G5 = POWER 5?

[MaCommentary]

<a href="http://www.theinquirer.org/?article=4266" target="_blank">http://www.theinquirer.org/?article=4266</a>

Check out that article... it's rather boring for most of it... but I'll pull something out of it that might back up all of the other claims of Apple going to IBM over MOT...

"As it (POWER4) sheds the remaining baggage that blocks fast speed ramps and moves towards the mainstream (Apple?), this architecture could be well positioned for the final 64-bit showdown."

"Don't be surprised to see Apple Macs or thin blade servers based on POWER5."

just more fuel for the fire...

[Rhumgod]

[quote]Originally posted by MaCommentary:
<strong>"Don't be surprised to see Apple Macs or thin blade servers based on POWER5."

just more fuel for the fire...</strong><hr></blockquote>

Oh man, just what we need in these days of pre-MWNY hype, more fuel.

In the words of James Hetfield, "gimme fuel, gimme fire, gimme that which I desire..."

[Nitride]

Because someone is a fan of a different CPU does not extend to Apple using it in its future hardware.

The G4 itself is not the problem. A computer it a complex system with many interactions between various discreet components including PCI bridges/memory controllers, CPUs, memory and more.

The biggest single issue in the G4 (computer) is that the memory bus is just not fat enough to allow the G4 to reach its full potential at ANY clockrate. Apple is handicapping it with a non-DDR &gt;133 MHz bus. The addition of heaps of ultra fast L3 cache is a stop-gap solution.

Part of the blame rests with Motorola, but Apple is the company designing the entire system. Apple should have been on DDR last year but they insist on holding it back for technical reasons (so they say) and cost issues blah blah blah. Slumping G4 sales should prove that the old tech is NOT what we want!

JUST SHIP it and let us see how well it will run Apple, geeze. Xserve proves they CAN do it, at least half-way. Now lets go ALL the way and blow away some DELLs.

[MaCommentary]

Nitride:
1) true to an extent. what you said is true, but Motorola has yet to produce 8500s that are consistantly stable above speeds that look like iBook and iMac processors.

2) you should concidder looking at the iBrick forum.

[Lemon Bon Bon]

Power 4.

Bring it on.

Apple are going to need some ammo'.

I think dark clouds loom. Salvo's fired to and from the x86/PPC platforms. Microsoft's recent statements sound ominous.

Having a processor that outguns the Wintel opposition is becoming paramount in my view. The G4 like the G3 is old. The G4 should go into the consumer lines and give up the ghost at being a workstation processor. Half cocked DDR/Bus won't save it from the beatings it gets in Lightwave benches.

IBM seem to be getting ready to do some kick ass with the Power4...forward winding...the idea of multicore designs for a workstation Apple seem to hold out the prospect of PPC finally delivering the desktop performance dominance it promised so long ago.

Bring it on, Microsoft.

Apple must be pretty confident that it has something up its sleeve to counter attack a Microsoft assault.

Lemon Bon Bon

[BRussell]

[quote]Originally posted by Nitride:
<strong>Because someone is a fan of a different CPU does not extend to Apple using it in its future hardware.

The G4 itself is not the problem. A computer it a complex system with many interactions between various discreet components including PCI bridges/memory controllers, CPUs, memory and more.

The biggest single issue in the G4 (computer) is that the memory bus is just not fat enough to allow the G4 to reach its full potential at ANY clockrate. Apple is handicapping it with a non-DDR &gt;133 MHz bus. The addition of heaps of ultra fast L3 cache is a stop-gap solution.

Part of the blame rests with Motorola, but Apple is the company designing the entire system. Apple should have been on DDR last year but they insist on holding it back for technical reasons (so they say) and cost issues blah blah blah. Slumping G4 sales should prove that the old tech is NOT what we want!

JUST SHIP it and let us see how well it will run Apple, geeze. Xserve proves they CAN do it, at least half-way. Now lets go ALL the way and blow away some DELLs.</strong><hr></blockquote>Apple does not control the bus capabilities of the G4. According to Motorla, the G4 has a maximum bus of 133. The best they could do is pull an xserve.

[torifile]

damn, the p.o.s. I've got in my office is so slow, I forgot what I was going to write....


Ah, yes. Let's say the next mac proc is a power5 (it wouldn't be a G5 b/c that's mot designation). How much would people bitch about it costing $5000+ for a low end tower? Fugetaboutit.

[Outsider]

The POWER4 has an impressive core, even if it was single. It has a 14 stage pipeline second longest only to the P4 and lots of room to grow. Already 64bit compliant, a proprietary but fast system bus (10GBps max transfer rate), excellent cache support, a first rate FPU, and a modular design. With the proper design (taking some elements from the 750FX design) this can even be made for use in a portable device.

IBM can do it. Does Apple want it?

In my opinion such a processor would have:
Single POWER4 derived core;
Altivec-clone (Velocity Engine II);
500MHz GX bus to memory/PCI/peripheral controller;
512KB L2 cache;
Up to 8MB L3 cache (optional);
state of the art 130nm SOI process.

The PowerPC 850.

[ 07-15-2002: Message edited by: Outsider ]</p>

[mooseman]

...man, I get tired of hearing about how Apple needs to introduce a DDR bus. Its not up to Apple. Apple can build all the DDR or RDRAM motherboards they want all day long, but if the processor doesn't support said buses, then the point is all moot. Motorola hasn't screwed Apple just with its miserably slow clock increases, but an equally slow implimentation of anything faster than 133MHz on the bus front. The IBM 750fx, IIRC, supports a 200MHz bus among other sweet features and is gonna be on a .13process, something Motorola only dreams about at this point.

Its plain a simple, Motorola is Apple's albatross.

[Anders]

Why anyone (beside people locking their computers in another room than where they sit) would want a CPU with more stages is beyond me. And no I'm not under the influence of Jobs Anti-Mhz-myth-myth-RDF-generator, but I prefer cool processors anytime.

You really cant find as cool laptops as the ice on the other side and if they have trouble putting a G4 in it what would they do with something with 14 stages.

Long pipelines is cheating and will hurt the laptops on the long run (and it ain't a mac if it ain't a laptop :cool: )

[Lemon Bon Bon]

You said it Outsider.

The IBM Power Chip would give Apple something it doesn't appear to have at the present time.

Options.

Oh. And industry leading performance.

Performance which could scale to Hollywood SGI level workstations...to yer standard 'power'Mac range. The 'multi-core' design of the Power 4 gives flexibility...if they are indeed designing it to ramp the mhz faster and lose some of the more expensive design baggage holding the chip back from being a great workstation/desktop cpu.

Paying those Apple premiums? Gladly if they got da performance mulla to go with it.

A Power 4 derivation would give Apple a roadmap of sorts. IBM aint going nowhere. Leading edge CPU is their business. And the Power 4 seems to have legs. It's got awesome power. Great FPU...and it's only at 1.3 ghz? A move to .13 etc should eek out more performance. Moto' on the other hand increasingly look like a struggling company in crisis...

The trickledown effect of IBM's cutting edge stuff (which need fridges to 'em, eh?) into desktop workstation performance. Well...that would seem less ambiguous than the love/hate 'ati' style relationship Apple seems to have with Moto'. IBM at least inspires confidence in me.

Lemon Bon BON

[ 07-15-2002: Message edited by: Lemon Bon Bon ]</p>

[Outsider]

Long pipelines is not cheating and if designed properly is an advantage. Given a full pipeline, it will generate the same IPC as a processor with a shorter pipeline. If you minimize stalls in the pipeline (by using a strong branch prediction unit) then you can keep it full and pumping out a high IPC at a high cycle rate.

This combination and smaller processes using SOI will keep heat issues under control, adequate for portable use.

[123]

[quote]Originally posted by Anders:
<strong>Long pipelines is cheating </strong><hr></blockquote>

Unfortunately, you are "under the influence of Jobs Anti-Mhz-myth-myth-RDF-generator". A longer pipeline simply allows you to run your processor at faster speeds. Hazards do occur, but in time critical sections not nearly as often as Steve makes you believe. Very often, you have big loops with a very good branch hit/miss ratio.

Your heat concerns are valid, but in general, more pipeline stages mean advanced technology and more computing power, not cheating (of course, 20 stages are a lot...)

123

[willywalloo]

I wanna believe that IBM would be powering the future's powermac's. My computer now runs off of IBM's chip and it screams.

Only problem is that a close contact that I know who works for IBM told me that their chip manufacturing has slowed and they now have multi-thousand dollared moto desktops coming in.

