Regarding IBM, they may be shifting focus but that does not mean they are leaving the hardware behind them.
<a href="http://www.macedition.com/nmr/nmr_20021126.php"; target="_blank">NMR</a> just posted a peek into IBM's future and the outlook is low-end and blade servers, using chips that will scale to 6GHz but run cooler than Intel's offerings.
OS X will be released for the x86 platforms next year, regardless what the processor road map is for Apple systems.
How could this cannibalize hardware sales, when there is a limited amount of hardware sales industry wide now? Think of the implications of releasing a fresh new UNIX-based OS for all of those corporate machines that are currently not being replaced by CIOs. Think of all the consumers who have no need or desire to upgrade their hardware. These are the same CIOs that are looking for an alternative to the steep licensing fees that they are paying for MS desktop applications. These are the same consumers who would easily spend $129 to update their hardware with an OS that provides them with all of the features OS X gives them.
Apple would make far more money replacing the existing OS's with OS X in these two markets alone than they would EVER lose in hardware revenue, period.
<strong>Further thought for all this Intel speed comparison.
Intel is shortly gonna have to swallow its own advertising.
The Itanium runs at a much lower clock speed than the current P4. They are even begining to sell the name 'Intel' and not so much the clock speed. Dell, et al, are still doing a mighty job of it though.
Long story short. The 970 will begin @ about 1.2 and scale from there. In that time frame Intel will need to start really pushing the Itanium. It currently stands in prototype form of 1Ghz. If they don't they are gonna be in a crunch. The P4 will not get bigger for ever.
<strong>Apple would make far more money replacing the existing OS's with OS X in these two markets alone than they would EVER lose in hardware revenue, period.
It will happen.</strong><hr></blockquote>
That is your opinion. So, instead of buying $4000 Mac computers, people will buy $129 Mac OS X + a PC.
So for these people, Apple will end up losing money with the switch. (profit margins on a $4000 computer being more than $129).
The question is whether enough people who would have never bought a mac will buy OS X. (and not pirate it). This is open for debate... and that's why it's not clearly a win for Apple to switch to an OS-only company.
That is your opinion. So, instead of buying $4000 Mac computers, people will buy $129 Mac OS X + a PC.
...
BeOS tried, and failed...
IBM tried, and failed (OS/2)...
NeXTSTep tried, and failed...
</strong><hr></blockquote>
I don't think it would be '$129'. Either the client version would be sold for more like $200, or it would be the Server version only for $500. By restricting it to the Server version, you are targeting just the people that are a little antsy about the Doze-server client license prices. (AKA STEEP).
One other issue is that MOSX has roughly ZERO copy protection/serial number verification built in. Doze is far more restrictive. Apple is able to survive a fairly robust level of piracy of OSX on _their_ hardware... because they already made _some_ money on the hardware itself. If OSX-86 was released with the same level of copy protection as OSX-ppc has right now, they'd sell a few copies before it hit the p2p net, and not so many after that. The fact that Darwin, the underpinnings of OSX, is available as source makes it tough to guarantee (or come anywhere near guaranteeing) that OSX-86 would be locked to the specific boxes it is licensed to.
The blade server CPU going towards 6 GHz in the future sound nice if they end up in some Mac towers as well <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
If Apple can scrap all Motorola CPUs by using the bottom of the barrel for the books and iMacs even 1.0 to 1.2 GHz single 970 would be nice. Apple have been raving about the Velocity engine since 1999 and they still sell computers without it <img src="graemlins/oyvey.gif" border="0" alt="[No]" />
So what if 7457 support a 200 MHz bus? I would like to see a comparison with a old G4 AGP upgraded with a dual 1GHz on its 100 MHz run alongside a current dual 1GHz (with a 167MHz bus). My guess is that the performance difference is marginal. The G4 does not need a new bus it needs to get on the other side of 2GHz.
Every 970 that Apple buy increase the importance they have for IBM and is a reason to work on the 9x0 line. Having a high end 970 a midrange G4+ and a low end of a G3 with Velocity Engine or other will pour money in CPUs that is not Apples future. The 970 may be the future and getting the bulk CPUs sales (iMac eMac, PB, iBook) in there is far better than dead ends. I assume that nobody assume that the future CPU for Apple is a G3 from IBM with a Altivec unit grafted on or a anything from Motorola <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
If Apple go for the 970 it is in their own interest to generate as much revenues for the 970 as possible.
