or Connect
AppleInsider › Forums › Mac Hardware › Future Apple Hardware › IBM unveils dual-core PowerPC chips up to 2.5GHz
New Posts  All Forums:Forum Nav:

IBM unveils dual-core PowerPC chips up to 2.5GHz - Page 2

post #41 of 280
Quote:
Originally posted by BRussell
How efficient would a dual dual-core be? I'm guessing that most users get little benefit right now from duals, let alone quads. I'm sure there are some who use apps that take advantage of them, or use multiple apps simultaneously. But quads?

If the program is able to use two cpu's it should work with four cores as well. Apple has worked on that in 10.4. Previous to that it could only use two.

Some programs won't get any benefit from four. Mostly games only use one. There's a great deal of talk about this one in PS3 and 360 threads, as well as the dual threads on Ars etc.
post #42 of 280
Quote:
Originally posted by melgross
If the program is able to use two cpu's it should work with four cores as well. Apple has worked on that in 10.4. Previous to that it could only use two.

Remember as well, the CPU isn't the only part of the equation, we have to see the motherboard Apple puts these new chips on.
post #43 of 280
Quote:
Originally posted by TofuTodd
IBM is a specialty chip manufactuer. They wouldn't spend millions on development of a mobile G5 unless there were guranteed customers. Apple is the only consumer of G5 chips...

uhh yea Apple and that other small company-- what was there name... micro something... errr.. squaresoft? oh i forgot, they make some OS or something and some game console, ive heard its popular in sweden, they are investing heavily in the PPC.
post #44 of 280
Quote:
Originally posted by mike12309
uhh yea Apple and that other small company-- what was there name... micro something... errr.. squaresoft? oh i forgot, they make some OS or something and some game console, ive heard its popular in sweden, they are investing heavily in the PPC.

That's very funny, but those aren't the same chips.
post #45 of 280
Quote:
Originally posted by mike12309
they are investing heavily in the PPC.

Investing heavily? Hows that? They went with IBM because they were the cheapest.
post #46 of 280
Quote:
Originally posted by TofuTodd
IBM is a specialty chip manufactuer. They wouldn't spend millions on development of a mobile G5 unless there were guranteed customers. Apple is the only consumer of G5 chips...

Oh really:

What about Momentum?
and
Terrasoft
and
Continuous Computing ???

Christ, those were on the first result page of a Google search for '970fx manufacturers'.
...we have assumed control
Reply
...we have assumed control
Reply
post #47 of 280
Anyone knows if those chips have any form of hyper/multithreading in them?
post #48 of 280
Quote:
Originally posted by Rhumgod
Oh really:

What about Momentum?
and
Terrasoft
and
Continuous Computing ???

Christ, those were on the first result page of a Google search for '970fx manufacturers'.

That's true, but all of those manufacturers together won't come close to equaling Apple's usage.
post #49 of 280
Quote:
Originally posted by melgross
Yeah, I saw that. I'm wondering who will want a dual 1.4GHz chip.

The enhancements are intriguing though. They don't tell us much about it.

We can SPECULATE

Well one thought comes to mind -- a little company named N-something with a product coming next year named R-something. We know it has to be compatible with an existing G3-based product.





Don't forget that both of these chips have internal improvements, including 1 MB of cache (per core). This should result in better performance at the same clock rate, compared to the 970FX. I expect Apple to use the 970MP in the towers until they switch to Intel sometime around 2007.

More and more software is going to be multi-threaded going forward, especially if quad-core machines are available. People that don't need that much horsepower will still benefit from a single 970MP due to the tighter link between cores and the increased cache sizes. Tiger's internal improvements (of which we can expect more in the future) also mean that we'll see better core utilization.

And remember that Apple's decision to go to Intel was made with full knowledge of these processors, and what IBM is going next. Intel's future roadmap is what has lured Apple from PPC to x86, not these chips or Intel's current chips. Intel is probably also talking a better story about their 65 and 40 nm migration math.
Providing grist for the rumour mill since 2001.
Reply
Providing grist for the rumour mill since 2001.
Reply
post #50 of 280
Quote:
Originally posted by Programmer
Well one thought comes to mind -- a little company named N-something with a product coming next year named R-something. We know it has to be compatible with an existing G3-based product.





Don't forget that both of these chips have internal improvements, including 1 MB of cache (per core). This should result in better performance at the same clock rate, compared to the 970FX. I expect Apple to use the 970MP in the towers until they switch to Intel sometime around 2007.

More and more software is going to be multi-threaded going forward, especially if quad-core machines are available. People that don't need that much horsepower will still benefit from a single 970MP due to the tighter link between cores and the increased cache sizes. Tiger's internal improvements (of which we can expect more in the future) also mean that we'll see better core utilization.

And remember that Apple's decision to go to Intel was made with full knowledge of these processors, and what IBM is going next. Intel's future roadmap is what has lured Apple from PPC to x86, not these chips or Intel's current chips. Intel is probably also talking a better story about their 65 and 40 nm migration math.

