Actually, you can purchase 2ghz 7447b g4's... I have a feeling apple only updated to 1.67 in order to have a little breathing room. Say they upgraded straight to 1.8 or 2ghz last jan... and the powerbook updates are due in November (or sooner)... apple would have NO WHERE TO UPGRADE TO. They made it this way so they can have at least a few more upgrades since they had NO idea when the 970gx was going to be ready. Now with the intel transition they really only need 1 more upgrade cycle with the 7447 or 7448... then its all said and done.
The gimmmicky 1.8 or the 2.0 or overclocked 1.67,Right, they yanked off the L3 and said presto see we have now something that can hit 1.8 Please just more Moto Crap.
The gimmmicky 1.8 or the 2.0 or overclocked 1.67,Right, they yanked off the L3 and said presto see we have now something that can hit 1.8 Please just more Moto Crap.
Umm l3 cache hasn't been present for quite a while. Have you noticed that the titanium powerbooks were the last ones to have l3 cache?
In any event... Overclock is still real clockspeed.
Umm l3 cache hasn't been present for quite a while. Have you noticed that the titanium powerbooks were the last ones to have l3 cache?
In any event... Overclock is still real clockspeed.
Whatever a 1.4 with L3 is more CPU then a 1.6 but who gives a crap about motostink? they were left in Intel 's Dust years ago just as G5 is now. PPC is slow crap compared to Intel/AMD. It is why Apple is using 2 cpu's in their machines for years.
Are they really saying that they'll debut a 1.7 GHz part sometime after October? And that's exciting? So we'll see a bump in the PB's from 1.67 to 1.70? In six months? The low-wattage is great, but this is exactly why Apple can't rely anymore on an embedded chip maker for PC parts. Bring on Yonah...
oh i forgot to add, for those who complain that we will see no upgrades to the PB line. i say its better that way, we could switch to intel today -- I mean do you really want a shit book like what Pc's are today... i mean who wants a 2 hour battery life "portable" (cough desktop cough) computer? Besides i doubt it would be much better as far as preformance... everyone knows Intel Mhz are worth real much.
I think this post makes a good point about the difficulty Apple is going to face overcoming their own marketing. Factually, every single thing this poster says is incorrect, but the perception Apple has built up of Intel as hot-running, not-as-fast-per-MHz, low battery-life portables is one that Apple is going to have to overcome to sell back to their established base next year.
The fact remains that this G4 is an extremely low-end chip with a very slow bus to RAM, so if it didn't draw extremely low power one would have to wonder. Even so, "typical" power consumption of 10W compared to the "max" Pentium M power consumption around 25W (probably half that "typical") isn't all that impressive. Consider that the Pentium M will outperform it computationally, has a bus to RAM almost 3x faster, and if you include not having to develop your own chipset, probably cheaper.
The bright side for Apple is that as long as people have a perception of the PowerPC as a good laptop chip, the intermediate "Osborne" effect won't be as bad.
Well this is an another reason, to be happy to see that Apple ditched the ppc chips.
Frankly, the release in late 2005 of a 1,7 ghz chip with 1 MB L2 cache, a stunning 200 mhz front bus, and an late 90 nm process, is not exciting at all.
The PPC chips could have been great, but the people behind it, do not wanted to make the necessary investissements in the desktop line. A Dothan is already 90 nm, has a 2 MB L 2 cache, a 400 mhz front bus and is clocked beyond 2 ghz now. Freescale is nearly one year behind ...
The G4 isn't a bad chip really. It's just that the infrastructure to support it has lagged way behind.
Consider that the Pentium M is a tweaked Pentium III with a massively updated infrastructure ... if someone at Motorola or Freescale had not just ignored the future and had actually done something similar, there'd have been a reasonably decent processor for laptops still!
Whilst this 90nm G4 will use less power than a Pentium M of the same speed, it could be held back because of the 200MHz bus. That's only 33MHz faster than now, unless they've DDR'd it. And I don't think they have.
From a Register story about the 'launch' on the 7448 dated September 2004 (how pitiful):
"The 90nm 7448 provides 0.9, 1.0 and 1.1V operating voltages, yielding power dissipation of 10W at 1.4GHz, Freescale claims, with compares well with the 7447A's typical and maximum dissipation figures of 21W and 30W, respectively, at 1.4GHz."
