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.
G6 Rumors
- Thread starter MacRumors
- Start date
- Sort by reaction score
Re: PPC vs PowerX
970 is based on Power4 technology according to IBM. When you do a search on thier website there is a picture of the two side by side. Power4 being twice the size physically as the 970. G5 is a PPC970 with Altivec added. From other stories I am guessing that this was not an easy task. It's my guess that the Power5 will not be the parent chip to the the 3.0 chips but will probably show up in the 4.0 chips.
side note:
RAM manufacturers are probably loving this because they can hike up prices for all the new kinds of ram.
970 is based on Power4 technology according to IBM. When you do a search on thier website there is a picture of the two side by side. Power4 being twice the size physically as the 970. G5 is a PPC970 with Altivec added. From other stories I am guessing that this was not an easy task. It's my guess that the Power5 will not be the parent chip to the the 3.0 chips but will probably show up in the 4.0 chips.
side note:
RAM manufacturers are probably loving this because they can hike up prices for all the new kinds of ram.
Re: Brother Mugga
With all due respect to you, Jaedreth, your statements are no more accurate or factual than mine are. We simply have differing opinions.
Your claim is that Apple could release the Power 5 Lite/PPC 980 and still call it a G5. Theoretically, that is true.
My claim is that if Apple were to release a Power 5 Lite/PPC 980, that it would probably be a significant enough difference that it would warrant the monicker G6. Theoretically, that is also true.
What am I using to base my claim on?
1 - If the Power4 and Power5 are essentially the same generation, then WHY is it called the Power5? Why not the Power 4++? (The Power4+ came first.) Do you have any evidence to support that the Power5 is NOT a significant upgrade from the Power4? Quoting from IBM's own web site:
"Talk about my generation: POWER5-based servers spring to life
IBM has announced that the first servers based on its next generation POWER5 microprocessors are up and running in IBM's Poughkeepsie labs. Initial internal performance tests indicate that POWER5 based eServer systems are expected to offer four times the system performance over the first POWER4 based servers."
Right there, in IBM's own words...the Power5 is GENERATION beyond the Power4.
2 - The claims that Power5 is 4 times faster than Power4 certainly seem to indicate that it is a significant redesign. So if a Power5 is a new generation from Power4 (which IBM clearly states it is) then a Power5 Lite/PPC 980 would be a new generation from the Power4 Lite/PPC 970. And thusly, Apple would MOST LIKELY call it the G6, because it would NOT be the same as a G5.
3 - By your definition, the G4 chip is really same generation as the G3 chip. (And in many ways I agree with you...the original G4 was a glorified G3 with Altivec tacked on.) BUT Apple called it the G4 because it was different ENOUGH to warrant a new monicker.
4 - Today's G4 chips are very different from the original 7400...but still based on the 7400. So every new version has been called the G4. (By Apple...I never claimed these Gx monickers have anything to do with Motorola or IBM.) BUT the G5 is a whole new ballgame, and Apple is shouting "Hey guys, this is really new!". They used "G5" as the machine name to clearly say "this is the next generation...it's not a G4".
So if the PPC 980 is up to 4 times faster than the PPC 970, with other benefits, you are correct that Apple COULD sell it as the G5 if they wanted to. BUT I can't see why Apple would do so. It would be AT LEAST as big a jump from the PPC 970 to the 980 as the G3 to G4 transition was.
So I stand by my assertion that any chip based on the Power5 that Apple sells will be called a G6. I don't care what IBM calls it internally. If IBM felt the Power5 warranted a whole number increment vs. calling it the Power4(something), then Apple's marketing team is just as likely to increment the G5 to G6. They'd be foolish not to.
And to quote Dennis Miller: "Of course, that's just my opinion. I could be wrong..."
With all due respect to you, Jaedreth, your statements are no more accurate or factual than mine are. We simply have differing opinions.
Your claim is that Apple could release the Power 5 Lite/PPC 980 and still call it a G5. Theoretically, that is true.
My claim is that if Apple were to release a Power 5 Lite/PPC 980, that it would probably be a significant enough difference that it would warrant the monicker G6. Theoretically, that is also true.
What am I using to base my claim on?
