That's what I'm hoping, at least until late 07 or early 08. It makes sense too, with all the rumors about the CHUD tools and quad processors (dual dual-core). Hopefully Apple does have a dual dual-core PowerMac G5 in the pipeline from IBM and plans to release it early next year.wdlove said:
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
970MP - Still Coming?
- Thread starter MacRumors
- Start date
- Sort by reaction score
No new PPC Mac
Steve Jobs has declared the PPC for the Mac as being dead. To me that means no really new PPC Mac will come out. Just some models until the Intel chips & Intel models are ready. Do you expect Apple to produce a new PPC board for the PowerMac? They didn't make any changes when they brought out the last couple of G5 PowerMacs, so why would they change the motherboard after the model has been declared DBA (Dead Before Arrival.)
Bill the TaxMan
Flynnstone said:
Steve Jobs has declared the PPC for the Mac as being dead. To me that means no really new PPC Mac will come out. Just some models until the Intel chips & Intel models are ready. Do you expect Apple to produce a new PPC board for the PowerMac? They didn't make any changes when they brought out the last couple of G5 PowerMacs, so why would they change the motherboard after the model has been declared DBA (Dead Before Arrival.)
Bill the TaxMan
heisetax said:
Steve also said that there is some fantastic PPC products coming out. So...
Steve is a smart guy. He has positioned OS X to be cross platform. He is saying to IBM & Freescale, "If you don't come up with some good, we are going with Intel!" With the demonstration, he showed that he was serious.
So really Apple can't lose.
Apple has ~5% market share. This can be debated to the nth degree. But PC sales are flat, Apple is on the rise and has a lot of excitement with it too. Could be the next big growth area. They want a piece of that pie.
Flynnstone said:
*shakes head* You guys still haven't learned yet have you? Jobs is about show. About glitz. Year of the laptop! Year of HD! Year of this! Year of that! You take the man's words as literal and you will be let down.
Is Apple and IBM going to do some speedbumping over the next year and a half? Doubtless. I fully expect Apple to do what they have been doing on the PowerBook over the last 18 or so month. Upgrade the video card. Maybe PCI-Express. Drop prices. Upgrade the hard drives. Add some funky new thing to the PowerMac. Anything to show forward momentum but at the end of the day I really doubt there is going to be a dual core release. You can bet your year's salary that Apple's #1 priority right now is the PowerBook. #2 is the Mac Mini and #3 is the PowerMac and iMac. Add to the fact that I'm willing to be that there simply is a learning curve for most of Apple's engineers. Do they know X86? Yah. I'm pretty sure they do but they have been designing for the PPC. The transition isn't going to be overnight. Add to the fact that they have to figure out what exactly they are going to replace Open Firmware with. That alone is NOT going to be a trivial tasks. That alone needs be DONE before they even get to the first Macintel* out the door be it a PowerBook or an Mac Mini
What is the end result of all of this? Priorities. One thing I've learned about Jobs over the last few year of being around the Apple community and reading about him is that one he flips a switch and the light is green he nor Apple looks back. Unless the DC PPC is virtually completed it simply isn't going to happen. If its mostly completed. Maybe. If its all but done with the chipset and motherboard completed and Apple is simply waiting on IBM at this point then I'd say it will be here. It all depends on where both Apple and IBM are in the process.
*Sorry guys. I just like the word. Macintel almost sounds white. Like an ibook or an iMac. I don't know. It sounds soft and aesthetic.
Flynnstone said:
No dude. They ARE going to Intel. You don't make your entire keynote about moving to x86, roll out tools at WWDC to convert your apps, stress to ALL your developers that you need to start working on this now, bring up the head of you 2nd party software developers on stage, bring up the head of the manufactuer you are going to onstage and give him 10 minutes of talk time, and roll out development systems just to see if your hardware vendors blink first.
As hard as this is to admit people need to get this through their head:
THIS ISNT A JOKE. THIS ISNT A CON TO GET BETTER HARDWARE FROM IBM OR MOTO. THIS IS HAPPENING.
Sorry but it HAS to be said even if it has to be done another 100 times.