I suppose my daughter will want one of those NR's, as well as an MX and an SP

I did mention the other matters in an above post. The dual 1.4 does seem to have a limited use though.
post #51 of 280
As part of the question of what programs can utilize 4, or more, processors, here's one answer:

http://www.architosh.com/news/2005-0..._maxwell1.html
post #52 of 280
Being that a PowerMac with Dual - Dual core processors could arrive faster than a Dual - Dual core Intel PowerMac (probably by at least a year.) I would be buying one. As long at it pretty much smokes the current Daul 2.7GHz PowerMac.
onlooker
Registered User

Join Date: Dec 2001
Location: parts unknown




http://www.apple.com/feedback/macpro.html
Reply
onlooker
Registered User

Join Date: Dec 2001
Location: parts unknown




http://www.apple.com/feedback/macpro.html
Reply
post #53 of 280
In case some might be interested in a broken English translation of IBM's Japanese press release, here it is (I hope). If it doesn't work, I'll copy it to the post.

http://translate.google.com/translat...050707003.html
post #54 of 280
I want to see...

DUAL DUAL PPC PowerMac
DUAL PPC Powerbook
DUAL PPC iMac

I like the idea of DUAL everything! G5 Squared (dual dual)!

post #55 of 280
What if the 1.4Ghz range of dual core G5s are low power? Say 30W dual G5 1.4, maybe 35W dual G5 1.6. That'd be worthwhile for a powerbook certainly. Also, at 16W @ 1.6Ghz the chip is extremely cool; Apple had a tradition of getting Motorola to certify chips at higher clocks for Apple when they felt they could get awat with generating more heat.
post #56 of 280
Quote:
Originally posted by ChevalierMalFet
What if the 1.4Ghz range of dual core G5s are low power? Say 30W dual G5 1.4, maybe 35W dual G5 1.6. That'd be worthwhile for a powerbook certainly. Also, at 16W @ 1.6Ghz the chip is extremely cool; Apple had a tradition of getting Motorola to certify chips at higher clocks for Apple when they felt they could get awat with generating more heat.

Certainly. Freescale still say the 7447A only goes up to 1.4GHz officially, yet Apple are shipping laptops with them running at 1.67GHz. I wouldn't start thinking that Freescale can't get 2GHz 7448s to Apple, and likewise that IBM can't get 1.8 or even 2GHz G5s at under 20W.
post #57 of 280
Quote:
Originally posted by kenaustus
Looks like IBM is trying to show that Apple was wrong. Unfortunately the only major customer they will have for it is Apple, and only for a year or so. Just shows how important it is to be proactive instead of reactive.

No, CPU development doesn't work like that. If IBM developed a new CPU in reaction to Jobs' switch to Intel announcement, it would take one to two years to get even close to mass production. If anything, the CPUs announced today were the very reason for Apple's switch.

Consider that Apple has know of these two chips, the low-power 970fx and the dual core 970mp, since their early development stages. IBM most likely gave Apple a roadmap with these CPUs and predicted availability - then IBM flubbed everything by missing the timeline projections, and kept moving the expected release date farther and farther into the future. Meanwhile Apple's tower sales are tanking, and they're still cramming Motorola's 20th century turds into low end Macs and high end laptops. IBM repeatedly failed to meet Apple's needs and so Jobs decided to switch to Intel.

This scenario is pure speculation, but we do know that Apple has had CHUD tools for some time that were written for systems with 4 cores. Furthermore, this low power 970fx would have been killer technology 2 years ago. The 970mp keeps Apple in the game if 9they can ship 970mp-based Powermacs this year, but if IBM had introduced the dual core 970 chip a year ago, Apple would have dominated in desktop PC performance.

There are a host of missed opportunities by IBM and Apple, and these are only the ones we know about. My guess is that Jobs grew weary of IBM's glacial development of the PPC 970, and so pulled the trigger on Intel-based Macs,
post #58 of 280
Quote:
Originally posted by kim kap sol
The dual-core announcement is interesting, until you realize that it's not an improvement over the current lineup unless Apple decides to sell dual-dual-cores computers.

Agreed. I'm at a loss for predicting Apple's choice on this one. They can pad their profit margins by offering only single CPU Powermacs, and make up for the lame performance with RDF marketing. Alternatively, Apple can do the right thing; evidence in the form of CHUD tools configured for >2 cores suggests Apple is considering the right thing, and may follow through. Unfortunately, it seems like Apple has been very miserly over motherboard specs and other details of Powermac performance that could easily boost the performance of stale CPUs.

Quote:
This will not help gaming or any app that isn't multithreaded one bit. What it will let you do is run single-thread apps without a hiccup. Not bad but not ground-breaking.

Although the few well threaded apps will be monsters on those computers.

Dude, not true! Since when is a quad-core Powermac not "ground-breaking"? Tiger is ready for quad cores, a few apps are, like you said, and once quad-core Powermacs are available, many developers will exploit the new technology. And since Intel and AMD are moving to multi-core CPUs, games will follow before long. A twin dual-core Powermac G5 would be a fucking earthquake that hits the entire Mac computing world - Apple may not be the first to the multi-core party, but they have the potential to implement dual core technology better by leveraging OS X, and by putting such instruments in the hands of average Mac users.
post #59 of 280
i think we are all arguing about the same thing. anyway here's my view:

1. IBM has some promising stuff but it is a bit late to the game, nonetheless, at least there is some indication about the g5 in 2006. remember that 2006 will be the "Year of Transition" to make the Mac platform more dynamic, more scalable, albeit with the split-personality challenges of intel/powerpc universal binaries. for the most part, Intel is very likely to deliver, if they cant, then the whole CPU industry is "teh doomed" which in any case AMD would love to have apple come knocking on their door. whatever IBM's deal is, let them be on their way, let's get our bloody g5 chips from them in 2006 and at the end of 2006 put an end to this nonsense of IBM and Freescale holding up Apple in its time of greatest popularity and brand value

2. however, people are skeptical about IBM delivering these g5s in 2006.