What I got out of the whole thing was that they basically weren't going to power the low-end markets anymore, but rather surge with the higher-demanding super(duper)computers. This could include powermac's but then where would you go with alti-vec apps?

Ideally the servers and powermac lines would run off of IBM chips and have a separate moto-Alti-vec chip connected by a bus running at 1:1 clock speed.

[hotboxd]

I've been saying for a while that Apple will probably use the Power5 as it's next gen. chip. Paraphrased from an IBM article "Power5 with be faster, smaller, cooler, and cheaper than the Power4." Apple has more than enough money to convince IBM to design a custom derivative of the Power5 for use in Apple computers.

The Power5 has been under development for a while. If Apple realized that Moto wasn't going anywhere during the G4 debacle, theres been plenty of time for them to go to IBM and work out an arrangement for development of a cheaper, cooler Power5 derivative with an Altivec-compatible SIMD or co-processor. Subsequent iterations of this design would eventually move it into the consumer and laptop areas, thus freeing Apple from the sinking ship that is Moto.

Who knows whether this will happen or not, but it seems like the smart thing to do.

[Spartacus]

A custom Power 5 would be more likely than a Power 4. IBM don't want Apple to go in the same market segment as them with high end Servers.
IBM will never sell high end chips to Apple. A larger agreement would be possible like licence MacOS X Server to IBM for there high end server and Apple sells only low end servers. Apple could than use the same chip but in max. 2 cores chip.

what do you think about this?

What is the power consumption for the Power4 vs PentiumIV?

[Cobra]

I think this Power 5 stuff is a little more credible than the G5 stuff we have been hearing about.

My only concern deals with Altivec. IBM has not taken a shining to it. :(

[Barto]

"It should be much cheaper, cooler and less power-guzzling than POWER4."

I'm happy and sad. Happy that this will probably be the G5 (what else is there on the horizon?). Sad that it's coming in 2004, not 2003.

JYD won't be happy, he reckons that Apple will only survive if a new chip is introduced at MWSF :rolleyes:

Cobra: AltiVec is powerful, and SIMD is the future. If the Power5 is going to attack a larger market than than the Power4 (read: Apple Macs), then you will see a "Velocity Compatible" SIMD unit.

Spartacus: I've always thought that Apple licencing Mac OS X to IBM a real possibility. I mean, it would cost Apple huge $$ to "Steve" the mainframe market. Why not let IBM do it for them?


"It doesn’t matter whether you have Mac, Windows, UNIX or Linux clients — or a combination thereof — on your network. Mac OS X Server provides cross-platform support for native file sharing, as well as Apache web server and WebDAV server, POP and IMAP mail, ftp, QuickTime Streaming Server, DNS and DHCP — right out-of-the-box. Windows, UNIX and Linux are not treated any differently, and will enjoy benefits that Mac users take for granted.*"

Shake in dem booties Microsoft.

Barto

[ 07-16-2002: Message edited by: Barto ]</p>

[The Mactivist]

[quote]Originally posted by Cobra:
<strong>I think this Power 5 stuff is a little more credible than the G5 stuff we have been hearing about.

My only concern deals with Altivec. IBM has not taken a shining to it. :( </strong><hr></blockquote>

I thought IBM had already licensed it from MOTO. What does MOTO have to loose in licensing the instructions? They make money and they don't have to spend a penny on producing new chips. Sounds like a good deal to me. :D

[Telomar]

[quote]Originally posted by The Mactivist:
<strong>What does MOTO have to loose in licensing the instructions? They make money and they don't have to spend a penny on producing new chips. Sounds like a good deal to me. :D </strong><hr></blockquote>

The IBM G3 with Altivec would attack the very heartland of Motorola's semiconductor business, the embedded market.

[Addison]

[quote]Originally posted by Telomar:
<strong>

The IBM G3 with Altivec would attack the very heartland of Motorola's semiconductor business, the embedded market.</strong><hr></blockquote>

Not if the license was worded correctly to prevent that.

[hotboxd]

IBM has recently been taking a shine to the whole SIMD idea, the PPC750 Sahara successor is rumored to have a SIMD unit in it's design. Regardless, it's perfectly feasible for Apple to reengineer the Velocity Engine to something that is compatible with the Altivec instruction set without Motorola's permission as long as they follow the correct procedures.

I thought the Power5 was supposed to be released sometime in 2003?? <img src="confused.gif" border="0">

[A Random Walk]

[quote]This is posted at Mot's web site:
<strong>Session Topic: PowerPC SIMD Instructions (H1121)
Level: Advanced
Presenter: Craig Lund, Mercury Computer Systems
Abstract: Motorola's new e500 core introduces new SIMD instructions. Early in the week Motorola's Kumar Gala will introduce the e500 SIMD instructions for the first time. Mr. Lund's presentation is a follow-on that contrasts the e500 SMID functionality with the AltiVec SIMD functionality currently available in the G4 and G4+ cores. Mr. Lund's tutorial will help users decide which SIMD instruction set is best for their application's requirements. The tutorial will start at a high level and transition into a review of source code optimized for each processor.</strong><hr></blockquote>

This may be old news...but, I wasn't aware that Mot had designed brand new SIMD instructions to go with the 8540. BTW, the session is scheduled the week of 7/21.

[MaCommentary]

[quote]Originally posted by The Mactivist:
<strong>

I thought IBM had already licensed it from MOTO. </strong><hr></blockquote>

They liccensed it for only certain speeds and only on the implamentations that it's arleady... essentially they can't integrate into their own chips,,, :(

[dumpster]

This just in:

<a href="http://www.reuters.com/news_article.jhtml?type=businessnews&StoryID=1209475" target="_blank">Motorola beats analyst expectation and posts their biggest loss in history.</a>

At this rate, Apple could just buy them.

[MicrosoftOsXp]

Wouldn't it make sense to have a workstation - super mac for the high end video market. A magical place where money doesn't mean a whole lot for computing equipment. Apple's buying up a bunch of high end company's so why not have a power 4/5 workstation that blows the crap out of intel's performance, and has an awesome operating system, with awesome software.

[taboo]

[quote] ....currently available in the G4 and G4+ cores.... <hr></blockquote>

Um.....G4+?

[Programmer]

[quote]Originally posted by taboo:
<strong>

Um.....G4+?</strong><hr></blockquote>

The 7450 and 7455 are the G4+, although that term was never officially used. They are the lengthened pipeline versions.

[taboo]

[quote]Originally posted by Programmer:
<strong>

The 7450 and 7455 are the G4+, although that term was never officially used. They are the lengthened pipeline versions.</strong><hr></blockquote>

Many thanx. I remembered the term used around here by someone relating to a future CPU. Must have been an older post.

[Aphelion]

[quote]Originally posted by MicrosoftOsXp:
<strong>Wouldn't it make sense to have a workstation - super mac for the high end video market. A magical place where money doesn't mean a whole lot for computing equipment. Apple's buying up a bunch of high end company's so why not have a power 4/5 workstation that blows the crap out of intel's performance, and has an awesome operating system, with awesome software.</strong><hr></blockquote>

Oh yes, this is coming but at what price? Apple could have had it all in 2000 with the 7410 (low power G4 @ 500Mhz) in a Quad and then an 8 way Powermac (by now). My sig is from 2000 when I became a Switcher from 15 years on Windows.

[Jules]

Just a quick question with respect to changing to another CPU, how similar is the power architecture to the Power PC? They derived from the same original design I hear. It would surely behove Apple to make the transition as smooth as possible, but they cant just sit around on their chuffs and do nothing, shit, the g4 has been out for 3 years now with what, a doubling of clock speed. Even the more optimistic amongst us would have to admit that Apple has to do something here.

[MaCommentary]

[quote]Originally posted by Jules:
<strong>Just a quick question with respect to changing to another CPU, how similar is the power architecture to the Power PC? They derived from the same original design I hear. </strong><hr></blockquote>

The G3 (turned G4 with altivec) and the POWER4 both come from the same basic PPC design. This means that they both handle data the same way. The actual architecture is quite different, but in terms of Apple this means absoubtly nothing. Apple would have to change pretty much only it's driver software in order to use the POWER series instead of hte MOT 7450s.