Also the line up would be more easy to understand. We alredy have the problem with two Apple products with the same speedrating (MHz) is the same speed or very different to to +/-velocity engine. Adding a third CPU would make it even worse.
I hope that the G3 is out early next year (If Velocity Engine is the best thing since sliced bread why sell computers without it year after year?) Then in less than a year the G4 starts to be replaced by the 970. Replacing aggressively that is replacing even if the performance is only on par with the G4 just to invest in the new CPU! They did that in 1999 replacing the G3/350-400-450 with G4/350-400-450. Apple can not afford to repeat that experience <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
[quote]Intel is shortly gonna have to swallow its own advertising.
The Itanium runs at a much lower clock speed than the current P4.<hr></blockquote>
Intel don't sell Itanium boxen to consumers--where the GHZ speed play works. Better to see what happens when Banias comes out and they sell lower or equal clock speed, but much improved battery life.
I was reading through the posts in this thread and there are a couple of misunderstandings.
1) The AMD Opteron CPU has NOT been renamed to the Athlon 64. They are two separate CPUs. The Opteron is for Servers and the Athlon 64 is for desktop/mobile systems.
2) The PowerPC 970 CPU should hold its own just fine when it debuts. Many people in this thread are saying that at 1.8GHz, it will be too slow. You can't compare a 1.8GHz 64-bit PowerPC CPU to a 3.0GHz 32-bit Intel based CPU. Apples to oranges, big time. Look at all of the current 64-bit CPU's on the market. They all top out at just over 1GHz currently. So if IBM can ship this thing towards the middle of 2003 at the advertised 1.8GHz, it will be one of the fastest 64-bit CPU's out there.
That said, I really hope that IBM/Apple can get this chip out the door asap! Although, an Opteron based Mac WOULD be interesing!
I assume that nobody assume that the future CPU for Apple is a G3 from IBM with a Altivec unit grafted on...
<hr></blockquote>
This is exactly what I've been hoping for in early 2003. If IBM can mass produce these at over 1 GHz -- perhaps multi-core -- the power, cost, and size would be perfect for most of Apple's line. This would include iMac, eMac, iBook and perhaps even PowerBook. Only the Server and PowerMac should immediately get the 970 once it is available.
Sorry, but that just doesn't follow. Revisions explain why manufacturers push back a target date, but it doesn't explain a target date announcement that far away. For IBM to announce general distribution in 2H '03 would mean they either just taped out or expect to at the end of the year or so.
It just doesn't take a year to go from tape out to production.
Somebody has to be fibbing here.</strong><hr></blockquote>
Well, in one of my earlier posts on this thread, I said Apple has had samples since mid March of this year. I neglected to say that they got a second batch of samples in late July (these being a second revision of the chip), and that they're due to get a third batch very early next year. Now, I'm not saying that this last batch won't be just for various certifications and/or possibly sold as the acual item. After all, the project is supposed to end around the end of July or early August '03.
This is exactly what I've been hoping for in early 2003. If IBM can mass produce these at over 1 GHz -- perhaps multi-core -- the power, cost, and size would be perfect for most of Apple's line. This would include iMac, eMac, iBook and perhaps even PowerBook. . .
</strong><hr></blockquote>
A G3 with Alti-Vec added would essentially have G4 capabilities for single processor systems. It would run any software that requires a G4, so why not just call it that?
Another option would be to simplify the 970, maybe even a 32-bit version, and use that for consumer products. If the 970 dissipates just 19 Watts at 1.2 GHz, think what a simple version of the chip would do on the 90 nanometer process. Why do it? Well, Apple would have the same processor bus configuration for all models of the Mac, and other possible advantages of the 970 design.
No, blame the music/film industry who despite a complete lack of evidence wishes to blame the consumer for the drop in CD sales, and a drop in movie attendance.
</strong>
<hr></blockquote>
Actually, in the UK anyway, cinema attendance is up, and I believe CD sales are too. But that's not the issue. The issue is that music piracy has become very visible, and there's no way that people "sharing" music also go out and buy it - it just doesn't happen.