3. people that HAVE USED A G5 PERSONALLY will know that applications designed for g5 / PowerPC/ Altivec/ iApps with 1gb of RAM, it kicks ass.

4. however, value for money, the thrill of latest technology, i can tell you, AMD single and dual cores (4000-4800 series) give consumers a feeling of value for money that the powermac g5 does NOT

5. we can argue about GHZ 'till we're blue in the face but i can tell you a joe average looking at a iMac g5 1.8ghz will wonder, how come Dell is selling a 2++ghz for cheaper and it has more of that "RAM" thing. even if the new 1.6ghz low power g5 970fx runs cooler, better, and makes for a sexier iMac g5, steve and apple marketing will just be like, well, that's all bloody well and good but how do we MARKET this thing?

apologies if the CAPS got a bit annoying in this post.
........
post #60 of 280
Quote:
Originally posted by Junkyard Dawg
........ A twin dual-core Powermac G5 would be a fucking earthquake that hits the entire Mac computing world - Apple may not be the first to the multi-core party, but they have the potential to implement dual core technology better by leveraging OS X, and by putting such instruments in the hands of average Mac users.

yes. but for the 'average Mac user', such a earth-shattering quad-powermac g5 would have to deliver on the following points to deliver sales, revenues and profits:

A. value for money
B. 2-3ghz clock speed
C. high definition 720p, 1080p
D. smooth integration into the home theatre thing
E. easily obtainable 720p and 1080p video content (see C and D)
F. integration with legacy windows xp/2000 applications
(yes i am talking virtual PC, virtual PC on a quad-powermac g5 with 2gb ram
put an end to the "oh but my business needs this-and-that bullshit
legacy windoze application thing so i can't even consider macs")
G. Games
H. value for money
I. a certain level of upgradeability to convince customers that they are not
locked in to a certain proprietary system, which is one major turn-off
to switching

and last but not least,

J. Steve back in blue jeans and RDF marketing set to '11'
post #61 of 280
Quote:
Originally posted by sunilraman
yes. but for the 'average Mac user', such a earth-shattering quad-powermac g5 would have to deliver on the following points to deliver sales, revenues and profits:

A. value for money
B. 2-3ghz clock speed
C. high definition 720p, 1080p
D. smooth integration into the home theatre thing
E. easily obtainable 720p and 1080p video content (see C and D)
F. integration with legacy windows xp/2000 applications
(yes i am talking virtual PC, virtual PC on a quad-powermac g5 with 2gb ram
put an end to the "oh but my business needs this-and-that bullshit
legacy windoze application thing so i can't even consider macs")
G. Games
H. value for money
I. a certain level of upgradeability to convince customers that they are not
locked in to a certain proprietary system, which is one major turn-off
to switching

and last but not least,

J. Steve back in blue jeans and RDF marketing set to '11'



G. Games
Yes, Apple seriously needs to work on this (opengl? more support for developrs? convince more developrs?). Luckily we have 2 of the best game makers happy to let mac ports exist- ID Software & Blizzard. But even then, the games run poor, and are far and few between.

Things are getting worse not better. And game engines are about to hit the 'next generation'- a sad tale when a sh*t hot powermac can only just run doom3.

The days of buying a nice well behaved mac solely to run quark and illustrator are over. Everything about macs kick ass, except the state of gaming.
post #62 of 280
If gaming is a top priority, you shouldn't buy a computer, not a mac, not a PC. Consoles are cheaper, better supported, and provide a better experience. There are some games that are a bit better on the computer, but they're incredibly monotonous cookie-cutter FPS games -- once you've played one, you've played them all.
IBL!
Reply
IBL!
Reply
post #63 of 280
Quote:
Originally posted by sunilraman
yes. but for the 'average Mac user', such a earth-shattering quad-powermac g5 would have to deliver on the following points to deliver sales, revenues and profits:

A. value for money
B. 2-3ghz clock speed
C. high definition 720p, 1080p
D. smooth integration into the home theatre thing
E. easily obtainable 720p and 1080p video content (see C and D)
F. integration with legacy windows xp/2000 applications
(yes i am talking virtual PC, virtual PC on a quad-powermac g5 with 2gb ram
put an end to the "oh but my business needs this-and-that bullshit
legacy windoze application thing so i can't even consider macs")
G. Games
H. value for money
I. a certain level of upgradeability to convince customers that they are not
locked in to a certain proprietary system, which is one major turn-off
to switching

and last but not least,

J. Steve back in blue jeans and RDF marketing set to '11'

For dog's sake,

What you're describing is a games console, not a PowerMac.

The 'average Mac user' doesn't buy a PowerMac for running games and watching video on their plasma screen. Heck, the average computer user full stop doesn't either. I was so glad when I got rid of my dual processor PC (Abit BP6, dual celeron) from my front room for an iMac.

