okay, did i miss something massively important? last i looked, this thread was a handful of pages. now, unless i am mistaken, it's the longest thread E-V-E-R. (i'd bet it's even longer than the pre-macworld-sanfrancisco thread earlier this year, and that's goin' some...)
Gee I wonder how threads get so long. I am sure it is because there are a lot of people with interesting input about the subject that is stated in the thread name. I am sure it wouldn't include pages of people commenting on such off the subject things like "Oh my god this is a long thread."
You know what my big prediction is, that the mods will wisen up and lock this monster.
Why? Actually, we won't really know until next fall and Moto's releasing it's G4++ in Jan. So actaully, we have almost 11 more months to go with this subject before we get an answer.
So why should it be locked? This thread is about the future of Apple and it's major hardware.
<strong>Just to add a few things the next PIV produced on the 0.09µm process is around 100 million transistors. Is it possible what they will do is start with a single core on a 0.13µm process then switch across?
</strong><hr></blockquote>
could it be possible that the new IBM Desktop Power4 will be produced at 0.09µm ? i remember there was some rumor about IBM to open a new fab where chips shall be produced at 0.09µm....
could it be possible that the new IBM Desktop Power4 will be produced at 0.09µm ? i remember there was some rumor about IBM to open a new fab where chips shall be produced at 0.09µm....</strong><hr></blockquote>
Amorph already reply to this question one page above : speaking of Fishkill the fab will start at 0,13 before migrating to 0,09 some months later.
<strong>Amorph already reply to this question one page above : speaking of Fishkill the fab will start at 0,13 before migrating to 0,09 some months later.</strong><hr></blockquote>
IBM is manufacturing chips using the 130-nanometer process at the plant. The company plans to increase manufacturing quickly and reach full capacity early next year. It also plans a quick move to 90-nanometer production, which will reduce the size of the chips, allowing them to reach higher speeds and increase its manufacturing capacity yet again.
<img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" /> I have general question that pops in my head in regards to 64bit and OS X when I read the various articles about the new IBM PowerPC and its possible use by Apple.
Every article I read make it sound as though there is a ton of work to do in order to bring the Mac OS to full 64 bitness.
However, the G5 from Motorola was always suppose to have a 64bit version and in fact there was even some mention in the past that the G5 would be 64bit(with 32bit compatibility) only.
It's not as though IBM's offering of a 64bit PowerPC part is so outlandish that Apple would have to all of a sudden think about 64bit. The public have been hearing about a 64bit G5 for years.
It would seem to me that Apple would have something well along 64bit native.
This is of course unless the resources for such a thing would be too great to take away from debugging, optimizing and making the 32bit version useable as the mainstream Mac OS release.
<strong>Every article I read make it sound as though there is a ton of work to do in order to bring the Mac OS to full 64 bitness.
</strong><hr></blockquote>
I don't know where you've been reading, but I don't agree with it. MacOSX should port quite nicely to 64-bit, and I expect that they have been experimenting with it for quite some time on POWER4, POWER3 and PPC620 based machines. Possibly even IA-64 since Apple actually said they'd support that back when jumping on the IA-64 bandwagon was fashionable.
Certain APIs, especially in Carbon, might have a more difficult time going 64-bit but they don't have to go full 64-bit on day 1 since the hardware will fully support 32-bit mode. Cocoa is probably 64-bit already. I wouldn't be at all surprised to see a staged roll-out of the 64-bit OS. It might even just start with the 32-bit OS knowing how to put the new chip into 32-bit mode and leave it there. The OS isn't likely to hold up the introduction of 64-bit capable hardware.
<strong>The OS isn't likely to hold up the introduction of 64-bit capable hardware.</strong><hr></blockquote>
That's a good thing... Also from what I remember reading from you and others far more up on that side of things than me most applications will not see any real benefit from being on a 64-bit processor / OS either... Well yes they will have access to GOBS more memory space but it isn't gonna work magic (as I was once led to believe).
<strong>Certain APIs, especially in Carbon, might have a more difficult time going 64-bit but they don't have to go full 64-bit on day 1 since the hardware will fully support 32-bit mode. </strong><hr></blockquote>
I'm not sure most apps will _ever_ go fully 64 bit. It is clearly not an improvement for a large number of things, and there are a lot of things that are simply never going to need that much memory devoted to them. I mean, who hear thinks 'ls' or 'more' or _any_ of the command line tools will be faster in 64-bit mode? Hmm, perhaps some of the networking stuff.