Steve did say that there are new PPC products in the pipeline. We'll know soon enough what that means exactly as several products (iBook, Mini) are approaching the end of their update cycle. Yes, the move to Intel will be a total and permanent move. But it will take 2.5 years. So Apple needs to release some new products attractive enough to support their bottom line. I imagine the iBooks and PowerBooks will get at least one update before they're moved to Intels. Same with the eMacs and Mac Minis. The Power Macs and iMacs are more difficult to figure. I'd imagine that there are at least a couple of revisions in the works. The 970fx is tapped out, so we'll most likely see the 970gx or 970mp in the next six months or so. The 970mp should give a pretty good speed boost and entice enough on-the-fence buyers to help the bottom line. My guess for the updates:SiliconAddict said:
Summer:
iBooks: up to 1.50 ghz using 7447
Minis: up to 1.67 ghz using 7447
Fall:
PBooks: up to 1.80 ghz using 7448
Power Macs: up to 3.0 ghz (finally!) using 970gx/970mp
MWSF 2006:
Mac Minis move first to Intel
Spring:
iMacs up to 2.2 ghz using 970gx
WWDC 2006:
iBooks move to Intel
Fall:
PBooks move to Intel
eMacs move to Intel
Last update for G5 PMacs (up to 3.3 ghz)
MWSF 2007:
iMacs move to Intel
WWDC 2007:
PMacs move to Intel
Fall:
Xserves move to Intel; the transition is complete
dongmin said:
Actually I fully expect a Pentium M PowerBook in the first half of 2006 if not at MacWorld. Considering they are demoing duals NOW: Intel Demos dual core Pentium M laptop or more accurately two months ago.....
You can bet that Apple probably has their hands on those puppies NOW. I can easily see Apple springing a PowerBook update on us at MW 2006. Why so fast? Because of the fact that laptops are now outselling desktops. Apple knows this is HIGH on their priorities. The PowerMac is getting a little long in the tooth but its still got fangs. The PowerBook? Can you say wimptastic? The only "issue" I can imagine is dealing with Firmware on the x86. This could take a fair amount of time. I could easily see that holding up the process.
SiliconAddict said:
I believe you are right.
Apple is committed to the Powerbook being updated first.
It is by far the most out of date product in their whole line.
And it is their biggest seller.
wish you were there.
Hahaha.. funny read guys.
How about this. let me give my perspective.. I was there at WWDC and all.
* Jobs said they were transitioning from PPC to X86, I don't recall him saying that they were REPLACING all PPC with x86 though I could be mistaken.
* Jobs did not Say that PPC was dead to Apple.
* Jobs said there are exciting new PPC products in the pipe.
* Apple's Press release said All Macs will transition to Intel in 2 years time.
* Apple did not say that they would be exclusively Intel. Hmn.
* Apple did NOT mention the xServe.
* The xServe is NOT A MACINTOSH.
* Apple has taken great pains to avoid mentioning AMD in any way. why?
* I've seen Apple's plans for x86/PPC support on Macintosh. There is no reason why x86/PPC can't co-exist. It WILL co-exist for years.
* Apple is STILL SELLING AMP for PPC systems!!
OK, more on that last point.
AMP is Apple's software licensing program to keep your OS up to date. Pay a very reasonable fee and you get free OS updates for 3 years.
If you buy AMP on a PPC mac today, that guarantees you OS updates that are PPC native through the next two OS updates (10.5 and 10.6).
I cornered an Apple rep and questioned them about support because it's a big enterprise issue. Apple's policy is full support for the previous OS. 10.3 is fully supported. Apple will update packages. Apple will fix bugs. Apple will patch critical security holes.
Apple provides Critical support for OSes that are two revisions old. That means OS 10.2 still gets critical patches.
By this, Apple is already planning full OS Support for PPC for 3 years. Add in support for the previous OS and that means they'll keep supporting 10.6 on PPC 4.5 years out. After that, critical support on 10.6 on PPC 6 years out. Apple won't obsolete PPC for 6 years at the earliest.
Is it shocking to think that Apple is keeping their options open?
chew on all that and let the arguing begin anew.
Hahaha.. funny read guys.
How about this. let me give my perspective.. I was there at WWDC and all.
* Jobs said they were transitioning from PPC to X86, I don't recall him saying that they were REPLACING all PPC with x86 though I could be mistaken.
* Jobs did not Say that PPC was dead to Apple.
* Jobs said there are exciting new PPC products in the pipe.