[Programmer]

[quote]Originally posted by MaCommentary:
<strong>
The G3 (turned G4 with altivec) and the POWER4 both come from the same basic PPC design. This means that they both handle data the same way. The actual architecture is quite different, but in terms of Apple this means absoubtly nothing. Apple would have to change pretty much only it's driver software in order to use the POWER series instead of hte MOT 7450s.</strong><hr></blockquote>

This explanation seems a bit confusing, I'll try to restate it:

The POWER4's instruction set is compatible with the PowerPC instruction set used in the G4, except that it doesn't include AltiVec and there are a few system-level issues that only the OS would care about. Application software and most of the OS could run on a POWER4 unchanged. The parts of the OS that would need to change have been carefully isolated by Apple in the MacOSX kernel and so supporting a new processor is a fairly straightforward matter. I'd be a little surprised if Apple didn't have a POWER3 and/or POWER4 machine somewhere in its labs running MacOSX as an experiment.

[Gamblor]

Hell, I'd be surprised if Apple didn't have OSX running on at least one of HP9000, MIPS, and Sparc processors, and possibly all three. Gotta keep them options open. ;)

[Outsider]

JObs recently told analysts that the PowerPC future looked bright. They were most likely referring to an IBM design in the pipeline, perhaps even ready to go.

[Appleworm]

[quote]Originally posted by Gamblor:
<strong>Hell, I'd be surprised if Apple didn't have OSX running on at least one of HP9000, MIPS, and Sparc processors, and possibly all three. Gotta keep them options open. ;) </strong><hr></blockquote>

I always dream of Apple and SGI working together...
:)

[Appleworm]

[quote]Originally posted by Aphelion:
<strong>

OSX + Duals, Quads & Octos = World Domination

</strong><hr></blockquote>

Yes !!!!
If we have the G5, what the G5 is (PWR4/5, Hammer, Moto G5,...)
Hope Steve doesn't think different for this ;)

MWNY02 : <img src="graemlins/oyvey.gif" border="0" alt="[No]" />

[anakin1992]

[quote]Originally posted by Spartacus:
<strong>A custom Power 5 would be more likely than a Power 4. IBM don't want Apple to go in the same market segment as them with high end Servers.
IBM will never sell high end chips to Apple. A larger agreement would be possible like licence MacOS X Server to IBM for there high end server and Apple sells only low end servers. Apple could than use the same chip but in max. 2 cores chip.
</strong><hr></blockquote>

even though it sounds that apple and ibm might be competing for high end server market when apple uses power4/5/6 line chip, in reality, they are working on different market sectors. people who want to use mac mostly likely will not buy an AIX or linux line of ibm machines. most important of all, apple is focusing on consumer market. so it should be ok.

as to the power comsumption on power4, somewhere i read and it is around 100W.

[ 07-20-2002: Message edited by: anakin1992 ]</p>

[anakin1992]

[quote]Originally posted by Appleworm:
<strong>

I always dream of Apple and SGI working together...
:) </strong><hr></blockquote>

apple and sgi are two jewels of workstation computing in my mind since apple promotes the ease of use and funky gui into computation and sgi really shows the world the wonder of graphical effect. they shed the light on computer box...

it would be great if they work together. sgi will bring talents to apple on high end media server platform for hollywood while apple will show sgi how to populate its tech advance into consumer market.

sgi can also lead apple into defense market where the taxpayer money go to and is wasted...

now sgi is a $241m market cap company while apple is a $5.26n market cap corporation with almost $5b in hand... why not?

[anakin1992]

[quote]Originally posted by Programmer:
<strong>

This explanation seems a bit confusing, I'll try to restate it:

The POWER4's instruction set is compatible with the PowerPC instruction set used in the G4, except that it doesn't include AltiVec and there are a few system-level issues that only the OS would care about. Application software and most of the OS could run on a POWER4 unchanged. The parts of the OS that would need to change have been carefully isolated by Apple in the MacOSX kernel and so supporting a new processor is a fairly straightforward matter. I'd be a little surprised if Apple didn't have a POWER3 and/or POWER4 machine somewhere in its labs running MacOSX as an experiment.</strong><hr></blockquote>

with the power of two 64-bit cpu cores cranching the number, i really don't see the need for altivec, as i always regard it as a workaround for apple/moto. but since apple and moto promoted it for a while, i guess many developers have used this feature. so it would generate certain bad response if apple's new cpu does not have altivec.

power4 is compatiable with powerpc architecture with some improvement on the parts that hinders the faster cycle design. also, keep in mind, this chip is powerful and intented to be used in server. therefore, it has many features which does not exist in consumer market: reliability and fault tolerance management. if apple has its will to use it and move consumer product to server level (not everything but certain features), i am not surprisde to see a future mac can run and run and run...

apple: answer is out there, please get it and fire it up and move on...

[anakin1992]

[quote]Originally posted by Outsider:
<strong>The POWER4 has an impressive core, even if it was single. It has a 14 stage pipeline second longest only to the P4 and lots of room to grow. Already 64bit compliant, a proprietary but fast system bus (10GBps max transfer rate), excellent cache support, a first rate FPU, and a modular design. With the proper design (taking some elements from the 750FX design) this can even be made for use in a portable device.

IBM can do it. Does Apple want it?

In my opinion such a processor would have:
Single POWER4 derived core;
Altivec-clone (Velocity Engine II);
500MHz GX bus to memory/PCI/peripheral controller;
512KB L2 cache;
Up to 8MB L3 cache (optional);
state of the art 130nm SOI process.

The PowerPC 850.

[ 07-15-2002: Message edited by: Outsider ]</strong><hr></blockquote>

not sure the process to make a single core version of power4, since the min. config for a power4 has two cores. the biggest archievement on pqwer4, in my mind, is that ibm created cpu switch within the die. ibm uses this type of crossbar switch between multiple cores to achieve on-die parellel processing. this is quite different from what intel does on its itanium chip and what sun does it on its ultrasparc v. of course it could use one cpu while another one is turned off. well, i think we lost some points or sense to do so, don;t we?

but i do agree that two core power4 will be too expensive for consumer. i can easily see a targeting price for such machine as $7000+.

my bet on what apple might do;
1: keep g4 line and improve it on motherboard. build low end/entry level power mac. price ~ $1000-2000. imac line

2: have g4 with two cpu on a SMP machine with higher clock rate. this is for middle level power mac. price ~ $2000 to 4000. high end imac and low end desktop

3: have pwer4 with two core on one die for a high end power mac. price ~ $4000 to 10,000. high end desktop

4: have power4 with 2, 4, 8, or 16 dies for high end server line. long gone the cheap server, replica of existing power mac line. price ~ $4000 to 100k

well, we have the right to speculate, don't we:-)?

[ 07-20-2002: Message edited by: anakin1992 ]</p>

[MaCommentary]

[quote]Originally posted by hotboxd:
<strong>The Power5 has been under development for a while. If Apple realized that Moto wasn't going anywhere during the G4 debacle, theres been plenty of time for them to go to IBM and work out an arrangement for development of a cheaper, cooler Power5 derivative with an Altivec-compatible SIMD or co-processor. Subsequent iterations of this design would eventually move it into the consumer and laptop areas, thus freeing Apple from the sinking ship that is Moto.

Who knows whether this will happen or not, but it seems like the smart thing to do.</strong><hr></blockquote>

I'd recon that they figured that out when the G4 was stuck at 500MHz for a year.

[vinney57]

I'm confused MaCommentary, is your comic supposed to be funny? Or are we supposed to make make allowances 'cos you are really young?

[Programmer]

[quote]Originally posted by anakin1992:
<strong>with the power of two 64-bit cpu cores cranching the number, i really don't see the need for altivec, as i always regard it as a workaround for apple/moto. but since apple and moto promoted it for a while, i guess many developers have used this feature. so it would generate certain bad response if apple's new cpu does not have altivec.
</strong><hr></blockquote>

Almost every processor these days includes a SIMD unit... they provide a significant performance advantage in many kinds of calculations. Even IBM seems to have come around on this one and will be including SIMD units in their future processors -- hopefully AltiVec compatible.

<strong> [quote]
power4 is compatiable with powerpc architecture with some improvement on the parts that hinders the faster cycle design. also, keep in mind, this chip is powerful and intented to be used in server. therefore, it has many features which does not exist in consumer market: reliability and fault tolerance management. if apple has its will to use it and move consumer product to server level (not everything but certain features), i am not surprisde to see a future mac can run and run and run...
</strong><hr></blockquote>

Why is it so many people insist on talking about the POWER4 being directly used by Apple? This is very unlikely. Instead we are likely to see IBM apply the POWER4 technologies to the design of a new processor which is specifically intended for Apple's use in desktop (and possibly notebook) computers. In this new processor a different set of design trade-offs can be made -- it could be single, or dual (or more) core. Its server oriented bus would be replaced with something more appropriate to Apple's machines. Caches would be resized. The supposed fault tolerant / reliability features could be removed if they don't make sense. Features like AltiVec and the rumoured Apple Pi could be added.