Note I'm not saying nothing needs to address 64 bits of memory, I'm just saying that unlike 8 bit, and 16 bit, there's a LOT that can be done in 32bit apps before forcing everyone up to 64 bit minimums.
In any case, I would think in the end there will be an API 'make a 32bit aware thread' and an API 'make a 64bit aware thread' where the programmer gets to use the right tool for the job with things coexisting relatively peacefully.
It details some interesting info and might provide some hard 'proof' (something I was looking for) that shows another MOT based Apple might show up before a GPUL one steps in to take it's place.
Interesting thing I saw was on the 6th page where it shows the 7451/41 and 7455/45 having/supporting RapidIO. Is that TRUE?!?! I was under the impression that the 7451/41 and 7455/45 were still MaxBus based. If they are RapidIO then why couldn't Apple support full DDR support from the Memory to the CPU?
*or*
Is the riser card RapidIO and Apple had full intentions to move the MB to RapidIO as well but Apple had a problem with that motherboard (Moki said something about that) and had to move to a 'less agressive' motherboard in the 11th hour that did some kinda 'RapidIO to MaxBus translation.
Can you tell I'm talking about stuff that I know very little about?
Anyone wanna take a stab at clearing this stuff up?
Oh also the PDF indicats moving to a .13 process and bumping the onboard L2 cache to 512k as well as a few other things.
Interesting thing I saw was on the 6th page where it shows the 7451/41 and 7455/45 having/supporting RapidIO. Is that TRUE?!?! I was under the impression that the 7451/41 and 7455/45 were still MaxBus based. If they are RapidIO then why couldn't Apple support full DDR support from the Memory to the CPU?</strong><hr></blockquote>
The current Apollo chips that Apple use do not support RapidIO. That doesn't mean Motorola hasn't been planning to add it though.
I don't even know what to say to that...</strong><hr></blockquote>
Oh I agree 110% Matsu! Then again Telomar could be on to something... The date this PDF was made (well as per the PDF Doc Properties) was July 24th of this year.
Oh I agree 110% Matsu! Then again Telomar could be on to something... The date this PDF was made (well as per the PDF Doc Properties) was July 24th of this year.
<strong>Oh I agree 110% Matsu! Then again Telomar could be on to something... The date this PDF was made (well as per the PDF Doc Properties) was July 24th of this year.
Dave</strong><hr></blockquote>
Yes but it was obviously made from a Powerpoint presentation that was created much earlier than that. Did you notice that the "Apollo program" was referred to as a "new technology"? Sorry, but it looks to me like someone updated a presentation from last year's Smart Networks Developer Forum (SNDF) for presentation at this year's Forum. (2002 Recap).
Also, you'll note that the HiPerMOS perfromance roadmap has .13 parts targeted for the end of 2001. Yet here we are at the end of 2002, and I'm not seeing any 0.13 products out from Moto. Unless I'm missing something?
Comments
It has a while to go before it reaches all-time-high status. <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
[EDIT: It was the <a href="http://forums.appleinsider.com/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=1&t=000401"; target="_blank">Apple Hypes MWSF 2002</a> thread, at 1757 posts, started and ended by Fran.]
[ 10-10-2002: Message edited by: Spart ]</p>
<strong>No...Whatcha listening to? thread and some other thread in FH beat it by quite a bit...1400 and 1700ish respectively.
It has a while to go before it reaches all-time-high status. <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
[EDIT: It was the <a href="http://forums.appleinsider.com/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=1&t=000401"; target="_blank">Apple Hypes MWSF 2002</a> thread, at 1757 posts, started and ended by Fran.]
[ 10-10-2002: Message edited by: Spart ]</strong><hr></blockquote>
sigh, how quickly i forgot.... FOURTY-FOUR PAGES?!?! damn, apple had us all by the unmentionables, huh?
damn. how many days until MWSF???
<strong>
sigh, how quickly i forgot.... FOURTY-FOUR PAGES?!?! damn, apple had us all by the unmentionables, huh?
damn. how many days until MWSF???</strong><hr></blockquote>
44 pages and all of it (well 99.99% of it) was posted from Dec 31st and Jan 7th. Talk about a web-board stress test.
Dave
You know what my big prediction is, that the mods will wisen up and lock this monster.
So why should it be locked? This thread is about the future of Apple and it's major hardware.
