity Engine, it is the key to the Powerbooks great speed!
Switcher: <walks away>
No more of this. Apple is never behind Wintels again. They're never ahead of Wintels, but in the history of PPC, Apple has been behind about 95% of the time, so it seems like a good trade off to just switch to x86.
Exactly.
With the switch, the question of performance will become irrelevant, and people will focus on others subjects where Apple brings innovation :
- design
- the mac os X user experience.
- quality
It will be a mess, in the next coming two years, but after that it will be great. May be I will buy one of the first centrino powerbook
Comments
Originally posted by JLL
But all the HPC people are nervous as Intel doesn't have something that can replace AltiVec at the moment.
This is a question that I was wondering about myself.
I was reading this info from the register, and thinking about the possible advances that Intel can make regarding the line-up. I then had a look at this line:
But it's clear from the tone of Gelsinger's comments that the Pentium M architecture is going through something of a complex redesign to incorporate these features, which is possibly why Intel is reserving it for second- rather than first-generation dual-core processors.
It has always been rumoured that Intel have been trying to make their processors more elegant, however this is hampered somewhat by the entrenchment that M$ is committed to due the software dependencies. Is it possible that in addition to their standard chips, Intel may be considering a small tweak to the processor line (stripping out the 'dead wood' if you will) to accomodate Apple and give a boost to the architecture?
Maybe even develop a new VMX unit that is compatible with Altivec?
(I don't know anything about the practical implications of this.)
A bit melodramatic to my taste.
Anyway, and finally it all boils down to the
processor roadmap intel, ibm and others
explained to apple.
Yes that was me on a bad day. I finally took a look at forthcoming Intel products and had to agree. While the 970MP would have been competitive enough at the desktop level IBM simply didn't have an answer for laptops and eventually Apple's product mix will likely be 70/30 portable to desktop sales ratio.
I'm not overly concerned with optimation tools. Apple is more serious about Xcode than I ever thought they would be. It is a distant memory of yesteryear when Apple used to charge big money for dev tools.
I'd love to see a new SIMD implementation from Intel that is every bit the equal of Altivec. Who knows maybe it'll happen. Apple will definitely support it because they love offering that extra something that the LCD PC market doesn't want to.
Originally posted by hmurchison
Yes that was me on a bad day.
Welcome back, thoughtful and salient hmurchison
Originally posted by hmurchison
I'd love to see a new SIMD implementation from Intel that is every bit the equal of Altivec. Who knows maybe it'll happen. Apple will definitely support it because they love offering that extra something that the LCD PC market doesn't want to.
And yet, even if they came up with a killer Altivec replacement, it would mean that MS Longhorn and Windows programs have access to the same instruction set--so it no longer gives Apple any competitive advantage. That's the dual edge nature of this move. It guarantees parity, but it also guarantees we'll never have anything more than parity.
Apple does own a lot of the IP to Altivec, though. I wonder if they'll share with Intel.
Originally posted by JLL
Shrinking? Where on earth did you get that idea from?
Largest WWDC attendance ever - over 500,000 ADC members - more and more enterprise software appearing on Mac OS X.
Shrinking?
Watch again. He only said it was the largest in a decade. I worked on WordPerfect for the Mac starting in the early 90's, and can remember it being much larger. It's also skewed because Apple backed out of the summer MacWorld, so I know a lot of people who go to WWDC now that would never have attended a decade, or even five years ago. Now, it's the place to go and do business in the Mac world weather you are a developer or not.
Try finding a job as a Mac developer, then you will see what I'm talking about. At least around here, there is almost nothing left.
Microsoft has generally refused to add any specific SIMD support in Windows because they don't want to link themselves to one processor. Apple can take that risk if the reward is good enough.
Originally posted by D.J. Adequate
Watch again. He only said it was the largest in a decade.
I was there! I know what he said. He said that he don't remember if it has been bigger.
Anyway, the numbers of ADC members are growing, and you're seeing new apps coming to the platform almost every week.
A japanese bank recently chose to go the Apple route, and they convinced TIBCO to port all their apps to Mac OS X which they are currently in the process of doing.
Originally posted by DaveLee
Welcome back, thoughtful and salient hmurchison
Indeed.
Originally posted by D.J. Adequate
And yet, even if they came up with a killer Altivec replacement, it would mean that MS Longhorn and Windows programs have access to the same instruction set--so it no longer gives Apple any competitive advantage. That's the dual edge nature of this move. ...
Or, since apple is only a prestige custumer to intel -
apple could have convinced intel to design custom chips
with all the trimmings ... just for prestige. Possible?
I was hoping for low-power, high-speed Cells, and going to 32-bit old-technology Intel chips is definitely a step backward. I don't think I can contemplate buying an Intel Mac, nor can I consider buying an obsolescent PPC Mac either. (You guys do know the difference between "obsolescent" and "obsolete" I hope.) Maybe Steve will recant his "transition complete by 2007" statement and Cell-based Macs will come out and be supported in parallel. What a weird world that would be...
I don't think I can contemplate buying an Intel Mac, nor can I consider buying an obsolescent PPC Mac either.
So what are you going to buy? It sounds like you have given up on computers all together (unless you would rather buy an Intel-windows box rather than an Intel-mac box, which would be strange...)