[quote]Originally posted by cowerd:
<strong>Anytime an industry acts with the assumption that all their customers are engaged in criminal behavior, that is a good indication that the industry is intellectually bereft of ideas, unable to deal with the future and morally suspect.</strong>
<hr></blockquote>
I don't think that's really true. The music industry doesn't think all of its customers are pirates, but they do, correctly I think, see a lot of piracy that's very visible. Even if that only amounts to 5% of their turnover (quite likely a lot more than that), that's 5% that the rest of us have to subsidise. They want to try to prevent the piracy, and in doing so they "hurt" some of their customers too. If the piracy wasn't so blatent then the means to prevent it would be less invasive.
[quote]Originally posted by cowerd:
<strong>These are the same people that said that VCR's would kill the industry (it instead opened up a lucrative post-theater sales and licensing market) and who pushed digital (CD's) to replace vinyl because they saw more profits in CD sales. There is still no direct evidence that mp3 downloads kill CD sales--crappy RIAA sponsored studies notwithstanding.</strong>
<hr></blockquote>
VCR, and TV, had a huge impact on cinema attendances, to say otherwise ignores the obvious facts.
I think it should also be obvious to anyone that once someone has a pirated version of any material it's unlikely that they will ever buy an original. I'm not pretending to be an Angel here, I've got a few tapes people have given me over the years, and live concerts I've recorded from the radio, why would I go and buy another copy?
It's naive to believe that people are basically honest about such things, which seems to be your premise. People buy CDs or DVDs because that's the only way to get a copy. Once the physicality of possessing an item breaks down then people no longer need to seek out a piece of merchandise, they download it for free if they can.
[quote]Originally posted by cowerd:
<strong>Consumers are sheep up until a certain point, if that were not the case MS would not have made substantial DRM changes to the XP Media OS. The consumer, in general, still believes in the tangibility of goods and purchase as ownership. Anything that messes with that basic assumption will meet with failure.</strong>
<hr></blockquote>
How can they believe in the tangibility of goods that they can never hold?
You've already written that MS and the music film industry will work hand in hand, there will be no other standard, there will be no other way to buy music, at least in the mainstream, digital rights management will be adopted, because there will be no choice.
Apple with have no choice other than to adopt that same standard, or forget all about the digital hub.
<strong>and there's no way that people "sharing" music also go out and buy it - it just doesn't happen
</strong><hr></blockquote>
Actually a friend of mineis doing just that. He does use sharing software to get the latest music - but he also always then goes and buy the CD. Why? Because he can not bother to leave the PC on all the time and he can't take his mp3 collection to work and not fit it into the CD player of his car. And he likes buying CDs.
So - people who were buying music before are still doing it. The music industry only has a problem because they spend too much fscking money on advertising sheety music that then doesn't sell well. Red Hot Chilli Peppers never had a problem selling their stuff - because people like it.
Well, in one of my earlier posts on this thread, I said Apple has had samples since mid March of this year. I neglected to say that they got a second batch of samples in late July (these being a second revision of the chip), and that they're due to get a third batch very early next year. Now, I'm not saying that this last batch won't be just for various certifications and/or possibly sold as the acual item. After all, the project is supposed to end around the end of July or early August '03.</strong><hr></blockquote>
Uhmm, no. Fabrication of a chip is not a project with start and end dates. It's an ongoing process with target dates for acheiving certain milestones. Milestones like a profitable yield. Once a design is taped out, the only revisions you make are in aid of improving yields. Usually this isn't going to require more than one or two sampling runs to work out kinks in your process. (Unless the design is so fubar'ed you can't get decent yields with your process.)
You then work on tweaking the design to acheive other targets (like expected clockspeeds) and to iron out any bugs.
So if IBM has already provided two sets of samples to Apple they should be close to acheiving production level yields. And if Apple has had samples this long, they should have a chipset by now and be ready to ramp up when they get final product.
A G3 with Alti-Vec added would essentially have G4 capabilities for single processor systems. It would run any software that requires a G4, so why not just call it that?
