I doubt it; if it is released in the next few months and is the chip that comes at .09 process. Otherwise, it sounds like it could be the "G5" for the PowerBook, while a higher clocked 980 would push the Power Macs to "G6". I think they would stick with G5 for the 980-based models.
I don't see why not. It is the next generation, and IBM have acknowledged that by calling the design on which it is based the Power5, the 980 is a derivitave of the Power5.
Jumping from the G5 to the G6 in just over a year would show the world that Apple is on the move. Intel moved quickly from the Pentium to the Pentum II. The 980 based machine is clearly going to be a much more compattitive machine thatthe G5. The G5 is intouch buit isn't all conquring. The G6 just might really have that crown.
It is my understanding that the 980 will be based off of the Power 5, not the Power 4 as the 970 is. If this is the case then it is a new generation and warrents the new G number. If the 980 is just an evolution of the 970, then it there is no need for an incramental G rating.
Unfortunately there is little (none) official information on the 980. The name itself implies , however, more than a simple change in manufacturing process and cache upgrade. It suggests that the 980 will be based on the Power5 and include SMT, upgraded cache scheme (faster and larger), maybe an on die memory controller, and other advanced features slated for release with the Power5.
The question is if the 980 is running parallel to the power5's schedule, then IBM may either skip making a 90nm 970, or produce both a 90nm 970 and 980. Adding more power saving features to the 970 at 90nm would make it an ideal portable processor as well as a blade or 1U server processor. For workstations, you can also use this processor, but for higher end workstations the 980 would be the monster.
No way. They will call it the G5, or the G5+. Too much marketing has gone into G5 and it is WAY too soon to ditch it for the G6. "Yeah, that G5 chip was good, bit it only lasted for half a year". Besides, the G means generation and the 980 is still Power4 derived, so technically its parent has not undergone a generation increment, so it won't either.
edit: I may be worong in terms of what generation the 980 is a part of, but I doubt that if it is coming in march that it is derived from the Power5.
I think it depends a lot on what the 980 is (if they dub a chip '980'). If it's a 970, with 1MB L2 cache on 90 nm, it will be G5 still.
I do believe that they'll rename it to G6 if it turns up to be directly derived off the POWER5 as the 970 is derived off the POWER4. It will then have good SMT (which is a major change), more rename registers, more L2 cache, ++, which is enough difference for dubbing it The Next Generation, G6. IBM does it (POWER4 -> POWER5), why shouldn't Apple?
I think Apple has historically jumped a generation when one of three things happened:
1. There was a major architectural change in the processor design. The G4 added Altivec, the G5 added 64 bit support.
2. A generation has gotten a reputation for sloth (the G4)
3. A significant performance enhancement at the same clock speed when similar clock speeds exist (the G3 over the G2).
Since we suspect that the 980 doesn't add anything architecturally, the only way it will be called a G6 IMO is if it is significantly faster clock for clock than the G5, and it has versions running at similar clock speeds. That will allow Apple to differentiate between product lines without confusion.
So it is going to be a bit longer than a few months before we see 90nm processors from IBM. And further, these 90nm parts will be 970's. What does this say? Perhaps the 3.0GHz processor will be the first on the 90nm process? If that is true, we might not see a 980/G6/whatever for a year+, as the 90nm 970+ would be able to probably scale quit well (at least to 4GHz I woudl imagine).
The G3 name applies to the PPC 740, 750, 750cx, 750cxe, and 750fx. It actually applies to more processors but those are the ones that Apple has called G3s.
Likewise, the G4 name is used for any Mac using the 7400, 7410, 7447, 7450, or 7455. I think there are more as well (7441? 7457?), but you get the idea.
So it is going to be a bit longer than a few months before we see 90nm processors from IBM. And further, these 90nm parts will be 970's. What does this say? Perhaps the 3.0GHz processor will be the first on the 90nm process? If that is true, we might not see a 980/G6/whatever for a year+, as the 90nm 970+ would be able to probably scale quit well (at least to 4GHz I woudl imagine).
Based on how Apple has dramatically over-engineered the current G5 tower, I think it's a good guess that the next iteration of shipping G5 desktops will use the older fabbing process. The switch to 90mm will hopefully allow Jobs to deliver on his promise of dual 3GHz G5s for Summer 04.
Unfortunately there is little (none) official information on the 980.
AFAIK even the moniker "980" is unofficial at best but probably even non existant. It appeard the first time in a The Register-artichle if my memory serves me right, and their record of PowerPC reports are spotty at best (remember G5 reports from Motorola?) The only hard evidence of the successor to 970 that is readily available is this PDF in witch an IBM technician clearly uses the designations "GR-UL" and "97x" for this processor. For me.. there is no 980, only GR-UL or 97x. I for one expect the "x" to become a "5" when the design is official, ie PowerPC 975. But this is just me.
I really hope that the designation "980" would die. I consider any report from "insiders" concerning a PowerPC 980 fake. I'd really like to hear from the AI folks just why we should believe "980" when IBM them selves seem to use another designation.