1 - If the Power4 and Power5 are essentially the same generation, then WHY is it called the Power5? Why not the Power 4++? (The Power4+ came first.) Do you have any evidence to support that the Power5 is NOT a significant upgrade from the Power4? Quoting from IBM's own web site:
"Talk about my generation: POWER5-based servers spring to life
IBM has announced that the first servers based on its next generation POWER5 microprocessors are up and running in IBM's Poughkeepsie labs. Initial internal performance tests indicate that POWER5 based eServer systems are expected to offer four times the system performance over the first POWER4 based servers."
Right there, in IBM's own words...the Power5 is GENERATION beyond the Power4.
2 - The claims that Power5 is 4 times faster than Power4 certainly seem to indicate that it is a significant redesign. So if a Power5 is a new generation from Power4 (which IBM clearly states it is) then a Power5 Lite/PPC 980 would be a new generation from the Power4 Lite/PPC 970. And thusly, Apple would MOST LIKELY call it the G6, because it would NOT be the same as a G5.
3 - By your definition, the G4 chip is really same generation as the G3 chip. (And in many ways I agree with you...the original G4 was a glorified G3 with Altivec tacked on.) BUT Apple called it the G4 because it was different ENOUGH to warrant a new monicker.
4 - Today's G4 chips are very different from the original 7400...but still based on the 7400. So every new version has been called the G4. (By Apple...I never claimed these Gx monickers have anything to do with Motorola or IBM.) BUT the G5 is a whole new ballgame, and Apple is shouting "Hey guys, this is really new!". They used "G5" as the machine name to clearly say "this is the next generation...it's not a G4".
So if the PPC 980 is up to 4 times faster than the PPC 970, with other benefits, you are correct that Apple COULD sell it as the G5 if they wanted to. BUT I can't see why Apple would do so. It would be AT LEAST as big a jump from the PPC 970 to the 980 as the G3 to G4 transition was.
So I stand by my assertion that any chip based on the Power5 that Apple sells will be called a G6. I don't care what IBM calls it internally. If IBM felt the Power5 warranted a whole number increment vs. calling it the Power4(something), then Apple's marketing team is just as likely to increment the G5 to G6. They'd be foolish not to.
And to quote Dennis Miller: "Of course, that's just my opinion. I could be wrong..."
IBM Power5
http://www-1.ibm.com/servers/eserver/pseries/news/pressreleases/2003/jun/milestone.html
Interesting note, IBM has had a contract to supply increasingly fast supercomputers on a factor of 10, contract with the Gov't that is...
The IBM G5 was used in the previous Supercomputer, promised to reach 10 Teraflops, whereas this new Power5 will be replacing the IBM G5 chips in ASCI Purple designs to produce 100 Teraflops.
(And the Power5 was listed by other sources as more than four times the processing power as Power4 on an individual chip basis. Further note, IBM must, in order to complete the contract, develop a 1000 Teraflop supercomputer within the next few years. Most likely will be a Power6, and thus may have a derivative. That derivative, which is increasingly hypothetical, could be used in a G6 mac.)
Power5 is coming next year, which means greater boots to power and speed on G5's, in the form of a possible PPC 980.
Jaedreth
http://www-1.ibm.com/servers/eserver/pseries/news/pressreleases/2003/jun/milestone.html
Interesting note, IBM has had a contract to supply increasingly fast supercomputers on a factor of 10, contract with the Gov't that is...
The IBM G5 was used in the previous Supercomputer, promised to reach 10 Teraflops, whereas this new Power5 will be replacing the IBM G5 chips in ASCI Purple designs to produce 100 Teraflops.
(And the Power5 was listed by other sources as more than four times the processing power as Power4 on an individual chip basis. Further note, IBM must, in order to complete the contract, develop a 1000 Teraflop supercomputer within the next few years. Most likely will be a Power6, and thus may have a derivative. That derivative, which is increasingly hypothetical, could be used in a G6 mac.)
Power5 is coming next year, which means greater boots to power and speed on G5's, in the form of a possible PPC 980.