</strong><hr></blockquote>
because the upcoming G3 from IBM will have DDR-Ram support and Rapid-IO and will be multicore superscalar...? (and this all besides the fact that it'll sport a SIMD unit)
Uhmm, no. Fabrication of a chip is not a project with start and end dates. It's an ongoing process with target dates for acheiving certain milestones. Milestones like a profitable yield. Once a design is taped out, the only revisions you make are in aid of improving yields. Usually this isn't going to require more than one or two sampling runs to work out kinks in your process. (Unless the design is so fubar'ed you can't get decent yields with your process.)
You then work on tweaking the design to acheive other targets (like expected clockspeeds) and to iron out any bugs.
So if IBM has already provided two sets of samples to Apple they should be close to acheiving production level yields. And if Apple has had samples this long, they should have a chipset by now and be ready to ramp up when they get final product.
So someone is still fibbing.</strong><hr></blockquote>
You're right about "target dates", that's all schedules realy are. I used the term 'project', because that's the term someone else used to describe the development schedule (the person that created it actually). Granted, there can be "other projects" to enhance processors after the initial development has been completed. IBM is very Project, Milestone, and process driven ... and yes, their projects do have estimated completion dates. They're NOT "never-ending-stories".
Apple started work on the chipset around the end of 2000 ... beginning of 2001
As far as production, IBM has yet to test full automated production on 300mm wafers. They've only just completed production tests in full manual mode.
Fibbing, not really. Just remember back to one of the original "Star Trek" episodes where Scotty advised expanding the length of time to complete repairs to a crew-mate, so that he could beat the time and look good (under normal circumstances), or give himself a cushion in case he ran into problems.
because the upcoming G3 from IBM will have DDR-Ram support and Rapid-IO and will be multicore superscalar...? (and this all besides the fact that it'll sport a SIMD unit)
edit: i'd call this one a "G5" and the 970 "G6"
</strong><hr></blockquote>
Uh, we are talking about different things. I was responding to the usual request that IBM simply add Alti-Vec (an SIMD unit) to the G3 and use that for consumer Macs. I simply pointed out that it would then run any software that the G4 could run, so why not call it a G4 so consumers would not be confused about its capabilities. Apple could use these only in single processor Macs, however.
What I pointed out in the second half of the reply was an alternative. If IBM is going to the trouble of taping out a new processor, why not base it on the 970, but leave things out that are unnecessary for consumer Macs? (An ultra lite ultra lite so to speak) The advantage of doing this is that some beneficial things for Apple could be left in. It may be advantageous to have certain similarities between all Mac processors.
If Apple departs from the G3, G4 naming sequence, the consumer chip could be X1 and the 970 could be X2. Use anything you wish in place of the X. It would be made known that the X1 is like a G4 plus, and the X2 is the 64-bit super chip.
Uh, we are talking about different things. I was responding to the usual request that IBM simply add Alti-Vec (an SIMD unit) to the G3 and use that for consumer Macs. I simply pointed out that it would then run any software that the G4 could run, so why not call it a G4 so consumers would not be confused about its capabilities. Apple could use these only in single processor Macs, however.
What I pointed out in the second half of the reply was an alternative. If IBM is going to the trouble of taping out a new processor, why not base it on the 970, but leave things out that are unnecessary for consumer Macs? (An ultra lite ultra lite so to speak) The advantage of doing this is that some beneficial things for Apple could be left in. It may be advantageous to have certain similarities between all Mac processors.
If Apple departs from the G3, G4 naming sequence, the consumer chip could be X1 and the 970 could be X2. Use anything you wish in place of the X. It would be made known that the X1 is like a G4 plus, and the X2 is the 64-bit super chip.</strong><hr></blockquote>
well,sorry - i just wanted to point out that there'll never be a ibm-G3 with just a simd unit added and nothing more. they'll make a generation-switch next year. one of the resulting products will be the mentioned G3successor and the other is the 970. two cpus are ok i think. but the X1, X2 naming convention is a step backwards in my opinion. i like the G5/6 better or even something like G4-X/G5-X64 ? ... i don't know...
the only thing i really have on my wish list is that at least one of those two new cpus will be out in february ...
Comments
<a href="http://www.macedition.com/nmr/nmr_20021126.php"; target="_blank">NMR</a> just posted a peek into IBM's future and the outlook is low-end and blade servers, using chips that will scale to 6GHz but run cooler than Intel's offerings.
6GHz Xserve in a few years, tasty.