If Apple release a Quad PowerMac with these chips in then it's for the top end creative professionals using FinalCut and stuff like that. In that world - time = money. Even if they cost $4000+, it's value-for-money there.

An iMac is more than enough for the 'average Mac user' which is why they sell hand over fist more of them than PowerMacs. I use one for web work and photoshop - it's more than enough. Even the much maligned FX5200 in mine is more than adequate for desktop use. Games - no, but then I don't play games.

However, if someone wants to drop me off a Quad PowerMac and 30 inch screen to play Enigmo on, I won't say no. ;-)
post #64 of 280
It is one thing to manufacture it is another thing to sell your product. Do you really see a huge percentage of the consumer market here.

I mentioned the consumer market because there isn't a big enough market outside that space to sustain the manufacture and development of a high performance processor. Maybe IBM / mentioned vendors could make headway into the commercial market with Linux, but Intel is intrenched there. Actually it is the same deal in consumer space where Intel compatibility is held as an important check off option.

Thanks
dave


Quote:
Originally posted by Rhumgod
Oh really:

What about Momentum?
and
Terrasoft
and
Continuous Computing ???

Christ, those were on the first result page of a Google search for '970fx manufacturers'.
post #65 of 280
Quote:
Originally posted by Hattig
Wow, such a lot of negativity here.


Well deserved negativity.
Quote:

The low power G5s will be cherry picked dies, like AMD cherry picks Athlon 64 dies that run cool and brands them as Turion (35W and 25W max). Intel's max is 27W. I don't know what the max is of this new processor, but if 13W is typical then I imagine that it'll be around 26W or so.

I don't believe this is the case at all, but haven't gotten official confirmation. I beleive the chips are built on a new low power process. In any event they don't run fast enough for anybody to get excited about. While it might be a bit early to say so it does look like Freescales chips would out perform these in many applications.
Quote:

So IBM announces 1.2, 1.4 and 1.6? So what! Previously Apple has had faster machines out than IBM's product line would suggest. Maybe IBM are making 1.8GHz 16W G5s for Apple?

Maybe - maybe not, we can only make our opinions based on reality of the release statments. Wishful thinking will not produce gold.
[quote]

The real issue is the power management features on the latest processors. 13W AVERAGE would be rather poor.
[\\quote]
While it is difficult to compare power numbers (typical and max and everything in between) 13 watts might not be that bad if the processors had arrived at a reasonable performnace zone. On the surface it doesn't look like the chips out perform Freescales old stuff not to mention what Freescale is now delivering or will deliver soon. It is not the 13 watts that are the problem it is the performance that comes from that 13 watts that is the problem.
Quote:

The dual-core looks quite interesting however.

The dual core is interesting no doubt. The problem is that it suffers from the same problems as the other chips which is the lack of real performance in single thread applications. The larger cache and other improvements might make up for some of the clock looses but we have yet to see how well that will play out.

Dave
post #66 of 280
Quote:
Originally posted by mike12309
i personally would have prefered these annoucements a day after steve jobs intel lovin speech :-) Sorry, steve is wrong this time

A reasonable examination of the release information would not allow you to make that statement!!! Especially in th ecase of the low power FX variants, which won't compete with Freescales hardware and certainly won't compete with current Intel hardware. That doesn't even take into account new intel hardware that is comeing very soon.

I hate to say it but Steve did the right thing and this announcement pretty much verifies his speach.

Dave
post #67 of 280
Quote:
Originally posted by Altivec_2.0
I don't know about anyone else, but I would buy a Dual 2.5Ghz Dual core PowerMac before I would ever consider buying one from intel.

man, i hate the idea apples switching to intel could have been at least AMD....sad

They will use AMD. Not right away though.

Remember, Apple is/was used to controlling almost everything, hard and soft. It's only been in the last decade that they stopped using proprietary interfaces (NuBus, ADB, etc.) When they dropped SCSI in favor of ATA, that was a clear signal. Then we got USB (an Intel technology!!), and more and more "standardization." The whole ADC thing was a weird backstep (even if it is a cool technology that I like), but one that was doomed.

I've said this before, but I may as well say it again.

Apple announced "Intel-only" and "OS X only on Macs" because it is being prudent. They want to be able to make sure that OS X is ultra stable in x86. My guess is that during the transition, (roughly March 06 - July 07), they will begin "certifying" various hardware with a "Works with Macs" logo program (or even "Designed for Macs" !). Mostly these will be various PCIe cards. And their drivers will be available from Apple (apple.com and SWU) or the manufacturer. Within 12 months of the switch, Apple will begin offering AMD chips as well. The Mac will be indistinguishable from an Intel box, except that the Apple's will be much more elegant. Then the magic moment will come when the "OS X only on Macs" will go away. The new switcher box will go from being a Mac mini ($499) to being a box of OS X ($129) installed on a Dell.

What about the need for hardware dollars? Well I for one don't see Apple's revenue falling. Perhaps lots of people will buy Dells or DIY boxen and a copy of OSX. Certainly geeks like us will. But the average users like my parents and parents-in-law don't like doing anything weird to their computers. They'll buy Macs with OSX installed. But even if you do see a hardware catastrophe for Apple in terms of CPUs sales on the horizon, Apple will not die. Why? Digital hub, digital spokes, digital lifestyle products of course. The iPod is showing Apple the way. Slightly modified minis (with 5.1 output and HDMI) attached to a 30-inch display or an HDTV, iPods, Airport Express, iSights, (and a remote for it all), iTunes, etc. All linked with iMacs and *Books all over the house. (Stream TV to any one of them.) PowerMacs will be the choice for businesses and artists (graphics, movies, music, etc.) wanting Macs.