Originally posted by e1618978
So what are you going to buy? It sounds like you have given up on computers all together (unless you would rather buy an Intel-windows box rather than an Intel-mac box, which would be strange...)
Hey, there's always Linux on a AMD machine.
Personally, I'll consider an Apple/Intel once I get a chance to see them. Until I have that chance, I'll hold with the Macs I've got.
Originally posted by onlooker
...
Not to exclude Adobe, or any of the other great developers. They all helped make the Mac what it is.
The opposite around. Without Apple Adobe (and other
developers) wouldn't exist. Adobe grew up with Apple.
Originally posted by JLL
I was there! I know what he said. He said that he don't remember if it has been bigger.
The exact quote, "It may be the largest ever, but I know its the largest in the last decade."
Jobs was at Next during the early 90's, so he wouldn't have been there at the peak. WWDC seemed bigger and more focused to me then. No big Keynote speeches back then, though.
Hey, there's always Linux on a AMD machine.
No thanks! Tried it and hated it (both Red Hat and Mandrake). I don't think that many mac users would be happy with Linux (but many Linux users like the mac) because Linux is a lot more involved and non-intuitive.
Originally posted by e1618978
No thanks! Tried it and hated it (both Red Hat and Mandrake). I don't think that many mac users would be happy with Linux (but many Linux users like the mac) because Linux is a lot more involved and non-intuitive.
Hence the smiley. I've been working on getting some multimedia training working on Linux. It has promise, but mostly it just drives me completely insane. Everything almost works, at least if you know how, but intuitive it is not.
Originally posted by cubist
As for the kernel, there's no reason why that couldn't be improved - either into a true Mach microkernel, or with other improvements. Apple will find that sufficiently important to address on the server side at some point. In any case, I doubt it's a show-stopper for anyone.
I was hoping for low-power, high-speed Cells, and going to 32-bit old-technology Intel chips is definitely a step backward. I don't think I can contemplate buying an Intel Mac, nor can I consider buying an obsolescent PPC Mac either. (You guys do know the difference between "obsolescent" and "obsolete" I hope.) Maybe Steve will recant his "transition complete by 2007" statement and Cell-based Macs will come out and be supported in parallel. What a weird world that would be...
Where did Apple or Intel say they were going to be 32-bit? The only thing we know is that the development box is running a 3.6ghz Pentium 4. We don't even know if its the 560, 560J, or 660.
This move to Intel seems strange to most Mac users (BASED ON CURRENT TECHNOLOGY). Hopefully the two companies will develop a new chip together (knowing Apple, they will). Like: "the new Apple PowerMac with an Intel X-TRON 6.2 MHz inside...." maybe.
One thing I do agree with: It's going to be hard for them to sell PowerPC Macs until the Intel switch. I don't wanna buy a new Mac thats going to be useless in 2 years (am I missing something there? that just seems to be common sense).
One thing is obvious - Apple needs to be more specific about this move to Intel. A lot of people are freaking out.
-BUBBA
Originally posted by hmurchison
I can see Apple trying to woo a switcher in 2007.
Apple-"Come over to Mac OS Leopard and get all these new whizzbang features!!" .
Switcher- "Great and I have a P4 3.8Ghz computer! How much is Leopard I want to install it today"
Apple- "uh sorry...Leopard won't run on your computer but we have a shiny new P4 4.2Ghz for you!"
Switcher- "but...they are the same platform. Why should I buy a new computer to run your OS on the same hardware?"
Apple- "Hey I gotta eat man"
Switcher- "Wow a $400 premium you must be eating filet mignon"
This stinks folks.
Yeah, and so today it's,
Apple: Come buy a Powerbook!
Switcher: But they're only at 1.7 GHz. Isn't that kinda slow?
Apple: Naw, that's a myth. MHz don't mean anything.
Switcher: <tries the powerbook out a bit> Well it sure seems slow. Much slower than my Sony Vaio. I think I'll stick to the Vaio.
Apple: No, wait! The Powerbook has the Velocity Engine! You must understand the Velocity Engine, it is the key to the Powerbooks great speed!
Switcher: <walks away>
No more of this. Apple is never behind Wintels again. They're never ahead of Wintels, but in the history of PPC, Apple has been behind about 95% of the time, so it seems like a good trade off to just switch to x86.
Originally posted by Junkyard Dawg
Yeah, and so today it's,
Apple: Come buy a Powerbook!
Switcher: But they're only at 1.7 GHz. Isn't that kinda slow?
Apple: Naw, that's a myth. MHz don't mean anything.
Switcher: <tries the powerbook out a bit> Well it sure seems slow. Much slower than my Sony Vaio. I think I'll stick to the Vaio.
Apple: No, wait! The Powerbook has the Velocity Engine! You must understand the Velocity Engine, it is the key to the Powerbooks great speed!
Switcher: <walks away>
No more of this. Apple is never behind Wintels again. They're never ahead of Wintels, but in the history of PPC, Apple has been behind about 95% of the time, so it seems like a good trade off to just switch to x86.
Exactly.
With the switch, the question of performance will become irrelevant, and people will focus on others subjects where Apple brings innovation :
- design
- the mac os X user experience.
- quality
It will be a mess, in the next coming two years, but after that it will be great. May be I will buy one of the first centrino powerbook