Screed
[ 11-26-2002: Message edited by: sCreeD ]</p>
How could this cannibalize hardware sales, when there is a limited amount of hardware sales industry wide now? Think of the implications of releasing a fresh new UNIX-based OS for all of those corporate machines that are currently not being replaced by CIOs. Think of all the consumers who have no need or desire to upgrade their hardware. These are the same CIOs that are looking for an alternative to the steep licensing fees that they are paying for MS desktop applications. These are the same consumers who would easily spend $129 to update their hardware with an OS that provides them with all of the features OS X gives them.
Apple would make far more money replacing the existing OS's with OS X in these two markets alone than they would EVER lose in hardware revenue, period.
It will happen.
<strong>Further thought for all this Intel speed comparison.
Intel is shortly gonna have to swallow its own advertising.
The Itanium runs at a much lower clock speed than the current P4. They are even begining to sell the name 'Intel' and not so much the clock speed. Dell, et al, are still doing a mighty job of it though.
Long story short. The 970 will begin @ about 1.2 and scale from there. In that time frame Intel will need to start really pushing the Itanium. It currently stands in prototype form of 1Ghz. If they don't they are gonna be in a crunch. The P4 will not get bigger for ever.
[ 11-26-2002: Message edited by: zaz ]</strong><hr></blockquote>
All the info I've seen says the 970 will start at 1.4ghz-1.8ghz not 1.2ghz.
<strong>Apple would make far more money replacing the existing OS's with OS X in these two markets alone than they would EVER lose in hardware revenue, period.
It will happen.</strong><hr></blockquote>
That is your opinion. So, instead of buying $4000 Mac computers, people will buy $129 Mac OS X + a PC.
So for these people, Apple will end up losing money with the switch. (profit margins on a $4000 computer being more than $129).
The question is whether enough people who would have never bought a mac will buy OS X. (and not pirate it). This is open for debate... and that's why it's not clearly a win for Apple to switch to an OS-only company.
BeOS tried, and failed...
IBM tried, and failed (OS/2)...
NeXTSTep tried, and failed...
arn
<strong>
That is your opinion. So, instead of buying $4000 Mac computers, people will buy $129 Mac OS X + a PC.
...
BeOS tried, and failed...
IBM tried, and failed (OS/2)...
NeXTSTep tried, and failed...
</strong><hr></blockquote>
I don't think it would be '$129'. Either the client version would be sold for more like $200, or it would be the Server version only for $500. By restricting it to the Server version, you are targeting just the people that are a little antsy about the Doze-server client license prices. (AKA STEEP).
One other issue is that MOSX has roughly ZERO copy protection/serial number verification built in. Doze is far more restrictive. Apple is able to survive a fairly robust level of piracy of OSX on _their_ hardware... because they already made _some_ money on the hardware itself. If OSX-86 was released with the same level of copy protection as OSX-ppc has right now, they'd sell a few copies before it hit the p2p net, and not so many after that. The fact that Darwin, the underpinnings of OSX, is available as source makes it tough to guarantee (or come anywhere near guaranteeing) that OSX-86 would be locked to the specific boxes it is licensed to.
If Apple can scrap all Motorola CPUs by using the bottom of the barrel for the books and iMacs even 1.0 to 1.2 GHz single 970 would be nice. Apple have been raving about the Velocity engine since 1999 and they still sell computers without it <img src="graemlins/oyvey.gif" border="0" alt="[No]" />
So what if 7457 support a 200 MHz bus? I would like to see a comparison with a old G4 AGP upgraded with a dual 1GHz on its 100 MHz run alongside a current dual 1GHz (with a 167MHz bus). My guess is that the performance difference is marginal. The G4 does not need a new bus it needs to get on the other side of 2GHz.
Every 970 that Apple buy increase the importance they have for IBM and is a reason to work on the 9x0 line. Having a high end 970 a midrange G4+ and a low end of a G3 with Velocity Engine or other will pour money in CPUs that is not Apples future. The 970 may be the future and getting the bulk CPUs sales (iMac eMac, PB, iBook) in there is far better than dead ends. I assume that nobody assume that the future CPU for Apple is a G3 from IBM with a Altivec unit grafted on or a anything from Motorola <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
If Apple go for the 970 it is in their own interest to generate as much revenues for the 970 as possible.