October is 2H 05. Right smack in the middle, actually.
Whoops! Mental transposition. The original schedule for volume production of the 7448 was 1H 05. An October 05 volume production date is 3 months late from that schedule.
Quote:
The question is whether Apple would use these if you believe their switching to Intel within 6 months of these things in production.
I think Apple will use these in Powerbooks up to 1.8 GHz, iBooks up to 1.5 GHz and Mac minis up to 1.5 GHz.
Tiger/Intel likely has another 7 to 8 months of debugging and optimizations left to do, so the earliest Mac/Intel machine that can ship is late Q1 06 or early Q2 06. That's enough time for one more PowerPC update of each Mac line.
The bright side for Apple is that as long as people have a perception of the PowerPC as a good laptop chip, the intermediate "Osborne" effect won't be as bad.
The thing is that to the general public, they don't even know that there's an intel switch, if you go out on the streets and asks people today, the majority won't know that Apple is switching to Intel, go to the Apple webpage, the only place that you'll find information that Apple has switched is if youwatched the keynote, or go into the press release section.
Whoops! Mental transposition. The original schedule for volume production of the 7448 was 1H 05. An October 05 volume production date is 3 months late from that schedule.
I think Apple will use these in Powerbooks up to 1.8 GHz, iBooks up to 1.5 GHz and Mac minis up to 1.5 GHz.
Tiger/Intel likely has another 7 to 8 months of debugging and optimizations left to do, so the earliest Mac/Intel machine that can ship is late Q1 06 or early Q2 06. That's enough time for one more PowerPC update of each Mac line.
A good timeframe, let's not forget that Apple will want some of their developers to have products ready to ship when these are released as well, if not it will look bad for Apple, indicating that the "switch" was not as easy as Apple said it was.
I would hope that the Mini tops out at 1.6, a ~200 mhz over their current offerings. 1.5 Ghz is a pittiful upgrade for a 1.42 that they currently offer.
I tend to agree. At this late hour, I highly doubt that this chip will make its way into PowerBooks. Perhaps future iBooks, minis, or emacs, but the race is on against intel. I have no idea who'll win this one, but either way its only a short term win as we know that intel is the big winner in the long run.
The G4 isn't a bad chip really. It's just that the infrastructure to support it has lagged way behind.
Consider that the Pentium M is a tweaked Pentium III with a massively updated infrastructure ... if someone at Motorola or Freescale had not just ignored the future and had actually done something similar, there'd have been a reasonably decent processor for laptops still!
Whilst this 90nm G4 will use less power than a Pentium M of the same speed, it could be held back because of the 200MHz bus. That's only 33MHz faster than now, unless they've DDR'd it. And I don't think they have.
The G4 was a great chip in its day. While it's unclear if the "10 Watts" is "typical" or "max", that probably is at least in the ballpark of the Pentium M under typical usage (word processing, web browsing, coding, etc.) And while the Pentium M branched from the Pentium III separately from the Pentium 4, there has been a lot of technology going back and forth. Benchmarks with similar desktop chipsets show the Pentium M 2.13GHz part in the same performance ballpark as the Pentium 4 3.6GHz except for heavy SIMD usage.
The 200MHz FSB definitely hurts the G4. The Pentium M currently has a 533MHz DDR2 bus, and is going to 667MHz with the next revision. Combined with the fact that x86 code is smaller than PowerPC and the Pentium M has 2MB of L2 cache, and any CPU at that bus speed that could have competed with Pentium M would simply starve.
Your point is taken, though. What could have been, had the G4 been developed seriously... It probably could have competed against the Pentium M with a better bus, cache, and a few other minor updates.
Whoops! Mental transposition. The original schedule for volume production of the 7448 was 1H 05. An October 05 volume production date is 3 months late from that schedule.
Let me refresh a little your memory . Look at slide 41. For the 7448 it says samples 1H-05, production 2H-05. For the 8641 family, samples 2H-05, production 1H-06.
So, the fact that Freescale comes late in technologies like 90nm (and other important improvements), should not prevent us from seeing that the 7447A/B and 7448 chips are produced on schedule. Most probably the 8641 too. Which is worth mentioning given their dark past of the Motorola times.