When OS X's marketshare begins to grow to 5-10% (which it will), MS will start showing its teeth. Luckily, MS Office is no longer the "killer app" it was in the 90s. Apple is probably just weeks away from being able to annouce a full-blown office suite that would meet the needs of 90% of businesses out there, and would be much more elegant, simple, (and therefore productive) than MS Office. (Not that it will announce one anytime soon. But iWork will get a spreadsheet in its next iteration ('06), and by '07 or '08 it will be a pretty nice piece of software. And by that time, it's only a step away from buffing iWork, Mail, Address Book, iCal, iChat into "Pro" versions, integrating them with Filemaker, and shrinkwrapping the whole thing. Charge a fiver for an unlimited liscence and voila.)

Anyway, the point is this. The Apple we know and love will continue to evolve, the marketplace will change, and blah blah blah. By '08, you'll have your AMD in your Mac.
post #68 of 280
Quote:
Originally posted by DCQ
blah blah blah

Sounds like we need to revive the Kormac threads...

;^p
Late 2009 Unibody MacBook (modified)
2.26GHz Core 2 Duo CPU/8GB RAM/60GB SSD/500GB HDD
SuperDrive delete
Reply
Late 2009 Unibody MacBook (modified)
2.26GHz Core 2 Duo CPU/8GB RAM/60GB SSD/500GB HDD
SuperDrive delete
Reply
post #69 of 280
Quote:
Originally posted by DCQ
They will use AMD. Not right away though.

Remember, Apple is/was used to controlling almost everything, hard and soft. It's only been in the last decade that they stopped using proprietary interfaces (NuBus, ADB, etc.) When they dropped SCSI in favor of ATA, that was a clear signal. Then we got USB (an Intel technology!!), and more and more "standardization." The whole ADC thing was a weird backstep (even if it is a cool technology that I like), but one that was doomed.

I've said this before, but I may as well say it again.

Apple announced "Intel-only" and "OS X only on Macs" because it is being prudent. They want to be able to make sure that OS X is ultra stable in x86. My guess is that during the transition, (roughly March 06 - July 07), they will begin "certifying" various hardware with a "Works with Macs" logo program (or even "Designed for Macs" !). Mostly these will be various PCIe cards. And their drivers will be available from Apple (apple.com and SWU) or the manufacturer. Within 12 months of the switch, Apple will begin offering AMD chips as well. The Mac will be indistinguishable from an Intel box, except that the Apple's will be much more elegant. Then the magic moment will come when the "OS X only on Macs" will go away. The new switcher box will go from being a Mac mini ($499) to being a box of OS X ($129) installed on a Dell.

What about the need for hardware dollars? Well I for one don't see Apple's revenue falling. Perhaps lots of people will buy Dells or DIY boxen and a copy of OSX. Certainly geeks like us will. But the average users like my parents and parents-in-law don't like doing anything weird to their computers. They'll buy Macs with OSX installed. But even if you do see a hardware catastrophe for Apple in terms of CPUs sales on the horizon, Apple will not die. Why? Digital hub, digital spokes, digital lifestyle products of course. The iPod is showing Apple the way. Slightly modified minis (with 5.1 output and HDMI) attached to a 30-inch display or an HDTV, iPods, Airport Express, iSights, (and a remote for it all), iTunes, etc. All linked with iMacs and *Books all over the house. (Stream TV to any one of them.) PowerMacs will be the choice for businesses and artists (graphics, movies, music, etc.) wanting Macs.

When OS X's marketshare begins to grow to 5-10% (which it will), MS will start showing its teeth. Luckily, MS Office is no longer the "killer app" it was in the 90s. Apple is probably just weeks away from being able to annouce a full-blown office suite that would meet the needs of 90% of businesses out there, and would be much more elegant, simple, (and therefore productive) than MS Office. (Not that it will announce one anytime soon. But iWork will get a spreadsheet in its next iteration ('06), and by '07 or '08 it will be a pretty nice piece of software. And by that time, it's only a step away from buffing iWork, Mail, Address Book, iCal, iChat into "Pro" versions, integrating them with Filemaker, and shrinkwrapping the whole thing. Charge a fiver for an unlimited liscence and voila.)

Anyway, the point is this. The Apple we know and love will continue to evolve, the marketplace will change, and blah blah blah. By '08, you'll have your AMD in your Mac.

This is all very optimistic, but you're getting ahead of yourself.

No one is going to manufacture Macs if they don't have control of what they make. As was shown with the first round of clones, it can get out of Apple's control .

Apple won't allow clones unless thet have enough business from other product lines to keep their computer business down to perhaps 25% of the total. If that occurs they could think of licensing. The business lost in hardware could then be made up by OS sales and royalty payments. But only if this would serve to increase the share of the market for the OS . This also failed to occur before.

There is no guarantee that this could work. I doubt very much that we would see anything occuring in 2006, or even until after the transition is complete at the end of 2007.
post #70 of 280
Quote:
Originally posted by melgross
If the program is able to use two cpu's it should work with four cores as well. Apple has worked on that in 10.4. Previous to that it could only use two.