Also the line up would be more easy to understand. We alredy have the problem with two Apple products with the same speedrating (MHz) is the same speed or very different to to +/-velocity engine. Adding a third CPU would make it even worse.
I hope that the G3 is out early next year (If Velocity Engine is the best thing since sliced bread why sell computers without it year after year?) Then in less than a year the G4 starts to be replaced by the 970. Replacing aggressively that is replacing even if the performance is only on par with the G4 just to invest in the new CPU! They did that in 1999 replacing the G3/350-400-450 with G4/350-400-450. Apple can not afford to repeat that experience <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
[ 11-26-2002: Message edited by: DrBoar ]</p>
The Itanium runs at a much lower clock speed than the current P4.<hr></blockquote>
Intel don't sell Itanium boxen to consumers--where the GHZ speed play works. Better to see what happens when Banias comes out and they sell lower or equal clock speed, but much improved battery life.
I was reading through the posts in this thread and there are a couple of misunderstandings.
1) The AMD Opteron CPU has NOT been renamed to the Athlon 64. They are two separate CPUs. The Opteron is for Servers and the Athlon 64 is for desktop/mobile systems.
2) The PowerPC 970 CPU should hold its own just fine when it debuts. Many people in this thread are saying that at 1.8GHz, it will be too slow. You can't compare a 1.8GHz 64-bit PowerPC CPU to a 3.0GHz 32-bit Intel based CPU. Apples to oranges, big time. Look at all of the current 64-bit CPU's on the market. They all top out at just over 1GHz currently. So if IBM can ship this thing towards the middle of 2003 at the advertised 1.8GHz, it will be one of the fastest 64-bit CPU's out there.
That said, I really hope that IBM/Apple can get this chip out the door asap! Although, an Opteron based Mac WOULD be interesing!
Cheers!
[quote]
I assume that nobody assume that the future CPU for Apple is a G3 from IBM with a Altivec unit grafted on...
<hr></blockquote>
This is exactly what I've been hoping for in early 2003. If IBM can mass produce these at over 1 GHz -- perhaps multi-core -- the power, cost, and size would be perfect for most of Apple's line. This would include iMac, eMac, iBook and perhaps even PowerBook. Only the Server and PowerMac should immediately get the 970 once it is available.
<strong>
Sorry, but that just doesn't follow. Revisions explain why manufacturers push back a target date, but it doesn't explain a target date announcement that far away. For IBM to announce general distribution in 2H '03 would mean they either just taped out or expect to at the end of the year or so.
It just doesn't take a year to go from tape out to production.
Somebody has to be fibbing here.</strong><hr></blockquote>
Well, in one of my earlier posts on this thread, I said Apple has had samples since mid March of this year. I neglected to say that they got a second batch of samples in late July (these being a second revision of the chip), and that they're due to get a third batch very early next year. Now, I'm not saying that this last batch won't be just for various certifications and/or possibly sold as the acual item. After all, the project is supposed to end around the end of July or early August '03.
<strong>From DrBoar:
This is exactly what I've been hoping for in early 2003. If IBM can mass produce these at over 1 GHz -- perhaps multi-core -- the power, cost, and size would be perfect for most of Apple's line. This would include iMac, eMac, iBook and perhaps even PowerBook. . .
</strong><hr></blockquote>
A G3 with Alti-Vec added would essentially have G4 capabilities for single processor systems. It would run any software that requires a G4, so why not just call it that?
Another option would be to simplify the 970, maybe even a 32-bit version, and use that for consumer products. If the 970 dissipates just 19 Watts at 1.2 GHz, think what a simple version of the chip would do on the 90 nanometer process. Why do it? Well, Apple would have the same processor bus configuration for all models of the Mac, and other possible advantages of the 970 design.
<strong>
No, blame the music/film industry who despite a complete lack of evidence wishes to blame the consumer for the drop in CD sales, and a drop in movie attendance.
</strong>
<hr></blockquote>
Actually, in the UK anyway, cinema attendance is up, and I believe CD sales are too. But that's not the issue. The issue is that music piracy has become very visible, and there's no way that people "sharing" music also go out and buy it - it just doesn't happen.