<strong>Just to add a few things the next PIV produced on the 0.09µm process is around 100 million transistors. Is it possible what they will do is start with a single core on a 0.13µm process then switch across?
</strong><hr></blockquote>
could it be possible that the new IBM Desktop Power4 will be produced at 0.09µm ? i remember there was some rumor about IBM to open a new fab where chips shall be produced at 0.09µm....
<strong>
could it be possible that the new IBM Desktop Power4 will be produced at 0.09µm ? i remember there was some rumor about IBM to open a new fab where chips shall be produced at 0.09µm....</strong><hr></blockquote>
Amorph already reply to this question one page above : speaking of Fishkill the fab will start at 0,13 before migrating to 0,09 some months later.
If it starts at 2 gig on .13.
...then .9 GPUL is going to be frightening.
It seems that this processor will have much more headroom mhz and performance wise than the shorter pipelined chips of recent PPC years.
Lemon Bon Bon
<strong>Amorph already reply to this question one page above : speaking of Fishkill the fab will start at 0,13 before migrating to 0,09 some months later.</strong><hr></blockquote>
The URL is <a href="http://news.com.com/2102-1001-947358.html"; target="_blank">here</a> in case anyone hasn't read it. I commented several lines in another thread on AI <a href="http://forums.appleinsider.com/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=1&t=002537&p=3"; target="_blank">here</a> but for completeness:
IBM is manufacturing chips using the 130-nanometer process at the plant. The company plans to increase manufacturing quickly and reach full capacity early next year. It also plans a quick move to 90-nanometer production, which will reduce the size of the chips, allowing them to reach higher speeds and increase its manufacturing capacity yet again.
Early next year!
Every article I read make it sound as though there is a ton of work to do in order to bring the Mac OS to full 64 bitness.
However, the G5 from Motorola was always suppose to have a 64bit version and in fact there was even some mention in the past that the G5 would be 64bit(with 32bit compatibility) only.
It's not as though IBM's offering of a 64bit PowerPC part is so outlandish that Apple would have to all of a sudden think about 64bit. The public have been hearing about a 64bit G5 for years.
It would seem to me that Apple would have something well along 64bit native.
This is of course unless the resources for such a thing would be too great to take away from debugging, optimizing and making the 32bit version useable as the mainstream Mac OS release.
Next, Inc. probably had a 64bit, no?
Any comments?
-tink
<strong>Every article I read make it sound as though there is a ton of work to do in order to bring the Mac OS to full 64 bitness.
</strong><hr></blockquote>
I don't know where you've been reading, but I don't agree with it. MacOSX should port quite nicely to 64-bit, and I expect that they have been experimenting with it for quite some time on POWER4, POWER3 and PPC620 based machines. Possibly even IA-64 since Apple actually said they'd support that back when jumping on the IA-64 bandwagon was fashionable.
Certain APIs, especially in Carbon, might have a more difficult time going 64-bit but they don't have to go full 64-bit on day 1 since the hardware will fully support 32-bit mode. Cocoa is probably 64-bit already. I wouldn't be at all surprised to see a staged roll-out of the 64-bit OS. It might even just start with the 32-bit OS knowing how to put the new chip into 32-bit mode and leave it there. The OS isn't likely to hold up the introduction of 64-bit capable hardware.
<strong>The OS isn't likely to hold up the introduction of 64-bit capable hardware.</strong><hr></blockquote>
That's a good thing... Also from what I remember reading from you and others far more up on that side of things than me most applications will not see any real benefit from being on a 64-bit processor / OS either... Well yes they will have access to GOBS more memory space but it isn't gonna work magic (as I was once led to believe).
Dave
<strong>Certain APIs, especially in Carbon, might have a more difficult time going 64-bit but they don't have to go full 64-bit on day 1 since the hardware will fully support 32-bit mode. </strong><hr></blockquote>
I'm not sure most apps will _ever_ go fully 64 bit. It is clearly not an improvement for a large number of things, and there are a lot of things that are simply never going to need that much memory devoted to them. I mean, who hear thinks 'ls' or 'more' or _any_ of the command line tools will be faster in 64-bit mode? Hmm, perhaps some of the networking stuff.
Note I'm not saying nothing needs to address 64 bits of memory, I'm just saying that unlike 8 bit, and 16 bit, there's a LOT that can be done in 32bit apps before forcing everyone up to 64 bit minimums.