Jaedreth
that would be really good if the G6 was out sometime in the next year. as well as keeping the momentum going, which will change peoples general view of Apple, if they bring out the G6, or release a "G6" labelled Mac that will only be good for competition.
for example switchers will look at PC's and think "OK, that's got a P4" then they'll look at the Mac and go "whoa, that things got a G6!!"
for example switchers will look at PC's and think "OK, that's got a P4" then they'll look at the Mac and go "whoa, that things got a G6!!"
G6
Well, I've been arguing for a while now that this supposed 980 chip will *not necessarily* be called G6 by Apple Marketing.
Now let me say this: I hope they do.
Because if they release the 980 in 2005, and then a 990 based on a theoretical Power6 that I speculated on in another post, in 2007, cutting the generation gap in half, and providing cutting edge technology constantly, IBM and Apple can do nothing but win.
So now I'm just gonna shut up.
Jaedreth
Well, I've been arguing for a while now that this supposed 980 chip will *not necessarily* be called G6 by Apple Marketing.
Now let me say this: I hope they do.
Because if they release the 980 in 2005, and then a 990 based on a theoretical Power6 that I speculated on in another post, in 2007, cutting the generation gap in half, and providing cutting edge technology constantly, IBM and Apple can do nothing but win.
So now I'm just gonna shut up.
Jaedreth
The G* Series
It seems to me that the G3 was only around for one and a half years before the G4 came out mid '99. And seriously, a jump from a 450 B&W G3 to a 450 G4 was not that great to warrant a G* change IMO. The G4->G5 is seeming like it will be a far greater increase in not just clock speed, but architecture and performance.
I'm betting that this so-called PPC980 will be the same 970 core with a few tweaks and a 90nm process. And, it will still be called G5, coming about next summer at 3GHz.
I think the Power5-derived chip will be like a PPC9700 or something different like that. We might see it in a year and a half to two years and it *might* take the G6 name in apple's line.
It seems to me that the G3 was only around for one and a half years before the G4 came out mid '99. And seriously, a jump from a 450 B&W G3 to a 450 G4 was not that great to warrant a G* change IMO. The G4->G5 is seeming like it will be a far greater increase in not just clock speed, but architecture and performance.
I'm betting that this so-called PPC980 will be the same 970 core with a few tweaks and a 90nm process. And, it will still be called G5, coming about next summer at 3GHz.
I think the Power5-derived chip will be like a PPC9700 or something different like that. We might see it in a year and a half to two years and it *might* take the G6 name in apple's line.
Re: G6
It continues to amuse me when people think that what's happened with the G4 is the absolute precedent for what will happen in the future.
Eighteen months between G5 and G6 is not cutting the generation gap in half. It is bringing it back to what it should be.
Consider:
(G1) 601 March 1994
(G2) 603/604 April/May 1995
G3 November 1997
G4 September 1999
G5 May 2003
Thus, since the release of the PPC line, we've seen G# upgrades once every one to two years (with the glaring exception of the time taken between the G4 and the G5). I think that it's not unreasonable to guess that we'll continue to see this, and that the debut of the G6 is due some time between next summer and the following summer. Personally, I tend to plant myself in the middle of that range and vote for Winter '04/'05.
It continues to amuse me when people think that what's happened with the G4 is the absolute precedent for what will happen in the future.
Eighteen months between G5 and G6 is not cutting the generation gap in half. It is bringing it back to what it should be.
Consider:
(G1) 601 March 1994
(G2) 603/604 April/May 1995
G3 November 1997
G4 September 1999
G5 May 2003
Thus, since the release of the PPC line, we've seen G# upgrades once every one to two years (with the glaring exception of the time taken between the G4 and the G5). I think that it's not unreasonable to guess that we'll continue to see this, and that the debut of the G6 is due some time between next summer and the following summer. Personally, I tend to plant myself in the middle of that range and vote for Winter '04/'05.
Think about what's hot on the rumor plate now; the latest rev to the G4 chip that Motorola is dragging their feet on. Take a look at the original ``G3'' versus the one in the latest iBooks. while the numbers might not increment by 1 (7455, 7456, 7457, etc), it still probably signifies a revised processor (cooler, more efficient, bigger cache, you know).
Consider the following:
Out of all the chips, ``Power PC'' is the only naming convention that has stuck with them all.