* Apple's Press release said All Macs will transition to Intel in 2 years time.
* Apple did not say that they would be exclusively Intel. Hmn.
* Apple did NOT mention the xServe.
* The xServe is NOT A MACINTOSH.
* Apple has taken great pains to avoid mentioning AMD in any way. why?
* I've seen Apple's plans for x86/PPC support on Macintosh. There is no reason why x86/PPC can't co-exist. It WILL co-exist for years.
* Apple is STILL SELLING AMP for PPC systems!!
OK, more on that last point.
AMP is Apple's software licensing program to keep your OS up to date. Pay a very reasonable fee and you get free OS updates for 3 years.
If you buy AMP on a PPC mac today, that guarantees you OS updates that are PPC native through the next two OS updates (10.5 and 10.6).
I cornered an Apple rep and questioned them about support because it's a big enterprise issue. Apple's policy is full support for the previous OS. 10.3 is fully supported. Apple will update packages. Apple will fix bugs. Apple will patch critical security holes.
Apple provides Critical support for OSes that are two revisions old. That means OS 10.2 still gets critical patches.
By this, Apple is already planning full OS Support for PPC for 3 years. Add in support for the previous OS and that means they'll keep supporting 10.6 on PPC 4.5 years out. After that, critical support on 10.6 on PPC 6 years out. Apple won't obsolete PPC for 6 years at the earliest.
Is it shocking to think that Apple is keeping their options open?
chew on all that and let the arguing begin anew.
dongmin said:
I think the other way around.
The Intels will be used for the "Big Iron" first. The mini will be last.
So I think Power Books first then Power Macs.
The professional series will get the performance first.
The PowerPC will fall to consumer (second tier) as an extra snub to IBM.
ffakr said:Hahaha.. funny read guys.
[snip]
After that, critical support on 10.6 on PPC 6 years out. Apple won't obsolete PPC for 6 years at the earliest.
Is it shocking to think that Apple is keeping their options open?
chew on all that and let the arguing begin anew.
Perhaps with dual core, the need for Altivec is removed. Move away from specialized hardware (Altivec/SSE), but still support in a more general way.
Keeping options open
Then Apple could use Intel or PowerPC or ... MIPS, Itanium, Alpha (dreaming) ARM, Pic .. oh scratch the last one.Flynnstone said:
Nope, this one I'll put money on. "Big Iron" won't go first.
Apple isn't moving to Intel for performance. It's moving for Performance/Wattage. I defy you to find where Jobs says they are moving to Intel for pure performance in the keynote.
The Mini is actually the most logical choice for the first transition for a number of reasons. Let me toss a few out:
* The mini is the perfect showcase for WHY Apple is doing this. It's a small, sleek design that will be a very computationally powerful computer with a dual core Yonah [Pentium-M] in it. It won't be a top of the line Yonah, it may not even be dual core but I wouldn't be surprised (on the .65 nm process it will be small and cheap, even as a dual).
* The Mini is the perfect development box. It's cheap. It will be plenty powerful.
* The mini is the switcher box (my mom wants me to buy her one to replace her POS athlon that keeps getting spyware even with AdAware installed.. even though I told her how much cheaper I could get her a Dell tower on sale). This is the box that will draw switchers. It's small. It's cheap. It will boot OS X or Windows XP when it goes x86.
The PowerPC won't stand still over the next year. I do expect a dual core PPC 970. We could easily have 3+ GHz dual dual-core PPC towers by next year. I don't see Apple abandoning that for a single dual Core 32bit Yonah system. i don't see them abandoning it even for a dual dual-core EMT-64 Xeon because the Xeon systems aren't all that cheap (often more than the PPC towers from Apple) and the Xeon will still be NetBurst P4 by then.. the PPC will likely be significantly faster.
The other consumer boxes and the Powerbooks will likely go after Minis. The iMac is a good candidate because it will become the enthusiast swithcher rig. It's all in one for LAN parties. It will be a fast dual core box (P-Ms are really fast with their bus slewing turned off). With PCI-E in the chipset Apple can put very nicely performing, very reasonable PCI-E native video chipsets in them. And they'll dual boot XP.