The key point is that IBM will likely provide a new processor to Apple, and they have a lot of existing technology to build from.

[MaCommentary]

First, let mee say that this is WAY off topic, but I'll respond to vinney57's comments none the less.

Reality High is a DRAMATIC comic. Sometimes a comic only tells a story. If you're looking for humor then don't look at annother veteran poster's 'comic' either. You might have heard of him: Leonis. I wouldn not compare myself to Leonis, he/she's on a whole different level... but the comic Sanmothy (http://www.sanmothy.com) isn't funny either.

Cartooning is a meduim and it's products have been labled 'comics' since before I can remember whether or not they are funny. Jeremy is the story of a very troubbled individual who affects the lives of the people around him. Those people go on to be the main characters in the main drag which will take off in September. At that point it may not be so heavy, but for now I'm tackling a serious issue in a 'comic' and so no; it is not meant to be funny.

I also resent the fact that my age seems to have affected my credibility.

Finally, I'm not asking for allowances. I've got a story to tell through my characters and if you don't like it (although I hope that most people do) then don't read it. Simple as that.

[Bigc]

you mean we don'tget 32 MB of cache

[Barto]

Assuming that Apple will use the Power5 in 2004.

1999-2001 G4
2001-2003 G4+
2003-2004 Still a G4+?

The scariest time for Apple might be 2003, when they fall further behind with an 7455.

What will they do for 2003?

Motorola says it will switch the G4 to 130nm, but all those layoffs and dodgey fabs that we hear about must slow up development sometime (after the huge amount of 1GHz CPUs they seem to be producing).

Motorola seems to not want Apple as a customer (every major customer except for Apple is mentioned on their web site), so maybe there is a deal when IBM fabs the G4+ for 2003 (with their superior fabs) to get higher clock rates for Apple, before moving on to the Power5 in 2004.

Also, a question. Would it be possible for a version of the G4 with dual MPX buses be possible, where there would be multiple pipelines from the northbridge and ring buses connecting each stage in multiple pipelines? Then, if there was 4-8mb of cache per CPU, in a quad CPU system there would be 16-32mb of total L3 cache, which the other CPUs could access over the dual MPX buses.

Barto

[MaCommentary]

A long time ago in a galaxy far far away, there was a lot of buzz about Apple buyiung MOT's semiconductor division. At MWNY, Jobs said that Apple will continue to innovate and innvest. It's the invest thing that I'm wondering about...

[Programmer]

[quote]Originally posted by Barto:
<strong>Assuming that Apple will use the Power5 in 2004.
</strong><hr></blockquote>

Bad assumption.

<strong> [quote]
Motorola says it will switch the G4 to 130nm, but all those layoffs and dodgey fabs that we hear about must slow up development sometime (after the huge amount of 1GHz CPUs they seem to be producing).
</strong><hr></blockquote>

That depends on who they are laying off. It is fairly normal for a company to protect its crown jewels, and Moto is headed toward out sourcing the fabbing so its possible that their G4 development team is unaffected. The first few layoffs out of a particular group tend to be culling of the weak anyhow, so they can actually improve a team until they get too deep.

<strong> [quote]
Motorola seems to not want Apple as a customer (every major customer except for Apple is mentioned on their web site), so maybe there is a deal when IBM fabs the G4+ for 2003 (with their superior fabs) to get higher clock rates for Apple, before moving on to the Power5 in 2004.
</strong><hr></blockquote>

IBM fabbing the G4 is possible (possibly has already happened in the past..?). And again, why do you assume that the POWER5 will end up in Macs? IBM is a large company and is capable of designing more than one processor at a time.

<strong> [quote]
Also, a question. Would it be possible for a version of the G4 with dual MPX buses be possible, where there would be multiple pipelines from the northbridge and ring buses connecting each stage in multiple pipelines? Then, if there was 4-8mb of cache per CPU, in a quad CPU system there would be 16-32mb of total L3 cache, which the other CPUs could access over the dual MPX buses.
</strong><hr></blockquote>

The MPX bus has a great many pins which would make such a design prohibative. Large amounts of cache provide diminishing returns. If they are going to work over the bus(es) like that, I would expect to see the bus replaced with something faster that uses fewer pins (HT or RIO). Then they could have multiple ports.

[Barto]

Programmer: I didn't mean that I though that Apple would start using the POWER5 in 2004, just as a theoretical senario (with any company, just replace IBM with x company), and what would happen in the meantime. "Assume" was a bad word, you're right.

Also, I was simply trying to get discussion going about what would happen if Apple choose to switch to a POWER5 or other chip not made by motorola, but with a launch date in 2004-2005. Would Apple licence the G4 from Motorola, and get a company with better fabs (like IBM) to fab it in the meantime to get higher clock-rates?

Thanks for the answers though

Barto

[ 07-24-2002: Message edited by: Barto ]</p>

[MaCommentary]

do the new cooling vents (see form on leaked photos) possibly have anything to do with a transition to a different processor... maybe a hotter one from the POWER series?

[hotboxd]

Maybe. If IBM is developing a derivative of the Power4 for Apple then that would make sense, but it doesn't make sense for a G4 on 0.13 process.

Dorsal and to a certain extent Moki have been hinting that early 2003 is the target for a bigtime power boost in the PowerMac line. I said earlier that theres been plenty of time for Apple to go to IBM to develop a derivative of the Power5 chip, well theres also been plenty of time for Apple to go to IBM for a derivative of the Power4 as well. This is maybe what we'll see.

Theres lots of people saying that the Power4 isn't going to happen because it's to costly, big, hot etc. IBM could theoretically take much of the basic design of the core processor from the Power4, slim it down significantly, lose the features that wouldn't be useful in a desktop (, and make it possible for single core versions to be manufactured. Again, Apple has had 2 years in which to organize this; that would be a decent amount of time for the development of a chip of this nature since it isn't being designed from scratch, but is merely a revising of an existing design. IBM definitely has the resources to do this (as well as the incentive, since Apple sells a lot of PowerMacs relative to IBMs server sales), all Apple would need to do would be to throw money at them, which they have plenty of. That would explain the heavily ventilated new case that has been floating around, and would also explain some of the pointers to early 03 as a big speed boost for the PowerMac line.

[BlueJekyll]

[quote]Originally posted by anakin1992:
<strong>

with the power of two 64-bit cpu cores cranching the number, i really don't see the need for altivec, as i always regard it as a workaround for apple/moto. but since apple and moto promoted it for a while, i guess many developers have used this feature. so it would generate certain bad response if apple's new cpu does not have altivec.

power4 is compatiable with powerpc architecture with some improvement on the parts that hinders the faster cycle design. also, keep in mind, this chip is powerful and intented to be used in server. therefore, it has many features which does not exist in consumer market: reliability and fault tolerance management. if apple has its will to use it and move consumer product to server level (not everything but certain features), i am not surprisde to see a future mac can run and run and run...

apple: answer is out there, please get it and fire it up and move on...</strong><hr></blockquote>

Shrey?

I agree, and with gcc now able to compile 64bit apps for ppc it seems like something along these lines will happen. The new processor could emulate altivec with 2 interger units...

Perhaps Apple and IBM have worked out a low power cuonsuming POWER4 for use in the power macs and xserve.

The thing that has me questioning a new processor being rolled out is xServe. That box only just started shipping, I could understand Apple just throwing a faster processor in it, but switching to a new processor? not yet. And why would Apple ship the PowerMac line with a more powerful processor before the xServe? I could be completely wrong though...

[Programmer]

[quote]Originally posted by BlueJekyll:
<strong>The thing that has me questioning a new processor being rolled out is xServe. That box only just started shipping, I could understand Apple just throwing a faster processor in it, but switching to a new processor? not yet. And why would Apple ship the PowerMac line with a more powerful processor before the xServe? I could be completely wrong though...</strong><hr></blockquote>

People keep saying this, but I disagree completely. The Xserve is a 1U rackmount and shoving a brand new processor that requires a monster cooling system into such a small space just doesn't make sense. There are still lots of PentiumIII U1 racmounts around -- these servers to not have to have the latest and greatest processor. The PowerMac is Apple's flagship Pro machine and people are constantly screaming for it to be as fast as possible, so it is actually a more sensible place to roll out a new processor.