You had your 601, 603, 604, G3, G4, and now G5. It's highly doubtful that the G6 will be called the PowerPC 9anything.
Remember, the G stands for generation.
Think about the G3. When it debuted in (1998?) it was the hottest, fastest Apple processor to date at a massive 233Mhz. That was the original revision of the generation 3 processor. Here we are now, mid-2003 and the ``G3'' still exists in a 900Mhz form. Though, even the producer of the chip (was Motorola, now IBM) has changed, it's still called a G3 because deep down inside it's apples third generation of technology. The G3's popularity grew quickly because of the introduction of the backside cache. It wasn't until the Altivec addition to the G4 happened (and correct me if I'm wrong on my Apple history here) that the G4 had a massive performance enhancement over the G3.
Isn't the original Intel Pentium classified as a P5? The code name for one of them, I know for a fact, was P55C, and then P55C-MMX, the Pentium Pro was the P6, and so forth.
Numbers, numbers, numbers. IMHO, Intel has dragged the Pentium name out too far. In theory, the Pentium 4, is what, a 786? 886?
If Apple kept the 68K chip names, what would the G5 be? 070?
Ho-hum
Sensationalism aside, what's the point of labeling this bit of hear-say as a "G6" rumor?
Think about it for a minute. The G3 family consists of a number of 700 series chips. The G4 family consists of a number of 7000 series chips. Why would the G5 family be limited to one lousy chip model, and why on Earth would Apple name an incremental upgrade of the design "G6"?
Yes, the 980 implements design changes from the Power5, but that doesn't make it a new chip generation. Heck, even IBM acknowledges this by not making it, say, "9800".
C'mon. Be sensible.
Sensationalism aside, what's the point of labeling this bit of hear-say as a "G6" rumor?
Think about it for a minute. The G3 family consists of a number of 700 series chips. The G4 family consists of a number of 7000 series chips. Why would the G5 family be limited to one lousy chip model, and why on Earth would Apple name an incremental upgrade of the design "G6"?
Yes, the 980 implements design changes from the Power5, but that doesn't make it a new chip generation. Heck, even IBM acknowledges this by not making it, say, "9800".
C'mon. Be sensible.
Re: Ho-hum
While you make a good point, I'd say that I see only one flaw in your logic. It's the rumor boards that have dubbed the Power5-Lite the "980", not IBM. If we had it on good authority from IBM that this was going to be called the 980, then, maybe, I'd agree with you.
The other side of the coin, however, is that going from Power4 to Power5 is considered a generational step. So, why wouldn't the Lite versions of these processors also go through a generational step, as well, regardless of what IBM chooses to number the chip as?
Food for thought...
While you make a good point, I'd say that I see only one flaw in your logic. It's the rumor boards that have dubbed the Power5-Lite the "980", not IBM. If we had it on good authority from IBM that this was going to be called the 980, then, maybe, I'd agree with you.
The other side of the coin, however, is that going from Power4 to Power5 is considered a generational step. So, why wouldn't the Lite versions of these processors also go through a generational step, as well, regardless of what IBM chooses to number the chip as?
Food for thought...
I don't think it's unreasonable that we'll have the G6 next year. After all, the 601 only lasted one year too. Great as it is, I can believe that the G5 is a transitionary chip.
I don't think Apple wants a two-tiered chip setup. With the G3 at least, they rapidly put one chip into every product. While Apple wants a pro/consumer distinction, they don't want the consumer product to be last year's pro product. They don't want the consumer to be a second class citizen. Problem is, with the G4 as the top chip, Motorola can't produce enough for Apple's entire product line. So, Apple's held back.
The 90 nm G5 and beyond will change that.
That's not the same. Apple actually changed architectures and gained IBM as a supplier. x86 didn't completely change architectures.
I don't think Apple wants a two-tiered chip setup. With the G3 at least, they rapidly put one chip into every product. While Apple wants a pro/consumer distinction, they don't want the consumer product to be last year's pro product. They don't want the consumer to be a second class citizen. Problem is, with the G4 as the top chip, Motorola can't produce enough for Apple's entire product line. So, Apple's held back.
The 90 nm G5 and beyond will change that.
Originally posted by BrandonRP0123
If Apple kept the 68K chip names, what would the G5 be? 070?