Imagine a current 17" or 20" iMac with a dual-Core P-M at 2GHz. Imagine it has a video chipset with 128 or better RAM and something that would (today) fit into the same product slot as an X600 or maybe an X700 from Nvidia. That's be a SWEET little machine. Perfectly capable of playing all games. Not a killer Alienware rig, but it would cost less. Right now a single P-M at 2GHz and an X600 or X700 would play Doom3 really nice and Doom3 is the ultimate Stress Test right now.
The Powerbooks are also obvious choices since that line has been languishing for a while. Don't get me wrong, I love my 17" PB but there are times I had more horsepower.. Usually no.. but occasionally I'll muck with Video and a dual P-M would be nice to have.
ffakr.
the rumors have said the low-end (i.e. consumer) macs will go intel first. several reasons i believe these rumors:Flynnstone said:
1. pro software won't be ready until later. the iapps are already running on x86, but apps like FCP and DVDSP will take longer to port and optimize for the x86 given their dependence on the g4/g5 architecture and altivec. that goes for 3rd party apps (photoshop, cinema3d) as well. emulation is simply not gonna get the job done for the pro apps, so until the universal binaries are released, 'pro' hardware will be useless.
2. the g5 is competitive with the fastest of the x86 workstations, as recently illustrated by tomshardware benchmarks. i think pros, who depend on fast, proven hardware will continue to buy the g5 powermacs. at least until the x86 desktops are well-tested and all the pro apps are proven to run well on these new boxes
3. as ffakr noted, where apple will benefit most with the switch to x86 is in portables and small-form-factor cpus, i.e. the mac mini. apple will most likely skip the p4/netburst architecture completely. desktop-class derivatives of the P-M won't arrive for a while. 64-bit versions will take even longer.
I disagree.
We can agree to disagree (These comments at best are speculation anyways )
Any program that uses the Altivec will require more work, granted.
But remember Apple has been running OS X for 5 years on Intel hardware behind closed doors. It would be reasonable to assume that they have programming all their apps to be cross platform. Photoshop is already on Intel, so port should be too bad. They likely have optimizations for SSE already.
The G5 is on par with the P4, at present. Next year is a long way off. IBM has their hands full with ... I'll call the game console silicon. Perhaps the G5 won't see 3 GHz. The P4 seems to have reached close to its limit. The G5 won't go much faster at the 90nm feature size. I don't see the 970 at its present form going to 65nm. A 970MP at 65nm ... yes. I haven't heard any real numbers for clock rates at 65nm. But I think there be dragons still. I know soft logic problems are more of a problem at 65nm.
I still think a dual core, hyperthreading Intel low power 32/64 bit Pentium M derivative will be put in the PowerBook first. Perhaps a single core for the iBook. Laptops are the things that are selling well. And that's where Apple's biggest problem area is.
We can agree to disagree
(These comments at best are speculation anyways ) Any program that uses the Altivec will require more work, granted.
But remember Apple has been running OS X for 5 years on Intel hardware behind closed doors. It would be reasonable to assume that they have programming all their apps to be cross platform. Photoshop is already on Intel, so port should be too bad. They likely have optimizations for SSE already.
The G5 is on par with the P4, at present. Next year is a long way off. IBM has their hands full with ... I'll call the game console silicon. Perhaps the G5 won't see 3 GHz. The P4 seems to have reached close to its limit. The G5 won't go much faster at the 90nm feature size. I don't see the 970 at its present form going to 65nm. A 970MP at 65nm ... yes. I haven't heard any real numbers for clock rates at 65nm. But I think there be dragons still. I know soft logic problems are more of a problem at 65nm.
I still think a dual core, hyperthreading Intel low power 32/64 bit Pentium M derivative will be put in the PowerBook first. Perhaps a single core for the iBook. Laptops are the things that are selling well. And that's where Apple's biggest problem area is.
hell yea, I said Apple wouldn't go Intel.. so there you goFlynnstone said:I disagree.
We can agree to disagree
That is a false statement. A program that leverages Altivec by hand coding Altivec code will require work. It may require reworking entire algorithms. Apps that access Altivec through Apple's accellerated Math frameworks, VecLib, they will compile straight to SSE#/MMX.
This is mostly true. The iLife apps were reportedly fully x86 native and optimised for Intel. Remember though, although Apple has built OS X on Intel for 5 years, they didn't put the same effort into the contingency plan as they did into the shipping OS. OS X x86 is still a Beta OS.