[BlueJekyll]

[quote]Originally posted by Programmer:
<strong>

People keep saying this, but I disagree completely. The Xserve is a 1U rackmount and shoving a brand new processor that requires a monster cooling system into such a small space just doesn't make sense. There are still lots of PentiumIII U1 racmounts around -- these servers to not have to have the latest and greatest processor. The PowerMac is Apple's flagship Pro machine and people are constantly screaming for it to be as fast as possible, so it is actually a more sensible place to roll out a new processor.</strong><hr></blockquote>

Good point. I hope I am wrong, along with the many other people. :)

And now that i think about it, even with the QuickSilver server boxes, Apple always took longer to update them than the PowerMac...

[The Pie Man]

[quote]Originally posted by BlueJekyll:
<strong>And now that i think about it, even with the QuickSilver server boxes, Apple always took longer to update them than the PowerMac...</strong><hr></blockquote>

Just have to play devil's advocate here - I think if you asked on Apple's X server board the response you would mostly get would be the slower updates on the server side comes more from a lack of focus on their server products than a need to keep their prosumers happy.

That being said throughput is obviously more important than screaming processors in the server world...

[ 07-24-2002: Message edited by: The Pie Man ]</p>

[taboo]

[quote]Originally posted by BlueJekyll:
<strong>

Shrey?

I agree, and with gcc now able to compile 64bit apps for ppc it seems like something along these lines will happen. The new processor could emulate altivec with 2 interger units...

Perhaps Apple and IBM have worked out a low power cuonsuming POWER4 for use in the power macs and xserve.

The thing that has me questioning a new processor being rolled out is xServe. That box only just started shipping, I could understand Apple just throwing a faster processor in it, but switching to a new processor? not yet. And why would Apple ship the PowerMac line with a more powerful processor before the xServe? I could be completely wrong though...</strong><hr></blockquote>

Also, when looking for a server, you want proven, reliable tech. Taking a chance on a new mobo or CPU before it's well-tested is kinda asking for trouble.

[jerk]

[quote]Originally posted by BlueJekyll:
<strong>
I agree, and with gcc now able to compile 64bit apps for ppc it seems like something along these lines will happen. The new processor could emulate altivec with 2 interger units...
</strong><hr></blockquote>

May I remind you that the altivec has 32 128b registers, not 1?

[Added:] The altivec can do operations on all its' registers at once, that is the point with it and what makes it differ from the ordinary register bank. It is a SIMD unit, "single instruction multiple data".

Or do you mean that the new cpus will be 32 times as fast (not calculating with the possible gain of doing something else while running the altivec unit)?

Saying that the altivec is a work around is just plain false. It is as saying as an floating point unit is a work around. They speed up certain kinds of programs by utilizing silicon for certain operations.

[ 07-25-2002: Message edited by: jerk ]</p>

[Powerdoc]

[quote]Originally posted by Outsider:
<strong>The POWER4 has an impressive core, even if it was single. It has a 14 stage pipeline second longest only to the P4 and lots of room to grow.


[ 07-15-2002: Message edited by: Outsider ]</strong><hr></blockquote>
strange 14 stage pipeline in the power 4 , the same number discribed in some previous rumors about the G5. It appears that the people who create the rumors have copied these specifications on the power 4, or the people already knews that the G5 will be based upon the power 4 core.

[Junkyard Dawg]

The new tower design has some mondo cooling capability. That tower isn't designed to cool a G4....it's designed to cool a POWER 5! Seriously, what G4 puts out the sort of heat that would need a case with that sort of cooling capacity?

I think the new Towers are going to surprise us all. Remember the new LCD iMac? What was the update just prior to that one? It was a silent iMac update that was ultra-lame, just like the last tower update. That's the pattern, we get a lame, quiet update, then BAM!!! Apple drops a new Powermac on us with a new CPU that SCREAMS!.

It all makes sense. We'll find out soon enough, but my guess is that the G5 is coming sooner than we think. Apple's kept security on this extremely tight, but it would make sense. Remember all the reports back in December and January about G5 Powermacs? If there was some hitch and they were delayed because of a bug, then NOW is the time they would come out. Apple would postpone their introduction for one more revision cycle to make sure all the bugs were ironed out, and also to ensure there were no supply problems like when they introduced the G4.

I think it's coming. Apple is going to come! One giant climax to change the face of computing. The Powermac G5 is going to eat Pentium 4s for breakfast, AMD Clawhammers for lunch....and for dinner, it's going to go hungry because there's nothing left for it to eat!! It's going to be the BEAST of all CPUs, the one that even Apple has trouble cooling.

Anyone got a better explanation for all that insane cooling capacity of the new towers?

[Barto]

Just don't set yourself up for a fall, JYD.

Barto

[Eupfhoria]

[quote]Originally posted by Junkyard Dawg:
<strong>

Anyone got a better explanation for all that insane cooling capacity of the new towers?</strong><hr></blockquote>

UH, it looks good? :rolleyes:

Nah, I really hope you are right JD, it would definatly want to be a ncei change.

[Lemon Bon Bon]

"I think it's coming. Apple is going to come! One giant climax to change the face of computing. The Powermac G5 is going to eat Pentium 4s for breakfast, AMD Clawhammers for lunch....and for dinner, it's going to go hungry because there's nothing left for it to eat!! It's going to be the BEAST of all CPUs, the one that even Apple has trouble cooling."

I hope so... :D I've been waiting a few years now for Apple to actually DO something exciting with the guts of their machines. It's the one area where they are guilty of neglect.

Lemon Bon Bon

[Outsider]

[quote]Originally posted by Powerdoc:
<strong>
strange 14 stage pipeline in the power 4 , the same number discribed in some previous rumors about the G5. It appears that the people who create the rumors have copied these specifications on the power 4, or the people already knews that the G5 will be based upon the power 4 core.</strong><hr></blockquote>


I heard this from a friend at a bar while drinking some brews, but he said 95% of all rumors are made up!

[Bigc]

[quote]Originally posted by Junkyard Dawg:
<strong>The new tower design has some mondo cooling capability. That tower isn't designed to cool a G4....it's designed to cool a POWER 5! Seriously, what G4 puts out the sort of heat that would need a case with that sort of cooling capacity?

</strong><hr></blockquote>

Mondo Cooling? Why, because it has 4 holes drilled in the case? A purported 7 lb. heat sink, how much does the dual gig heat sink weigh?

[Lemon Bon Bon]

"how much does the dual gig heat sink weigh?"

Perhaps you can tell us?

Lemon Bon Bon

[Capt. Obvious]

[quote]from Bigc:
<strong>Mondo Cooling? Why, because it has 4 holes drilled in the case?</strong><hr></blockquote>
No - because the metal interior has been mega-perforated, an obvious first step if one wants extreme ventilation.

Or did you not notice the resemblance between the case interior and a sieve? Maximum airflow!

Whether we see the recently-leaked case or no, Dawg is absolutely right: it is AT LEAST an experiment in extreme - even radical - ventilation; my guess (worst-case) is that this is (the fruit of) an effort to apply Cube-style cooling tech to a non-Cube form factor.

[ 07-25-2002: Message edited by: Capt. Obvious ]</p>

[Programmer]

[quote]Originally posted by jerk:
<strong>
May I remind you that the altivec has 32 128b registers, not 1?

[Added:] The altivec can do operations on all its' registers at once, that is the point with it and what makes it differ from the ordinary register bank. It is a SIMD unit, "single instruction multiple data".

Or do you mean that the new cpus will be 32 times as fast (not calculating with the possible gain of doing something else while running the altivec unit)?

Saying that the altivec is a work around is just plain false. It is as saying as an floating point unit is a work around. They speed up certain kinds of programs by utilizing silicon for certain operations.
</strong><hr></blockquote>

Not to mention that the AltiVec unit has a whole bunch of operations which the integer unit simply doesn't have, and that would have to be emulated by a sequence of instructions per data element (not just one). Add with saturate, shifting 128 bit values, etc.

[Programmer]

[quote]Originally posted by Junkyard Dawg:
<strong>I think it's coming. Apple is going to come! One giant climax to change the face of computing. The Powermac G5 is going to eat Pentium 4s for breakfast, AMD Clawhammers for lunch....and for dinner, it's going to go hungry because there's nothing left for it to eat!! It's going to be the BEAST of all CPUs, the one that even Apple has trouble cooling.