[quote]Originally posted by cowerd:
<strong>Anytime an industry acts with the assumption that all their customers are engaged in criminal behavior, that is a good indication that the industry is intellectually bereft of ideas, unable to deal with the future and morally suspect.</strong>
<hr></blockquote>
I don't think that's really true. The music industry doesn't think all of its customers are pirates, but they do, correctly I think, see a lot of piracy that's very visible. Even if that only amounts to 5% of their turnover (quite likely a lot more than that), that's 5% that the rest of us have to subsidise. They want to try to prevent the piracy, and in doing so they "hurt" some of their customers too. If the piracy wasn't so blatent then the means to prevent it would be less invasive.
[quote]Originally posted by cowerd:
<strong>These are the same people that said that VCR's would kill the industry (it instead opened up a lucrative post-theater sales and licensing market) and who pushed digital (CD's) to replace vinyl because they saw more profits in CD sales. There is still no direct evidence that mp3 downloads kill CD sales--crappy RIAA sponsored studies notwithstanding.</strong>
<hr></blockquote>
VCR, and TV, had a huge impact on cinema attendances, to say otherwise ignores the obvious facts.
I think it should also be obvious to anyone that once someone has a pirated version of any material it's unlikely that they will ever buy an original. I'm not pretending to be an Angel here, I've got a few tapes people have given me over the years, and live concerts I've recorded from the radio, why would I go and buy another copy?
It's naive to believe that people are basically honest about such things, which seems to be your premise. People buy CDs or DVDs because that's the only way to get a copy. Once the physicality of possessing an item breaks down then people no longer need to seek out a piece of merchandise, they download it for free if they can.
[quote]Originally posted by cowerd:
<strong>Consumers are sheep up until a certain point, if that were not the case MS would not have made substantial DRM changes to the XP Media OS. The consumer, in general, still believes in the tangibility of goods and purchase as ownership. Anything that messes with that basic assumption will meet with failure.</strong>
<hr></blockquote>
How can they believe in the tangibility of goods that they can never hold?
You've already written that MS and the music film industry will work hand in hand, there will be no other standard, there will be no other way to buy music, at least in the mainstream, digital rights management will be adopted, because there will be no choice.
Apple with have no choice other than to adopt that same standard, or forget all about the digital hub.
<strong>
601, 603, 603e 603ev, 604, 604e, 620, with 630 (G3) on the way, etc...
Steve</strong><hr></blockquote>
G3 was a "750" IIRC.
<strong>and there's no way that people "sharing" music also go out and buy it - it just doesn't happen
</strong><hr></blockquote>
Actually a friend of mineis doing just that. He does use sharing software to get the latest music - but he also always then goes and buy the CD. Why? Because he can not bother to leave the PC on all the time and he can't take his mp3 collection to work and not fit it into the CD player of his car. And he likes buying CDs.
So - people who were buying music before are still doing it. The music industry only has a problem because they spend too much fscking money on advertising sheety music that then doesn't sell well. Red Hot Chilli Peppers never had a problem selling their stuff - because people like it.
<strong>
Well, in one of my earlier posts on this thread, I said Apple has had samples since mid March of this year. I neglected to say that they got a second batch of samples in late July (these being a second revision of the chip), and that they're due to get a third batch very early next year. Now, I'm not saying that this last batch won't be just for various certifications and/or possibly sold as the acual item. After all, the project is supposed to end around the end of July or early August '03.</strong><hr></blockquote>
Uhmm, no. Fabrication of a chip is not a project with start and end dates. It's an ongoing process with target dates for acheiving certain milestones. Milestones like a profitable yield. Once a design is taped out, the only revisions you make are in aid of improving yields. Usually this isn't going to require more than one or two sampling runs to work out kinks in your process. (Unless the design is so fubar'ed you can't get decent yields with your process.)
You then work on tweaking the design to acheive other targets (like expected clockspeeds) and to iron out any bugs.
So if IBM has already provided two sets of samples to Apple they should be close to acheiving production level yields. And if Apple has had samples this long, they should have a chipset by now and be ready to ramp up when they get final product.
So someone is still fibbing.
<strong>
A G3 with Alti-Vec added would essentially have G4 capabilities for single processor systems. It would run any software that requires a G4, so why not just call it that?