Some programs won't get any benefit from four. Mostly games only use one. There's a great deal of talk about this one in PS3 and 360 threads, as well as the dual threads on Ars etc.

Actually, it all depends on how many threads any particular program decides to use. If they're written such as to say "Max threads = Num Cores", then they're OK. If they just max out at two threads (or just never use more than two threads), then the extra cores won't be used anyway. Except for tasks you want to run in the background that aren't affected by a user's use of a program, programs shouldn't be using many threads in their software, esp. more than the number of processors on the computer, as it'd actually take longer to process (due to task switching) then it is to run them serially.

This is why I always thought it was funny how everyone waited desperately for a pre-emptively multitasking OS, thinking you could do many things at once and save all that time, only to find out everything seemed slower. Gee, what a shock.
post #71 of 280
Perceptions:
The fact that these chips were announced prior to an Apple product announcement is telling. IBM has developed a business model like one of those paint your own clay pots and IBM will fire it in our fab (Power Everywhere). IBM makes Power4's and Power5's for internal consumption. If you want to put together a chip in our foundry IBM will work with your team and piece together a processor from parts. The first successful implementation of that strategy was the 970 and resulted from the following partnership:
http://www.apple.com/pr/library/2003/jun/23joint.html

Other examples of this strategy are XBox and PS3.

Speculation:
Based on the partnership Apple had employees working with IBM developing the next 970 chips, portable and dual core. The team is having less success than Steve would like and he has finally thrown in the towel. Steve decided it is time to get a little less hardware integrated and leave processor (and motherboard?) development to others.

These chips being on the market is a combination of a technical failure and a marketing failure. The GHz myth is real from a marketing standpoint. It is a hard sell to replace the existing G4's with 'improved' G5's with same GHz and so much doubt about the real advantage going from 32 to 64 bit.

On the dual core desktops, these will be easier to sell as many have stated here. I would also like a dual dual computer. However from a marketing standpoint it will still be a challenge. I have never seen Apple take a high end machine, in this case the Dual 2.7 and make it obsolete in one generation. This means that a dual dual line-up will probably be up-market (read expensive).

As many also realize the multi-processor capabilities of the dual dual makes it more suitable for XServe and specialized markets. It isn't really a home machine except for bragging rights. It has a limited market. (The Mac Mini and notebooks really are the bulk of the market.)

If the Dual Dual 970's do get released I will guess that they will have a long shelf life similar to the 9600.

How about an FX or a Dual in the Mac Mini? How does that fit in the line-up? Can Apple put a G5 in the Mac Mini?
post #72 of 280
Quote:
Originally posted by Louzer
Actually, it all depends on how many threads any particular program decides to use. If they're written such as to say "Max threads = Num Cores", then they're OK. If they just max out at two threads (or just never use more than two threads), then the extra cores won't be used anyway. Except for tasks you want to run in the background that aren't affected by a user's use of a program, programs shouldn't be using many threads in their software, esp. more than the number of processors on the computer, as it'd actually take longer to process (due to task switching) then it is to run them serially.

This is why I always thought it was funny how everyone waited desperately for a pre-emptively multitasking OS, thinking you could do many things at once and save all that time, only to find out everything seemed slower. Gee, what a shock.

it just takes time, for a single processor pre-emptively multitasking, from a consumer point of view, any Tiger 10.4.1 on a 1ghz or greater g4 with 512mb or more is pretty much the current pinnacle of single processor pre-emptive multitasking IMHO
particularly if you talk about iApps, iLife, iWork, Adobe/Macromedia, yes, even Microsoft Office 2004 for Mac, 11.1.1 (latest build is FAST)

apple's software geniuses will figure out the optimising on multiple cores/ cpus, from iApps, iLife, iWork, leading the way for Adobe/Macromedia, Microsoft Mac Unit and inspiring all those talented open-source kids out there...

2005: year of HD, year of Intel bombshell
2006: year of transition
2007: starting to reap great benefits of transition
2008: ready to hit 10% market share, entry into business market

.........
.........
post #73 of 280
Quote:
Originally posted by Louzer
Actually, it all depends on how many threads any particular program decides to use. If they're written such as to say "Max threads = Num Cores", then they're OK. If they just max out at two threads (or just never use more than two threads), then the extra cores won't be used anyway. Except for tasks you want to run in the background that aren't affected by a user's use of a program, programs shouldn't be using many threads in their software, esp. more than the number of processors on the computer, as it'd actually take longer to process (due to task switching) then it is to run them serially.

This is why I always thought it was funny how everyone waited desperately for a pre-emptively multitasking OS, thinking you could do many things at once and save all that time, only to find out everything seemed slower. Gee, what a shock.

Nonsense.

Sure, there's an overhead in managing multiple threads but it's not as big as switching tasks. A well written piece of software will use multiple threads to perform different things like updating the display, updating status information, spell checking as you type, printing, rendering thumbnails etc.

eg. in iPhoto, spawn a thread per thumbnail preview.

Then let the OS decide which core to run the thread on.

The problem is, OSX's threading leaves something to be desired when you compare it to other multi-threaded OSs like BeOS, QNX or even Windows.