Anyone got a better explanation for all that insane cooling capacity of the new towers?</strong><hr></blockquote>

I don't think its completely impossible, but I'm also not expecting it. We don't know that the prototype machine pictured will reach production in the near future. We don't know that it isn't a significantly overclocked advanced G4 with much higher heat output than previous G4s. We don't even know that its not just a prototype that has over-inflated cooling capacity just as a precaution because they hadn't yet worked out the real cooling requirements.

I wouldn't go to town expecting to come home with a G5 at any point this year... but I also wouldn't put any money down on the G5 not showing up this year. We just don't know for sure.

[OverToasty]

[quote]Originally posted by Capt. Obvious:
<strong>

Whether we see the recently-leaked case or no, Dawg is absolutely right: it is AT LEAST an experiment in extreme - even radical - ventilation; my guess (worst-case) is that this is (the fruit of) an effort to apply Cube-style cooling tech to a non-Cube form factor.

[ 07-25-2002: Message edited by: Capt. Obvious ]</strong><hr></blockquote>


... knowest ye not of the fan? It was indeed nobler by it's exclusion in previous cube's mortal coil.

But soft!

Whither the place in thy purest convection's philosophy of the super-sized mo-fo in this one?

Nay ... the case, in form and thought, is not "worst", nor holds it dreams of convection pure.

The chip is the thing, from which wind-blown forms spring.

<img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />

[ 07-25-2002: Message edited by: OverToasty ]</p>

[Ed M.]

A few things...


[[[The new processor could emulate altivec with 2 interger units...]]]

I'm not so sure.... Hop on over to distributed.net and check out the results of the Power4 vs the dual G4's. You might be surprised. Emulation is not an option IMHO. If Moto. and Apple spring a new 64-bit chip on us it will be able to run all 32-bit apps *natively*. And in case we need a refresher about why the industry is slowly moving to 64-bit, I cut a piece out of a previous post that talks precisely about such a move and pasted it here (works great for quick replies ;-)

I'm sure many of you remember it, but for those who haven't seen it....

[[[In many cases migrating to 64-bit could actually be *slower*. Why? It's fairly simple if you take the time to look at it and understand what's really going on. It isn't so obvious to people who have fallen for the "number's game" that Intel (with processors) and Micro$oft (with version numbers) have mastered in their Marketing campaigns. Here is where all the assumptions (by less informed people) started to arise:

During the transition from 16 to 32 bits applications and systems *did* gain speed. This was because many of the 16 bit machines often had to work with 32 bit numbers. A range from 0 to 65535 was simply not enough for common data. Therefore, in that case there is a *significant* difference between processing two 16 bit chunks or a single 32 bit chunk. However, here is the kicker (and something many people don't realize)... There are very few common applications that really need more than 32 bit wide data, or more than 32 bits of address space. Period. So in these cases, moving an app to 64-bit code would be a huge waste of time with little return and in many cases, it would even hurt performance. What this means is that there are *very few* kludges currently at work which process several chunks of 32 bit data. This, in turn, means that going to 64 bit will only speed up *those* very few kludges, while all pointer variables everywhere will consume double as much memory! In Apples case, having a single 64-bit CPU solution that can execute 32-bit code natively is a much smarter approach since it will give the developers of the *common* apps ample time to migrate their 32-bit apps to 64-bit. There isn't going to be a performance difference after migrating those apps to 64-bit anyway. So why the push toward 64-bit? We'll look at why Apple is moving to 64-bit. It probably isn't entirely for the reasons you might expect. More on that later...

As for the developers who cold use the extra horsepower. Guess what? The 64-bit CPU that is running 32-bit code natively is right there, just as happy to run 64-bit code. This means that developers will have vastly more flexibility when coding their apps. The key areas of the big "number-cruncher" apps can be moved to 64-bit, where 64-bit addressing would be an enormous boost. Here is how a trusted friend and PPC/AltiVec programmer explains it:

He expanded on this comment taken from a Darwin development board:

&lt;snip&gt; PPC uses a 16 bit offset from a register to determine the load/store address. The instructions don't change, nor do their operands when you go from 32 to 64 bit. Only the data in the registers differs, which would be controlled by whether you loaded a pointer using a load word or load double word. .... cont. &lt;end snip&gt;

Reply:

[[[Yes, the instruction format is identical. Yet there is a price for going to 64 bit: immediate values (constants embedded in machine instructions) are 16 bits wide. Filling a 32 bit register takes two instructions; filling a 64 bit register with an arbitrary bit pattern requires five instructions (synthesize two 32 bit values in four instructions, then combine them with a fifth).]]] - (Anonymous source)

By now you should understand why going from 32-bit to 64-bit doesn't automatically translate into more performance. Likewise, I'm sure you will see the added flexibility and advantages developers will have when coding for a a 64-bit processor that can run 32-bit code natively as well as true 64-bit code. And do keep in mind that OSX-64 will likely be able to run 32-bit apps natively as well. This is vastly different that Intel's approach in ITANIC which relies on emulation and M$'s approach with Win-64 doesn't seem to be any better. Anyway, 64 bit addressing won't help those few apps directly unless you've got gobs of RAM. Possibly 2Gig or greater. Now, The real reason for Apple moving to 64-bit...

Hmmm.. It sort of goes along this line of reasoning: In the very near future there will be desktops with more than 4Gig of RAM, so we will need to address that. That is to say going to 64 bits today is more of a forward-looking move. There are a few applications in existence that really benefit from a 64 bit machine, but those are server-type, "big iron" stuff. For example, The AltaVista search engine used to run its database on an Alpha machine with 16 gigs of RAM. This basically allowed it to service most of the requests without ever going to disk.

However, the sheer size of the problems and datasets being processed keeps increasing steadily. Some people claim that "typical" applications occupy 1.0 to 1.5 more address bits per year. If you just look at the 'default' amount of memory in simple, entry-level PCs, you can draw similar conclusions; Just do the math. Current off-the-shelf offerings range from 128MB to 512MB. This means at the high end, only three doublings are left before we reach the 4GB limit of 32 bit machines. Therefore, more addressing is needed. 64-bit will provide that.

On the desktop, video processing could end up being *the* application that drives the transition to 64 bit, not because it would make a huge difference in speed, but because it is much more convenient to handle files above 4GB on a 64 bit machine. ]]]

I Hope people have a slightly better understanding of what moving to 64-bit means.

Now, on the issue of the Power4... That chip is a 120-125 watt chip! The only hope is that IBM somehow develops a Power5 with significantly less power requirements. And the more chips you add the hotter things will become... So I really don't think we'll see a Power4 in a desktop Mac anytime soon. And since the Power5, 6, and 7 are a ways off..You can see what I mean.

The 7-pound heat-sink...

It is possible that Apple utilizes the ENTIRE CASE as such a heatsink. surely an empty husk of a case would be roughly 7 lb. if it were made from aluminum. The idea with a heatsink is surface area. Just like a car radiator and baseboard heating in homes. Actually I'd kinda like the idea of the entire case becoming the means by which the computer is cooled. If they can maximize airflow across all the surfaces of the case, it could work... The idea of a 7-pound block of aluminum sitting somewhere inside the case is preposterous! the volume of the heatsink alone would be enormous. Just something to think about.

--
Ed M.

[Rhumgod]

[quote]Originally posted by Programmer:
<strong>We just don't know for sure.</strong><hr></blockquote>

Thanks for enlightening us on that fact! ;)

However, I feel that with all of the high-end apps Apple has purchased recently, something BIG is in the mix. By big, I am thinking a new 64-bit chip. Jagwire uses GCC 3.1 and supports 64-bit executables, after all. Wouldn't surprise me in the least that Jobs announces Jagwire is 64-bit and, oh, by the way, so is the new PowerMac.