Adobe does actually use a common core codebase for apps like Photoshop. At least they did a few years ago. They optimize and slap a new interface on the Windows/Mac releases.
No.
The P4 is at the end of it's life. Perhaps you didn't mean to say that. Tejas is cancelled and Intel has announced that the P4 will never hit 4GHz. The current dual core P4s are as fast as the P4 will ever get, period.
The PowerPC will get better. Steve Jobs himself said they had exciting PPC updates in the pipe. IBM has developer documentation for the dual core PPC 970. It exists. Apple would be a fool to not ship it. IBM does continually refine their litho process. The PowerPCs, the Dual Core included will continue to speed bin faster with no other work other than what they continually do in improving their chip litho yield and performance.
The G5 is currently at 2.7. The BIGGEST problem with the current G5 (PPC970) is that it is too small. It's difficult to disipate heat quickly enough from the cpu surface even though it doesn't put out much compared to something bigger like a Prescott P4. Dual Core G5s, with NO additional modifications, will be more than 2x as large and will disipate heat better. The next gen PPC 970 will be produced on a revised process though. It should have little trouble hitting 3GHz, I suspect not that far off either.
You are right about the P4 as I stated above. Intel says it will never go any faster.
As for 65nm. Intel has also been quite open about how their 65nm transition is going much better than they expected. They aren't seeing the electron hopping/current leakage. Considering IBM had WAY less problems with 90nm than Intel, hopefully IBM's 65nm will go even better.
Um. Nope. The P-M is only 32 bit right now (Dothan). The NEXT P-M will only be 32-bit (Yonah). I don't believe Intel has an EMT-64 version of the P-M due till the second half of 2005 at the earliest. Most people seem to agree with me that Yonah will be the first Intel processor used in production Macs. I don't disagree that Apple will want to get the mobile line bumped asap, I just see equally strong reasons to bump other lines. The bottom line will likely be to get as many MacTel boxes out as quickly as posible. Large numbers of x86 Macs will provide additional pressure to developers to fully support the new platform (though I don't see that as a real problem with Apple's transition plan). If Apple thinks that Minis and iMacs will be better for development.. and will sell faster, than they will get more attention in the transition. If Apple thinks development and uptake will be faster with Powerbooks, that will add weight to getting powerbooks out first.
Consider one more thing though. The Powerbook is a professional product line. I've got people I support who rely on their powerbooks for presentations to NASA. It's probably not the best line to launch experiments with. Now, if you release Minis, you probably get a little more leeway on the occasional hiccup.
Just thinking here...
Guys I think you need to brush up on what the def of transition means.....
![Transition.jpg]()
There is no grey area here. None at all............
![Jobs.jpg]()
There is no grey area here. None at all............
kainjow said:
Naw, Steve's into vaporware. First he buys into Motorola's pipedreams, then IBM's and now Intel's. The first two companies didn't come close to producing what they promised so don't expect Intel to either. In fact, Freescale and IBM have the greatest potential to achieve their promises as they are close now. Intel is not even in the ballpark. Their 64 bit chip is basically two 32 bit chips welded together. Freescale's dual core G4 does a much better job of processing data. But, Steve is made at the other guys and he's bought into the megahertz myth apparently. Goodbye thin elegant Powerbooks, hello hot as hell Intel powered bricks.
You might have no idea what you're talking about.stockscalper said:
You just might. Let's check that out.
Apple HELPED DESIGN THE POWERPC. Linley Gwennap, currently of The Microprocessor Forum, worked for Apple on the PPC development team. This is why the PowerPC alliance was called AIM [Apple, IBM, Motorola].
Now, on to the other reason why that statement is just plain silly. IBM and Motorola DID, in fact, offer more powerful cpus than Intel and AMD for a long time. The 601 and 603 were very compareable to the Pentiums of the day, of course the 603 scaled much higher in Clock where the P1 could never go. The 604, partcularly the 604e revision compared very favorably to the PentiumPro.. oh yea, except that the PPro never got released over.. what was that.. 200Mhz? The 604e continued to scale to.. what was that again? 400MHz in Macs and 450MHz in IBM hardware? Then of course it all changed with the G3.. Intel and AMD blew past AIM.. wait, no it didn't. It wasn't until the G4 that Intel and AMD surpassed IBM/Moto in overall performance, and not even that was the case until the 2nd quarter of 2000. Remember, the G4 and Athlon actually shipped at relatively close Clock.. and their performance was similar (an edge to Athlon while Altivec performance was Much better than 3DNow and the implementation was more broad Allowing the G4 to remain competative in Apple's core markets for longer than it would have based on core CPU performance). And of course, the G5 leveled the playing field again.