That's not the same. Apple actually changed architectures and gained IBM as a supplier. x86 didn't completely change architectures.
I don't think that's quite true. When the G3 came out Apple didn't have their product line set up with the consumer/pro-sumer lines yet. They really got there when the iMac was released, and it was released with a 233MHz processor, which was, in fact, last years product. Even when they updated it, five months later, it was only bumped to 266MHz, while the Power Mac G3 was at 300MHz-450MHz. When the G4 iMac was released, it was introduced with processor speeds roughly equivalent to the Power Mac from a year earlier.
So, I think that, while Apple may not want the overall technology of their consumer line to be last year's, they have fairly consistently used last year's Power Mac's processors.
But the point was, the iMac had a G3. Same processor type, lower speed. And at the time of immediate release, it wasn't that far off the Power Mac. It would be like if the iMac had the 1.6 GHz G5 now. I'm not saying there'll be parity, but the consumer Macs are going to have faster processors to keep up with (but slightly behind) the Power Macs. That means same processor, lower frequency.
Well, actually, my point is that it would be like the iMac getting a 1.6GHz G5 next May. Yes, by then we might have 2.6GHz G5s in the Power Macs, but what the iMac gets is last years Power Mac processor. Not only that, but, if we are to believe what rumors this thread is talking about, the Power Macs could go to the G6 not long after that.
I'm inclined to agree with Phil here. What we shouldn't lost sight of is that Apple had a policy (at the time of introduction of the iMac) that clearly expressed a desire to have the same processor generation (if a much lower clock frequency) in as many products as possible, consumer and professional. This obviously lowers the cost of the chips - don't forget that it was around this time that they were moving to the Unified Motherboard Architecture as well. Both strands of this policy had clear economic benefits. You make more processors , you make more motherboards of the same generation , you lower costs. Sure, you do things to differentiate them, like bus speed, and other onboard technologies, but if the products are sufficiently similar, it's all good news for Apple, and for the consumer.
Then, along comes the G4 to spoil the party. Here's a chip that promised to scale quickly, but doesn't. Then Apple have to take the marketing decision to differentiate their products on the basis of what sort of chip is inside them, so you have a period where G4 says professional user, and G3 says consumer, getting away from the unified philosophy again.
The fact that Apple wants to have the same chip generation in as many products as possible was again reiterated by the introduction of the flat panel iMac. Again, they tried to bring the best processor tech they could to the consumer table, but were forced to cripple the level of this technology on that occasion due to Moto's inability to produce sufficiently faster chips for the Pro line.
I think what I'm trying to get at really here, is that Apple knows it's a boutique manufacturer, so it's consumer products need to be seen to be close enough to the cutting edge. That they've failed in this during the G4 fiasco is clear enough from the sales of FP iMacs over the last year (which have been pretty poor, esp in the last quarter).
They have had a tough lesson from the G4 to G5 period, and I'm pretty sure they'll be pressing on with readying as many new technologies as possible in readiness for the next economic boom, which means a G6 is probably a fair way into development. If there has been an upside to all the faffing about over the G4, I suppose it's that they did get the OS sorted out, whilst focussing resources on some great new software tech (the iApps etc) to get us away from thinking about processors.
It's all turned out quite nicely actually
Well, now there's an argument that I can understand. The economics of this does make sense. Although, on the flip side, Apple can leverage tech that it's already invested in by using the same chip that it used in the Power models a year ago for the consumer models.
I think it all comes down to the economics of it, and I don't think we know enough about what's in Apple's books to do anything other than guess at what Apple's choices in this area are going to be.
Yes, it does all come down to the economics of it. However, without us knowing the per-chip price of the 970, it'll be difficult to extrapolate what we can expect from Apple in terms of G5 (and later, G6) rollout, and, for that matter, dropping the G3 and the G4. Also, where will the dual-processor/dual-core strategy go? I for one won't be buying another single-processor machine because I love duals so much.
It's a very interesting time to be a Mac user.
Interesting times...
Interesting times to be a mac user... In a chinese sense?
Jaedreth
(An old Chinese curse says, "May you live in interesting times...")