[Rhumgod]

[quote]Originally posted by Ed M.:
<strong>It is possible that Apple utilizes the ENTIRE CASE as such a heatsink.</strong><hr></blockquote>

What? Is the skin on my hand going to burn off when I touch the case!? ;)

I think moving to 64-bit is a necessity! True, some 32-bit apps will trip on their dicks and be slower, but who cares! Seriously. The rest of the world is going 64-bit soon and does Apple really want to be stuck with 32-bit G4+++++'s that can only achieve 1.x GHz? Hell no. I think Jobs finally has gotten the idea that the PowerMacs of the past were the FASTEST machines on Earth. Remember when they achieved 300MHz? They were the first to do it. And these machines screamed over Wintels. I think it's time to get back on top, or to quote JYD above, climax in the face, or something like that :p

[Programmer]

[quote]Originally posted by Rhumgod:
<strong>

Thanks for enlightening us on that fact! ;)

However, I feel that with all of the high-end apps Apple has purchased recently, something BIG is in the mix. By big, I am thinking a new 64-bit chip. Jagwire uses GCC 3.1 and supports 64-bit executables, after all. Wouldn't surprise me in the least that Jobs announces Jagwire is 64-bit and, oh, by the way, so is the new PowerMac.</strong><hr></blockquote>

I like to keep people aware that these forums are pure speculation -- it seems like they forget sometimes and get awfully caught up in designing awesome theoretical Macs and really low prices which are going to ship in a week. Best to stay grounded. :)

I agree that there is a 64-bit Mac on the horizon, but I'm doubtful it'll show in '02 or that Jaguar includes 64-bit support. They've been busying making Jaguar fast, and adding full 64-bit support at the same time would detract from that. Not impossible, just improbable. Sure hope I'm wrong though. ;)

[Bigc]

[quote]Originally posted by Lemon Bon Bon:
<strong>"how much does the dual gig heat sink weigh?"

Perhaps you can tell us?

Lemon Bon Bon</strong><hr></blockquote>

The weight of a heat sink is not what controls it's cooling efficiency, it's the surface area of the heat sink that controls its themodynamics. 7 lbs ain't gonna help

[Bigc]

[quote]Originally posted by Programmer:
<strong>

......We don't even know that its not just a prototype that has over-inflated cooling capacity just as a precaution because they hadn't yet worked out the real cooling requirements.....

</strong><hr></blockquote>

That's would be a reasonable assumption .

[tsukurite]

[quote]Originally posted by OverToasty:
<strong>


... knowest ye not of the fan? It was indeed nobler by it's exclusion in previous cube's mortal coil.

But soft!

Whither the place in thy purest convection's philosophy of the super-sized mo-fo in this one?

Nay ... the case, in form and thought, is not "worst", nor holds it dreams of convection pure.

The chip is the thing, from which wind-blown forms spring.

<img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />

[ 07-25-2002: Message edited by: OverToasty ]</strong><hr></blockquote>


holy smoke!
:eek: :D

[othello]

[quote]Originally posted by OverToasty:
<strong>


... knowest ye not of the fan? It was indeed nobler by it's exclusion in previous cube's mortal coil.

But soft!

Whither the place in thy purest convection's philosophy of the super-sized mo-fo in this one?

Nay ... the case, in form and thought, is not "worst", nor holds it dreams of convection pure.

The chip is the thing, from which wind-blown forms spring.

<img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />

[ 07-25-2002: Message edited by: OverToasty ]</strong><hr></blockquote>

i'll have whatever he's drinking. or should that be smoking...

:D <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />

[KidRed]

The case is more then likely real. 1.4ghz tops if Mot can deliver. But next year is when X transition will be complate and Appke will have "more options".

So, let the cowards jump ship this year because of speed issues, they'll be doggy paddling to come back next year because Apple won't be relying on just one mobo from one company. Hey, I like options too :)

[jerk]

[quote]Originally posted by Rhumgod:
<strong>
I think moving to 64-bit is a necessity! True, some 32-bit apps will trip on their dicks and be slower, but who cares! Seriously. The rest of the world is going 64-bit soon and does Apple really want to be stuck with 32-bit G4+++++'s that can only achieve 1.x GHz? :p </strong><hr></blockquote>

As has been noted here before:
1. 32 bit apps won't be slower on a 64 bit ppc - there is just a bit that tells the processor how to handle the very few and simple cases it has to handle differently (Think Different? :-).

2. There are just a very few cases where a program would gain speed by being remade into a 64 bit program using more 64 bit data types. Most programs would just be slower, a computer that has to shuffle 64 bits where only 32 are needed will waste more resources.

I think the main reason to go 64 bit is to be able to have larger than 4G virtual address space (please correct me if I am wrong).

[MaCommentary]

If apple is going to "continue to innovate and invest..." (Jobs) then what's to say that Apple won't buy MOT's semiconductor linceses (they have an option at the end of 2002) and put a vector engine on a stripped-down POWER 4/5? AND then IBM can make all speeds in the G4 line, making them a complete solution for Apple.

[bunge]

My guess is IBM is much closer to having a Power4 replacement than some people think. As someone here said previously, they built the Gekko G3 variant for Nintendo. The Power4 replacement could easily be an Apple only chip like the Gekko. The production on the chip wouldn't follow IBM's Power 'roadmap' or timetables.

[Rhumgod]

[quote]Originally posted by jerk:
<strong>There are just a very few cases where a program would gain speed by being remade into a 64 bit program using more 64 bit data types. Most programs would just be slower, a computer that has to shuffle 64 bits where only 32 are needed will waste more resources.

(please correct me if I am wrong).</strong><hr></blockquote>

I an including a link to a great article called, <a href="http://www.ddj.com/documents/s=1038/ddj9301c/9301c.htm" target="_blank">Cray: 64-bit programming in a 32-bit world</a> Not too tough nor detrimental, speed-wise, if you just don't use the machine word size in code. Quoting the first paragraph, "Compared to 16-bit programming, 32 bits means faster programs, more memory with straightforward addressing, and better processor architecture."

Same can be said for 64-bit programming.

[Programmer]

[quote]Originally posted by Rhumgod:
<strong>I an including a link to a great article called, <a href="http://www.ddj.com/documents/s=1038/ddj9301c/9301c.htm" target="_blank">Cray: 64-bit programming in a 32-bit world</a> Not too tough nor detrimental, speed-wise, if you just don't use the machine word size in code. Quoting the first paragraph, "Compared to 16-bit programming, 32 bits means faster programs, more memory with straightforward addressing, and better processor architecture."

Same can be said for 64-bit programming.</strong><hr></blockquote>

Except for memory addresses, which are heavily used in modern C++ code. All of these must be 64-bits if the process is has a 64-bit address space.

[KidRed]

[quote]Originally posted by bunge:
<strong>My guess is IBM is much closer to having a Power4 replacement than some people think. As someone here said previously, they built the Gekko G3 variant for Nintendo. The Power4 replacement could easily be an Apple only chip like the Gekko. The production on the chip wouldn't follow IBM's Power 'roadmap' or timetables.</strong><hr></blockquote>

I'd look for a powr4-core mobo from IBM for G4s this time next year. The days are numbered with moto and anything with AMD would be last resort.

[MozillaMan]

Jumping from 32bit to 64bit does NOT mean a doubling of speed...

16 to 32 effectively did, though. With 16 bits, integers could only be in a range of –32768 to 32767, but with 32 bits, the range is -2,147,483,648 to 2,147,483,647.

There were plenty of times where 16 bit integers were too small for calculations, but the same can't be said of 32 bit integers, where the only applications that would really benefit from 64 bit integers would be crypto, and some niche engineering programs. If anything, I read from a study that compiling 64-bit code actually on average takes a 5% performance hit. And it makes sense, when you think about it, since 64-bit numbers take up much more space than their 32-bit counterparts.

The only real-world advantage I can see for mainstream 64-bit platforms is the addressing of memory greater than 4 GB. Anyone know how much memory 64-bit registers can accomodate? :D

[Programmer]

[quote]Originally posted by MozillaMan:
<strong>
The only real-world advantage I can see for mainstream 64-bit platforms is the addressing of memory greater than 4 GB. Anyone know how much memory 64-bit registers can accomodate? :D </strong><hr></blockquote>

16 billion gigabytes.

64-bit address spaces are useful for some sparse data structures -- i.e. its a huge structure but a very large amount of the space is never touched an thus doesn't actually need any memory pages at those locations. Pretty rare technique though.

[MozillaMan]

[quote]Originally posted by Programmer:
<strong>

16 billion gigabytes.

</strong><hr></blockquote>

Mmmm... Looks like I'd have to solder on a lot more RAM slots... :D

[hotboxd]

Well theres been plenty of time for Apple to realize that Mot isn't going anywhere, or at best is unreliable. They've had two years in which to sit down with IBM and design a custom next generation chip for Macs. That's plenty of time to create a less expensive, smaller, and cooler chip based on the Power4 architecture.

[jerk]

[quote]Originally posted by Rhumgod:
<strong>

I an including a link to a great article called, <a href="http://www.ddj.com/documents/s=1038/ddj9301c/9301c.htm" target="_blank">Cray: 64-bit programming in a 32-bit world</a> Not too tough nor detrimental, speed-wise, if you just don't use the machine word size in code. Quoting the first paragraph, "Compared to 16-bit programming, 32 bits means faster programs, more memory with straightforward addressing, and better processor architecture."