BTW.. let's not get in a tit for tat benchmark game. I guarantee you can find Athlon64 and P4/PM benchs that beat the G5 and I guarantee I can do the reverse.
Perhaps you need to spend more time with Chris Tom at AMD zone.
There's a lot here so I'll tick them off..
* so, the 64bit chip is two 32bit cores. Even if that's true in THIS release, how does that affect the roadmap 2 to 7 years down which is where Apple is betting the farm? Also, how is that worse than a single core? You can't run dual P4s, only Xeon in SMP. This is the only way, right now, to get a real SMP Intel box without buying a workstation/server chip. Oh yea, where's that IBM 970MP? You're berating Intel for beating IBM AND AMD out of the gate with a consumer Dual Core because the first revision wasn't good enough for you? Sure AMD is the real desktop Dual core leader now that the X2 is finally out.. but they cost up to $1000 retail just for the CPUs.
* FREESCALE'S DUAL-CORE G4 IS VAPORWARE. Do I even need to expound? Where is this chip? What bus will it run on?
* Apparently AMD and Intel have both bought into the MHz myth since the Athlon64 and the P-M both spank MUCH faster P4 processors. Could this be why AMD insists on Performance Ratings and codes NOT MHz to index their processors? In fact, the P-M 2.1GHz Kept up with the FX55 and P4 3.8 in most real world benchmarks and beat both processors when it was overclocked to only 2.5 GHz. Check out Tom's Hardware for the very thorough review.
* Hmn.. Hot as Hell Intel bricks? The only reason powerbooks run the way they do is because we're using a chip that we should have been using 2 years ago. Where's your G5 powerbook? Oh wait, the MAX thermal output of a P-M is equal to the typical thermal output of a lower clocked G5 right now. The typical thermal output of a P-M is 1/3 that. I'm a PPC and a Mac fan, no doubt of that, but the P-M is currently the best performing mobile processor Period. It looks to get much better too. Yonah will be dual core on 65nm this year.. and it will be in the same thermal envelope. Today's P-M Dell which would thoroughly spank the beautify Powerbook I'm typing on, will be twice as powerful by Q1 next year with no additional heat.
Also, don't confuse monster laptops with P4s or Athlon64s with laptops with the P-Mobility chip (Dothan right now).
FINALLY.. my powerbook gets REALLY hot sometimes.
ffakr.
If you take the heatsink off of a dual core Pentium, you just might get it to weld the cores together
When you put two 32 bit cpu cores together, its a logarithmic factor, so 32 + 32 is 33 bits
When you put two 32 bit cpu cores together, its a logarithmic factor, so 32 + 32 is 33 bits
Flynnstone said:If you take the heatsink off of a dual core Pentium, you just might get it to weld the cores together
When you put two 32 bit cpu cores together, its a logarithmic factor, so 32 + 32 is 33 bits
the pentium D is contains two 64-bit cores, it's like two 600 series botched together.
Ownership
I'm realy glad that Steve took ownership about the 3Ghz fiasco & even owning up to the fact that no G5 PowerBook has come. That was really important. I just want to see that there is a replacement for the performance that AltiVec/Velocity Engine gives us today when the Intels debut.
I'm realy glad that Steve took ownership about the 3Ghz fiasco & even owning up to the fact that no G5 PowerBook has come. That was really important. I just want to see that there is a replacement for the performance that AltiVec/Velocity Engine gives us today when the Intels debut.
dongmin said:
Hmm. Honestly I don't see the first Mactels at MWSF. Here is why: when Jobs explained the switch to the developers he said that the first product will be in the marketplace by that time. That means in about a year. It is really very important what he told the developers because they need to get their products out with universal binaries. I he wanted them to ship them beginning next year he would have said so.
My guess therefore: first Mactels at 06-06-06