Interesting times to be a mac user... In a chinese sense?
Jaedreth
(An old Chinese curse says, "May you live in interesting times...")
Re: Interesting times...
Chinese curse or not, I'd honestly rather live in interesting times than boring times....
Chinese curse or not, I'd honestly rather live in interesting times than boring times....
-Gents
This thread is a terrific example of trying to apply marketing to tech.
We're trying to out-guess the marketing arm of Apple if rumors of a Power5-derived 980 chip does hit the Mac.
Would they brand it the G6? If 980 is to 970 as Power5 is to Power4, I sure as hell would.
And 12-18 months before it hits, is right back on schedule. Remember IBM was the manufacturer of the G1(601) as Moto didn't have the facilities yet. IBM set the pace, Moto had it for a while, but started losing it. I suspect they never had the facilities they claimed to.
Well cudos to IBM for taking it back and pulling through for us.
This thread is a terrific example of trying to apply marketing to tech.
We're trying to out-guess the marketing arm of Apple if rumors of a Power5-derived 980 chip does hit the Mac.
Would they brand it the G6? If 980 is to 970 as Power5 is to Power4, I sure as hell would.
And 12-18 months before it hits, is right back on schedule. Remember IBM was the manufacturer of the G1(601) as Moto didn't have the facilities yet. IBM set the pace, Moto had it for a while, but started losing it. I suspect they never had the facilities they claimed to.
Well cudos to IBM for taking it back and pulling through for us.
IBM's Strategy
IBM is pulling Apple out of the Fire big time here. (Motorola is the one who threw diesel on us and screamed, "Fire hazard, don't anyone throw a match...")
By IBM making a PPC-like derivative of Power4, and having plans for a PPC-like derivative of Power5 in the works simultaneously with Power5, IBM is able to produce next generation chips faster than Motorola can update their old pieces of crap.
Now even though technically they named the 970 the PowerPC 970, it *isn't* PowerPC. But it isn't Power either. It's a Power derivative with PowerPC like features. So technically it should be just called the 970, which we've been calling it.
IBM will be able to keep producing new generation chips for Apple and its lower end workstations. Now since these chips are Power based and are not PPC (thus not PPC gen 1, gen 2, gen3, gen 4), Apple can decide when to pull out the hammer label "g6", but it won't matter if it's the 980 or the 990 if they really exist.
Would a 980 labeled G5 be a worse machine than a 980 labeled G6?
Would a 990 labeled G6 be a worse machine than a 990 labeled G7?
The name isn't important. The important thing is: We are in good hands.
Get rid of Moto, no one can keep up with what IBM is doing. Why do we need a second vendor? In case IBM screws us over? Yeah right, Apple has the potential to make them *lots* of money.
In IBM We Trust.
Jaedreth
IBM is pulling Apple out of the Fire big time here. (Motorola is the one who threw diesel on us and screamed, "Fire hazard, don't anyone throw a match...")
By IBM making a PPC-like derivative of Power4, and having plans for a PPC-like derivative of Power5 in the works simultaneously with Power5, IBM is able to produce next generation chips faster than Motorola can update their old pieces of crap.
Now even though technically they named the 970 the PowerPC 970, it *isn't* PowerPC. But it isn't Power either. It's a Power derivative with PowerPC like features. So technically it should be just called the 970, which we've been calling it.
IBM will be able to keep producing new generation chips for Apple and its lower end workstations. Now since these chips are Power based and are not PPC (thus not PPC gen 1, gen 2, gen3, gen 4), Apple can decide when to pull out the hammer label "g6", but it won't matter if it's the 980 or the 990 if they really exist.
Would a 980 labeled G5 be a worse machine than a 980 labeled G6?
Would a 990 labeled G6 be a worse machine than a 990 labeled G7?
The name isn't important. The important thing is: We are in good hands.
Get rid of Moto, no one can keep up with what IBM is doing. Why do we need a second vendor? In case IBM screws us over? Yeah right, Apple has the potential to make them *lots* of money.
In IBM We Trust.
Jaedreth
When the iMacs were announced in May 1998, there was a Power Mac G3 at 233 MHz. The iMacs were released in August. So the iMac gets the low end Power Mac processor of 3 months ago