Same can be said for 64-bit programming.</strong><hr></blockquote>

[ Note: I had to edit this, I misread your comment! I guess we in most parts agree. Sorry! I still let most of the text remain and hope it will give someone something, now that I have written it. ]

No, it can't. Actually, the "faster" statement in most cases is false, it just happens to be true for the PCs' old architecture and heritage and ugly solutions to its shortcomings some 15 years ago, and has therefore been taken as a truth by some. It is sad to see a cray researcher lend himself to such cheap a trick to earn some kind of points for cray hardware, what kind I don't know. I am bent to think that it isn't the researcher himself that wrote that part, since the rest of the article proves that he understands the differences of different register sizes.

The rest of the article is quite good though.

I'll try to keep this short:
Different processors have different shortcomings in handling different sizes of data, which the article also mentions. The only thing the article says regarding speed is that it is typically faster, and on some processors only possible with compiler tricks which means slowdown, to address data on its natural alignment. Nothing else. This is true for the ppc too. The ppc can handle 8, 16, 32 and soon probably 64 bit data very well with no special treatment or difference in speed, but reading a 64 bit word that not lies on its natural alignment would require two bus cycles on a 64 bit bus instead of one. It can actually handle larger data sizes too, depending on what you want to do with what of its units. The FPU has been using 64 bit data and the addition altivec 128 bit data as their basic types all along.

The size of the registers matters for the size of the data you want to handle is, that is all. If you want to handle data that is larger than the register size, you have to emulate larger registers with more instructions and registers, which is bad for speed. If we want to handle 64 bit address spaces, which we want more and more often, we want 64 bit registers.

Many programs of today are fine with 32 bit integer/fixed point registers and 64 bit floating point registers as they are in the ppc today, and are just using those datatypes, even when only 5 or 14 or whatever are really required. Shuffling those extra never used bits just takes bandwith from the bus. This is only if the programmer is sloppy, but they (we) most of the time are, and analyzing how many bits would really be needed would require a lot of work and be error prone (with the tools of today at least).

When switching to 64 bit architecture, many programs will use 64 bit data where 32 bits are used today, shuffling even more bits that will never be used.

This is probably not a very big deal, but that is why I say that there for many programs is no speed gain in going 64 bit, it can even slow things down.

The PPC is a derivate of the POWER family, which has long before the PPC and maybe from the very beginning been designed to be able to have both 32 and 64 bit implementations. The 32 bit implementations are just lacking half of the register bits and some instructions and instruction modes for handling those extra bits and a little more glue to make it emulate a 32 bit implementaion (maybe a chip designer would argue on how small the differences are :-). The 64 bit implementations can easily be switched to and from 32 bit mode, the logical differences are really small. On a 64 bit ppc, processes can be 32 or 64 bit running side by side with no performance hit (except for the instruction that has to be run to switch mode).

Many programs will probably go faster if they are just run in 32 bit mode, and we will probably see a lot of those for a long while.

If one would like to do comparisons (which often is dangerous, so I probably shouldn't), one could compare it to shipping a certain load with a larger lorry. It won't go faster just because the lorry is larger, it will only take up more space on the road - consuming more bandwith for transporting things. A larger load COULD fit, though, without the work of splitting it up and reassembling it, but if that isn't used, the it will just be wasted.

I hope it is now clear why a 64 bit ppc will not magically buy you speed, it won't more than in a few special cases and applications.

As mentioned here before, one good thing with 64 bit addressing is that you could have sparse mode addressing which can do some nice tricks, but not necessarily gain any speed.

[ 07-26-2002: Message edited by: jerk ]</p>

[Programmer]

[quote]Originally posted by jerk:
<strong>On a 64 bit ppc, processes can be 32 or 64 bit running side by side with no performance hit (except for the instruction that has to be run to switch mode).</strong><hr></blockquote>

Great post! One nitpick: the mode switch to/from 32/64 bit mode will probably be on a per-process basis and the mode bit will be carried for free in the context switch -- i.e. zero overhead.

32-bit applications will likely remain the norm for a long time since there will be 32-bit PPC chips around for a long time and developers will normally write 32-bit code to ensure they can run on all the PPC machines. Only in particular situations will developers target just the 64-bit machines. And as alluded to above, even when targetting a 64-bit machine it might be faster to compile as 32-bit to get better performance.

[rambo47]

Apple has said it's going with the G4 for now. So if an IBM-Power4/5 solution is to be, it will be several years before we see it. Years. Does Apple have the sales numbers to convince IBM to make a custom chip for PowerMacs? Plan on seeing Motorola chips in your Macs for the foreseeable future.

[taboo]

[quote]Originally posted by rambo47:
<strong>Apple has said it's going with the G4 for now. So if an IBM-Power4/5 solution is to be, it will be several years before we see it. Years. Does Apple have the sales numbers to convince IBM to make a custom chip for PowerMacs? Plan on seeing Motorola chips in your Macs for the foreseeable future.</strong><hr></blockquote>

Yes and no. Apple (and Moto) has said that the G4 "has plenty of life left in it". That doesn't neccesarily mean that the G4 will be the only proccessor used. They've had a "low-end", "high-end" solution pretty consistently for quite a while now....
Although, I do agree that we may very well have a modified G4 'til the desktop Power5 comes available. Early in 2004 I seem to recall reading... :)

Whoops. Just remembered that statement was made about the G3. That's what I get for reading/posting while drinking my first coffee. <img src="graemlins/oyvey.gif" border="0" alt="[No]" />
The 2 proccessor comment still stands, tho'....

[ 07-28-2002: Message edited by: taboo ]</p>

[Programmer]

[quote]Originally posted by taboo:
<strong>Whoops. Just remembered that statement was made about the G3. That's what I get for reading/posting while drinking my first coffee. </strong><hr></blockquote>

Actually Motorola said essentially that about the G4, and Apple said it about the G3 (for the iBook). I think IBM has an VelocityEngine-equipped RapidIO G3 on the way that would suit the iBook just fine (look at their roadmap). Just because everybody (including me) thinks the POWER4 will form the basis for the next high-end PowerPC chip, doesn't mean they'll suddenly drop all of their existing low-power stuff.

[Rhumgod]

An excellent article on IBM, Altivec, ppc64 and various 32-bit and 64-bit programming notes:

<a href="http://www-106.ibm.com/developerworks/library/l-ppc/" target="_blank">IBM PowerPC assembly</a>

[blabla]

[quote]Originally posted by Programmer:
[QB]

I think IBM has an VelocityEngine-equipped RapidIO G3 on the way that would suit the iBook just fine (look at their roadmap). [QB]<hr></blockquote>

That would be a G4... G3 is not a single chip-design, but a chip family, so is the G4.

[Programmer]

[quote]Originally posted by blabla:
<strong>
That would be a G4... G3 is not a single chip-design, but a chip family, so is the G4.</strong><hr></blockquote>

Okay, substitute "Sahara" for G3 then. IBM might number it in the 7xx series instead of the 7xxx series, which only Motorola has used.

[Programmer]

[quote]Originally posted by Rhumgod:
<strong>An excellent article on IBM, Altivec, ppc64 and various 32-bit and 64-bit programming notes:

<a href="http://www-106.ibm.com/developerworks/library/l-ppc/" target="_blank">IBM PowerPC assembly</a></strong><hr></blockquote>

Interesting to note that this paper, written by an IBM guy and on an IBM site, talks about AltiVec programming and refers to AltiVec articles. It also mentions the term "VMX".

Hmmm....

[rickag]

[quote]Originally posted by Programmer:
<strong>

Interesting to note that this paper, written by an IBM guy and on an IBM site, talks about AltiVec programming and refers to AltiVec articles. It also mentions the term "VMX".

Hmmm....</strong><hr></blockquote>

More specifically, the article says,"Altivec (also called VMX)"....interesting

[Moogs]

I think we're gonna get 1167MHz G4's from Motor-scro-la and everything from there on out will be from IBM, that's what I think (or what I hope, more specifically). Good riddance.

It's great to see IBM is working with Altivec on some level, has a new high-tech fab with a process that's more advanced than the one MOS13 uses (the new IBM fab operates at .09µ correct?), Apple is talking about "options", etc. The signs seem positive for a change....


Buh bye, Motorola!

[ 08-06-2002: Message edited by: Moogs ]</p>