Comments
Jumping from the G5 to the G6 in just over a year would show the world that Apple is on the move. Intel moved quickly from the Pentium to the Pentum II. The 980 based machine is clearly going to be a much more compattitive machine thatthe G5. The G5 is intouch buit isn't all conquring. The G6 just might really have that crown.
The question is if the 980 is running parallel to the power5's schedule, then IBM may either skip making a 90nm 970, or produce both a 90nm 970 and 980. Adding more power saving features to the 970 at 90nm would make it an ideal portable processor as well as a blade or 1U server processor. For workstations, you can also use this processor, but for higher end workstations the 980 would be the monster.
edit: I may be worong in terms of what generation the 980 is a part of, but I doubt that if it is coming in march that it is derived from the Power5.
Originally posted by Amorph
The POWER5 core is based on the POWER4 core. Obviously, there are a lot of improvements, but I'm not sure whether they quality as generational.
I think the biggest change is SMT and the addition of VMX.
Originally posted by Outsider
I think the biggest change is SMT and the addition of VMX.
There have been no mention of VMX for the POWER5 yet.
Originally posted by Zapchud
There have been no mention of VMX for the POWER5 yet.
I meant SIMD unit and assumed it would be VMX. But I do recall there being mention of a SIMD unit. Must do research...
Originally posted by DHagan4755
What do you think?
I think it depends a lot on what the 980 is (if they dub a chip '980'). If it's a 970, with 1MB L2 cache on 90 nm, it will be G5 still.
I do believe that they'll rename it to G6 if it turns up to be directly derived off the POWER5 as the 970 is derived off the POWER4. It will then have good SMT (which is a major change), more rename registers, more L2 cache, ++, which is enough difference for dubbing it The Next Generation, G6. IBM does it (POWER4 -> POWER5), why shouldn't Apple?
Originally posted by Outsider
I meant SIMD unit and assumed it would be VMX. But I do recall there being mention of a SIMD unit. Must do research...
IBM has mentioned no SIMD unit for the POWER5, and the die photo shows no SIMD unit either.
Originally posted by Zapchud
IBM has mentioned no SIMD unit for the POWER5, and the die photo shows no SIMD unit either.
Huh, I guess it was one of those rumors that was taken as fact.
edit: actually after further looking into thye matter, I'm getting my vapor processors mixed up. I may be thinking of the 980 (still vapor).
Airport -> Airport Extreme
Quartz -> Quartz Extreme
G5 -> G5 Extreme
1. There was a major architectural change in the processor design. The G4 added Altivec, the G5 added 64 bit support.
2. A generation has gotten a reputation for sloth (the G4)
3. A significant performance enhancement at the same clock speed when similar clock speeds exist (the G3 over the G2).
Since we suspect that the 980 doesn't add anything architecturally, the only way it will be called a G6 IMO is if it is significantly faster clock for clock than the G5, and it has versions running at similar clock speeds. That will allow Apple to differentiate between product lines without confusion.
Originally posted by Rhumgod
I doubt it; if it is released in the next few months and is the chip that comes at .09 process.
IBM to present info on 90 nm G5 PPC 970 at Feb. 2004 ISSCC meeting.
Original Japanese article & Babelfish translation
So it is going to be a bit longer than a few months before we see 90nm processors from IBM. And further, these 90nm parts will be 970's. What does this say? Perhaps the 3.0GHz processor will be the first on the 90nm process? If that is true, we might not see a 980/G6/whatever for a year+, as the 90nm 970+ would be able to probably scale quit well (at least to 4GHz I woudl imagine).
Likewise, the G4 name is used for any Mac using the 7400, 7410, 7447, 7450, or 7455. I think there are more as well (7441? 7457?), but you get the idea.
Originally posted by kupan787
IBM to present info on 90 nm G5 PPC 970 at Feb. 2004 ISSCC meeting.
Original Japanese article & Babelfish translation
So it is going to be a bit longer than a few months before we see 90nm processors from IBM. And further, these 90nm parts will be 970's. What does this say? Perhaps the 3.0GHz processor will be the first on the 90nm process? If that is true, we might not see a 980/G6/whatever for a year+, as the 90nm 970+ would be able to probably scale quit well (at least to 4GHz I woudl imagine).
Based on how Apple has dramatically over-engineered the current G5 tower, I think it's a good guess that the next iteration of shipping G5 desktops will use the older fabbing process. The switch to 90mm will hopefully allow Jobs to deliver on his promise of dual 3GHz G5s for Summer 04.
Originally posted by Outsider
Unfortunately there is little (none) official information on the 980.
AFAIK even the moniker "980" is unofficial at best but probably even non existant. It appeard the first time in a The Register-artichle if my memory serves me right, and their record of PowerPC reports are spotty at best (remember G5 reports from Motorola?) The only hard evidence of the successor to 970 that is readily available is this PDF in witch an IBM technician clearly uses the designations "GR-UL" and "97x" for this processor. For me.. there is no 980, only GR-UL or 97x. I for one expect the "x" to become a "5" when the design is official, ie PowerPC 975. But this is just me.
I really hope that the designation "980" would die. I consider any report from "insiders" concerning a PowerPC 980 fake. I'd really like to hear from the AI folks just why we should believe "980" when IBM them selves seem to use another designation.