There's been great pushes by Apple and it's developers to make the OS and applications more thread friendly and so multi-CPU friendly but at the same time OSX still has issues and the beachball of death still appears far too often because of the poor threading in OSX and Cocoa, especially on single-cpu Macs.

Threading, done well, really helps a system with multiple processors as anyone who has written software for systems with 1024 processors will tell you*.





* ok, they were all 1-bit processors but the tech still applies. ;-)
post #74 of 280
Quote:
Originally posted by g3pro
Do people forget that AMD maxes out at around 2.6ghz JUST LIKE THE G5? Of course they do.

2.8GHz now. They just bumped it up with the release of the FX-57.
post #75 of 280
Quote:
Originally posted by melgross
This is all very optimistic, but you're getting ahead of yourself.

No one is going to manufacture Macs if they don't have control of what they make. As was shown with the first round of clones, it can get out of Apple's control .

Apple won't allow clones unless thet have enough business from other product lines to keep their computer business down to perhaps 25% of the total. If that occurs they could think of licensing. The business lost in hardware could then be made up by OS sales and royalty payments. But only if this would serve to increase the share of the market for the OS . This also failed to occur before.

There is no guarantee that this could work. I doubt very much that we would see anything occuring in 2006, or even until after the transition is complete at the end of 2007.

Yes, it's optimistic. Apple could tank. So could 100 other companies with finances in far worse and some even in better position (other comp companies, Delta, GM, etc.). The transition could be a complete flop, and Apple could be reduced to rubble in a few years. Nothing is guaranteed.

What I'm saying is that Apple has been slowly learning to play with the rest of the PC industry over the last decade or so. It is now in the process of removing the last barriers to its self-imposed ghettoization. Personally, I feel if it tries to artificially keep a headlock on hardware, it'll be eaten alive by hackers and pirates, and limp along with 2-3% marketshare for quite a while. If, however, it opens itself up, and takes the opportunities provided by the iPod "halo effect" seriously, then things become very interesting.

Also, I was not necessarily talking about liscensing. That may happen. But I don't know if that's really an option (it involves too may legal and financial knots for my small brain to handle). I'm talking about just allowing OS X to work on an off-the-shelf PC. They would still have to buy a copy of OS X. If they want a computer with OS X preinstalled, they have to buy a Mac. (This has the added benefit of not pissing MS off so much right away, since all those Dells will still have a copy of Windows.)

Sure, you might get some PC mags complaining that you have to pay an "Apple tax." But then again, OS X isn't all that much more expensive than an anti-virus program.

DCQ
post #76 of 280
I think Apple will do the minimum necessary for the next revision of PowerMacs before they move to Intel Conroe in 2007.

This means, Apple would ship a PowerMac G5 with 2.3 to 2.5+ GHz 970mp processors in 1H 06 using the existing G5 chipsets, perhaps slightly modified for PCIe/DDR2, with probably 2 standard configurations: a 2.3 970mp config and a 2x2.5 970mp config. 1H 06 gives IBM another 6 months to optimize their 90 nm fab to get the 2.5 GHz 970mp down to ~100 Watt numbers, hopefully.

The iMac could use a 2 GHz 970mp, but it's going to be too hot for the enclosure I think. Maybe a 1.8 GHz 970mp though.

The low power 970fx chips are interesting. Whether they are low voltage (<1.0V) or low-k/DSL chips (or both), I don't know, but it appears to make a 1.8 GHz Powerbook G5 possible. This means Apple has choice to ship a 1.8 GHz Powerbook G5 or a 1.8 GHz Powerbook G4 in Q3/Q4 05. I wouldn't care to guess, but I think they have a realistic option of shipping a PB G5 instead of a 7448-based PB G4.

I would not worry about the usability of 4 cores and whether apps can use them or not. The vast majority (90+%) of Macs, including Mac/Intel machines, will be 2 core machines in the hands of Joe-consumer. This situation is well covered. The 4-core machines will be used by people who use software that will take advantage of them. This situation is the normal way of things.
post #77 of 280
Quote:
Originally posted by DCQ

. . . I'm talking about just allowing OS X to work on an off-the-shelf PC. They would still have to buy a copy of OS X. If they want a computer with OS X preinstalled, they have to buy a Mac. (This has the added benefit of not pissing MS off so much right away, since all those Dells will still have a copy of Windows.) . . .

DCQ


Apple gets most of its revenue from hardware. How would you suggest this be replaced if people begin buying cheap PCs to run Mac OS X? Regarding MS, did you ever think that Apple might have an agreement with MS to "not" run Mac OS X on generic PCs. This could be a threat to Windows dominance. In return for not running on generic PCs, MS may have given assurance of continued support from their Mac business unit.
post #78 of 280
Quote:
Originally posted by DCQ
Yes, it's optimistic. Apple could tank. So could 100 other companies with finances in far worse and some even in better position (other comp companies, Delta, GM, etc.). The transition could be a complete flop, and Apple could be reduced to rubble in a few years. Nothing is guaranteed.

What I'm saying is that Apple has been slowly learning to play with the rest of the PC industry over the last decade or so. It is now in the process of removing the last barriers to its self-imposed ghettoization. Personally, I feel if it tries to artificially keep a headlock on hardware, it'll be eaten alive by hackers and pirates, and limp along with 2-3% marketshare for quite a while. If, however, it opens itself up, and takes the opportunities provided by the iPod "halo effect" seriously, then things become very interesting.