In any case, I would think in the end there will be an API 'make a 32bit aware thread' and an API 'make a 64bit aware thread' where the programmer gets to use the right tool for the job with things coexisting relatively peacefully.
[ 10-13-2002: Message edited by: Nevyn ]</p>
<a href="http://e-www.motorola.com/collateral/SNDF2002RECAP_H1101.pdf"; target="_blank">http://e-www.motorola.com/collateral/SNDF2002RECAP_H1101.pdf</a>;
It details some interesting info and might provide some hard 'proof' (something I was looking for) that shows another MOT based Apple might show up before a GPUL one steps in to take it's place.
Interesting thing I saw was on the 6th page where it shows the 7451/41 and 7455/45 having/supporting RapidIO. Is that TRUE?!?! I was under the impression that the 7451/41 and 7455/45 were still MaxBus based. If they are RapidIO then why couldn't Apple support full DDR support from the Memory to the CPU?
*or*
Is the riser card RapidIO and Apple had full intentions to move the MB to RapidIO as well but Apple had a problem with that motherboard (Moki said something about that) and had to move to a 'less agressive' motherboard in the 11th hour that did some kinda 'RapidIO to MaxBus translation.
Can you tell I'm talking about stuff that I know very little about?
Anyone wanna take a stab at clearing this stuff up?
Oh also the PDF indicats moving to a .13 process and bumping the onboard L2 cache to 512k as well as a few other things.
Dave
[ 10-13-2002: Message edited by: DaveGee ]</p>
<strong>
Interesting thing I saw was on the 6th page where it shows the 7451/41 and 7455/45 having/supporting RapidIO. Is that TRUE?!?! I was under the impression that the 7451/41 and 7455/45 were still MaxBus based. If they are RapidIO then why couldn't Apple support full DDR support from the Memory to the CPU?</strong><hr></blockquote>
The current Apollo chips that Apple use do not support RapidIO. That doesn't mean Motorola hasn't been planning to add it though.
If Apple had a faster bus available to them and has been dogging it for the last 12 months...
<img src="graemlins/oyvey.gif" border="0" alt="[No]" /> <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> <img src="graemlins/oyvey.gif" border="0" alt="[No]" />
I don't even know what to say to that...
<strong>OH boy!
If Apple had a faster bus available to them and has been dogging it for the last 12 months...
<img src="graemlins/oyvey.gif" border="0" alt="[No]" /> <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> <img src="graemlins/oyvey.gif" border="0" alt="[No]" />
I don't even know what to say to that...</strong><hr></blockquote>
Oh I agree 110% Matsu! Then again Telomar could be on to something... The date this PDF was made (well as per the PDF Doc Properties) was July 24th of this year.
Dave
[ 10-13-2002: Message edited by: DaveGee ]</p>
<strong>
Oh I agree 110% Matsu! Then again Telomar could be on to something... The date this PDF was made (well as per the PDF Doc Properties) was July 24th of this year.
Dave
[ 10-13-2002: Message edited by: DaveGee ]</strong><hr></blockquote>
Dave, now apply some of the info from the "GPUL on Oct 15th" thread. Borrowing from your post there (with assistance from Barto and Outsider):
[quote]<strong>
MPF 1998 = MOT = 7400. = 1 Year 0 Months
MPF 1999 = MOT = 7450. = 1 Year 6 Months
MPF 2000 = MOT = 7455. = 1 Year 6 Months
MPF 2001 = IBM = 750fx = 0 Year 6 Months
MPF 2002 = IBM = GPUL. = ? Year ? Months
</strong><hr></blockquote>
And, Moto hasn't made an official public announcement (aside from the roadmap "briefing chart"). So following Moto's past performance ...
I still don't have any faith in Moto.
<strong>Oh I agree 110% Matsu! Then again Telomar could be on to something... The date this PDF was made (well as per the PDF Doc Properties) was July 24th of this year.
Dave</strong><hr></blockquote>
Yes but it was obviously made from a Powerpoint presentation that was created much earlier than that. Did you notice that the "Apollo program" was referred to as a "new technology"? Sorry, but it looks to me like someone updated a presentation from last year's Smart Networks Developer Forum (SNDF) for presentation at this year's Forum. (2002 Recap).
Also, you'll note that the HiPerMOS perfromance roadmap has .13 parts targeted for the end of 2001. Yet here we are at the end of 2002, and I'm not seeing any 0.13 products out from Moto. Unless I'm missing something?