</strong><hr></blockquote>
because the upcoming G3 from IBM will have DDR-Ram support and Rapid-IO and will be multicore superscalar...? (and this all besides the fact that it'll sport a SIMD unit)
edit: i'd call this one a "G5" and the 970 "G6"
[ 11-26-2002: Message edited by: Krassy ]</p>
<strong>
Uhmm, no. Fabrication of a chip is not a project with start and end dates. It's an ongoing process with target dates for acheiving certain milestones. Milestones like a profitable yield. Once a design is taped out, the only revisions you make are in aid of improving yields. Usually this isn't going to require more than one or two sampling runs to work out kinks in your process. (Unless the design is so fubar'ed you can't get decent yields with your process.)
You then work on tweaking the design to acheive other targets (like expected clockspeeds) and to iron out any bugs.
So if IBM has already provided two sets of samples to Apple they should be close to acheiving production level yields. And if Apple has had samples this long, they should have a chipset by now and be ready to ramp up when they get final product.
So someone is still fibbing.</strong><hr></blockquote>
You're right about "target dates", that's all schedules realy are. I used the term 'project', because that's the term someone else used to describe the development schedule (the person that created it actually). Granted, there can be "other projects" to enhance processors after the initial development has been completed. IBM is very Project, Milestone, and process driven ... and yes, their projects do have estimated completion dates. They're NOT "never-ending-stories".
Apple started work on the chipset around the end of 2000 ... beginning of 2001
As far as production, IBM has yet to test full automated production on 300mm wafers. They've only just completed production tests in full manual mode.
Fibbing, not really. Just remember back to one of the original "Star Trek" episodes where Scotty advised expanding the length of time to complete repairs to a crew-mate, so that he could beat the time and look good (under normal circumstances), or give himself a cushion in case he ran into problems.
[ 11-26-2002: Message edited by: MacJedai ]</p>
<strong>
because the upcoming G3 from IBM will have DDR-Ram support and Rapid-IO and will be multicore superscalar...? (and this all besides the fact that it'll sport a SIMD unit)
edit: i'd call this one a "G5" and the 970 "G6"
</strong><hr></blockquote>
Uh, we are talking about different things. I was responding to the usual request that IBM simply add Alti-Vec (an SIMD unit) to the G3 and use that for consumer Macs. I simply pointed out that it would then run any software that the G4 could run, so why not call it a G4 so consumers would not be confused about its capabilities. Apple could use these only in single processor Macs, however.
What I pointed out in the second half of the reply was an alternative. If IBM is going to the trouble of taping out a new processor, why not base it on the 970, but leave things out that are unnecessary for consumer Macs? (An ultra lite ultra lite so to speak) The advantage of doing this is that some beneficial things for Apple could be left in. It may be advantageous to have certain similarities between all Mac processors.
If Apple departs from the G3, G4 naming sequence, the consumer chip could be X1 and the 970 could be X2. Use anything you wish in place of the X. It would be made known that the X1 is like a G4 plus, and the X2 is the 64-bit super chip.
<strong>
Uh, we are talking about different things. I was responding to the usual request that IBM simply add Alti-Vec (an SIMD unit) to the G3 and use that for consumer Macs. I simply pointed out that it would then run any software that the G4 could run, so why not call it a G4 so consumers would not be confused about its capabilities. Apple could use these only in single processor Macs, however.
What I pointed out in the second half of the reply was an alternative. If IBM is going to the trouble of taping out a new processor, why not base it on the 970, but leave things out that are unnecessary for consumer Macs? (An ultra lite ultra lite so to speak) The advantage of doing this is that some beneficial things for Apple could be left in. It may be advantageous to have certain similarities between all Mac processors.
If Apple departs from the G3, G4 naming sequence, the consumer chip could be X1 and the 970 could be X2. Use anything you wish in place of the X. It would be made known that the X1 is like a G4 plus, and the X2 is the 64-bit super chip.</strong><hr></blockquote>
well,sorry - i just wanted to point out that there'll never be a ibm-G3 with just a simd unit added and nothing more. they'll make a generation-switch next year. one of the resulting products will be the mentioned G3successor and the other is the 970. two cpus are ok i think. but the X1, X2 naming convention is a step backwards in my opinion. i like the G5/6 better or even something like G4-X/G5-X64 ? ... i don't know...