Also, I was not necessarily talking about liscensing. That may happen. But I don't know if that's really an option (it involves too may legal and financial knots for my small brain to handle). I'm talking about just allowing OS X to work on an off-the-shelf PC. They would still have to buy a copy of OS X. If they want a computer with OS X preinstalled, they have to buy a Mac. (This has the added benefit of not pissing MS off so much right away, since all those Dells will still have a copy of Windows.)

Sure, you might get some PC mags complaining that you have to pay an "Apple tax." But then again, OS X isn't all that much more expensive than an anti-virus program.

DCQ

By opening up, you're inviting Apple to become the next Microsoft and Mac OS X to become the next Windows. No thanks.

I would personally rather see Apple limp along at 2-3% marketshare forever than see a beautiful OS 'reduced to rubble'.

I think the limping along is what pushes Apple to try things that other people don't dare. I'd hate to see Mac marketshare reach more than 10%.
post #79 of 280
Quote:
Originally posted by sunilraman
2005: year of HD, year of Intel bombshell
2006: year of transition
2007: starting to reap great benefits of transition
2008: ready to hit 10% market share, entry into business market

.........
.........

Apple just has a tough road ahead in the Mac business. Sure the iPod is going great and Mac sales as a percentage of increase in sales is astounding, but there are some serious problems.

Problem 1:

I used to think that games didn't matter. I bought my kids 2 Mac Minis to replace their WinAMD boxes. One was used for a few weeks, the other almost a whole month.

I walked into my son's room one night and quickly glanced and saw Windows on the monitor. I asked him where he got Virtual PC. I started lecturing him that downloading apps was bad. He said he put the Mini back in the box and into his closet. Why, Because he could not play games. Both my kids went back to their Windows boxes. Hello eBay.

I hardly think my situation is atypical. Just go to CompUSA. They have an 8' section of Mac games. Wahoo! However their are 2 - 40' isles for PC Games.

I just don't ever see a 10% market share. I don't think Microsoft sees it either, hence the Mac business unit is still around.

Problem 2:

I work for a fortune 100 company. It's an all WinTel setup. Most of corporate America is. Now my company does not have Intel twisting their arm, nor does Intel withhold kickbacks if they use AMD.

The vast business world won't even give AMD a chance, a simple hardware change that is transparent to the user and requires no changes in terms of IT support.

How can you even seriously think that any company is going to look at OS X if AMD can't even get traction?
post #80 of 280
Quote:
Originally posted by kresh
Apple just has a tough road ahead in the Mac business. Sure the iPod is going great and Mac sales as a percentage of increase in sales is astounding, but there are some serious problems.

Problem 1:

I used to think that games didn't matter. I bought my kids 2 Mac Minis to replace their WinAMD boxes. One was used for a few weeks, the other almost a whole month.

I walked into my son's room one night and quickly glanced and saw Windows on the monitor. I asked him where he got Virtual PC. I started lecturing him that downloading apps was bad. He said he put the Mini back in the box and into his closet. Why, Because he could not play games. Both my kids went back to their Windows boxes. Hello eBay.

I hardly think my situation is atypical. Just go to CompUSA. They have an 8' section of Mac games. Wahoo! However their are 2 - 40' isles for PC Games.

I just don't ever see a 10% market share.

Problem 2:

I work for a fortune 100 company. It's an all WinTel setup. Most of corporate America is. Now my company does not have Intel twisting their arm, nor does Intel withhold kickbacks if they use AMD.

The vast business world won't even give AMD a chance, a simple hardware change that is transparent to the user and requires no changes in terms of IT support.

How can you even seriously think that any company is going to look at OS X if AMD can't even get traction?

I think you're painting a very narrow-minded scenario...with only 2 problems without taking into account several other factors such as the halo-effect, the malware effect, and many other factors that have a good chance of pushing Apple beyond 5% and even reach as high as 10%.

Apple made a huge mistake in the mid 90s by allowing System 7 to stagnate and frustrate a lot of people. Sure, it was still superior to Windows 95 in many ways, but not enough to justify staying on a Mac...this is how Apple lost half its marketshare. Then Motorola problems made another chunk of people switch.

Some people are now starting to realize that MS OSs are the same no matter how many times you upgrade. You still get viruses, malware, etc...combined with the fact that Apple isn't letting OS X dwindle like they let System 7 dwindle and the iPod which is making heads turn towards Apple, combined with the fact that people are tired of malware (at least those aware enough that they've got malware installed on their computer somewhere), Apple has a brighter future than it ever had in 15 years.

With the Intel switch, Apple is giving themselves a break from defending their choice of chip architecture and will let the OS be the determining factor for consumers.

Macs will have the same hardware speed as everyone else in the market...but with an OS that has the polish of the iPod interface. And one that isn't affected (yet) by viruses and malware.

It won't take much before games start pouring onto the Mac like they did back in the late 80s, early 90s (Sierra, LucasArts, etc. all the big game developer names were right there with their Mac versions)...10% is plenty to make developers think twice before deciding not to port to Mac. The port will also be much more simple to do.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Future Apple Hardware
AppleInsider › Forums › Mac Hardware › Future Apple Hardware › IBM unveils dual-core PowerPC chips up to 2.5GHz