I tend to agree. At this late hour, I highly doubt that this chip will make its way into PowerBooks.
And what the Powerbook is supposed to get? It must be updated before the end of the year. The 7448 timing is perfect. And something tells me it will be the last PowerPC chip in the Powerbooks.
Stupid question ahead....I know what embedded processors are and I know what discrete processors are.....what I don't know is physically what's the difference between the two processors? Can embedded processors easily be adapted to discrete processors?
Comments
No wonder Jobs jumped ship, he had no options in PPC land.
Originally posted by emig647
Actually, you can purchase 2ghz 7447b g4's... I have a feeling apple only updated to 1.67 in order to have a little breathing room. Say they upgraded straight to 1.8 or 2ghz last jan... and the powerbook updates are due in November (or sooner)... apple would have NO WHERE TO UPGRADE TO. They made it this way so they can have at least a few more upgrades since they had NO idea when the 970gx was going to be ready. Now with the intel transition they really only need 1 more upgrade cycle with the 7447 or 7448... then its all said and done.
The gimmmicky 1.8 or the 2.0 or overclocked 1.67,Right, they yanked off the L3 and said presto see we have now something that can hit 1.8 Please just more Moto Crap.
Originally posted by Aurora
The gimmmicky 1.8 or the 2.0 or overclocked 1.67,Right, they yanked off the L3 and said presto see we have now something that can hit 1.8 Please just more Moto Crap.
Umm l3 cache hasn't been present for quite a while. Have you noticed that the titanium powerbooks were the last ones to have l3 cache?
In any event... Overclock is still real clockspeed.
Originally posted by emig647
Umm l3 cache hasn't been present for quite a while. Have you noticed that the titanium powerbooks were the last ones to have l3 cache?
In any event... Overclock is still real clockspeed.
Whatever a 1.4 with L3 is more CPU then a 1.6 but who gives a crap about motostink? they were left in Intel 's Dust years ago just as G5 is now. PPC is slow crap compared to Intel/AMD. It is why Apple is using 2 cpu's in their machines for years.
Originally posted by THT
Wow. Whatever happened to volume production in 2H 05? It's October now?
October is 2H 05. Right smack in the middle, actually.
The question is whether Apple would use these if you believe their switching to Intel within 6 months of these things in production.
Originally posted by mike12309
oh i forgot to add, for those who complain that we will see no upgrades to the PB line. i say its better that way, we could switch to intel today -- I mean do you really want a shit book like what Pc's are today... i mean who wants a 2 hour battery life "portable" (cough desktop cough) computer? Besides i doubt it would be much better as far as preformance... everyone knows Intel Mhz are worth real much.
I think this post makes a good point about the difficulty Apple is going to face overcoming their own marketing. Factually, every single thing this poster says is incorrect, but the perception Apple has built up of Intel as hot-running, not-as-fast-per-MHz, low battery-life portables is one that Apple is going to have to overcome to sell back to their established base next year.
The fact remains that this G4 is an extremely low-end chip with a very slow bus to RAM, so if it didn't draw extremely low power one would have to wonder. Even so, "typical" power consumption of 10W compared to the "max" Pentium M power consumption around 25W (probably half that "typical") isn't all that impressive. Consider that the Pentium M will outperform it computationally, has a bus to RAM almost 3x faster, and if you include not having to develop your own chipset, probably cheaper.
The bright side for Apple is that as long as people have a perception of the PowerPC as a good laptop chip, the intermediate "Osborne" effect won't be as bad.
Frankly, the release in late 2005 of a 1,7 ghz chip with 1 MB L2 cache, a stunning 200 mhz front bus, and an late 90 nm process, is not exciting at all.
The PPC chips could have been great, but the people behind it, do not wanted to make the necessary investissements in the desktop line. A Dothan is already 90 nm, has a 2 MB L 2 cache, a 400 mhz front bus and is clocked beyond 2 ghz now. Freescale is nearly one year behind ...
Pathetic
case closed
Consider that the Pentium M is a tweaked Pentium III with a massively updated infrastructure ... if someone at Motorola or Freescale had not just ignored the future and had actually done something similar, there'd have been a reasonably decent processor for laptops still!
Whilst this 90nm G4 will use less power than a Pentium M of the same speed, it could be held back because of the 200MHz bus. That's only 33MHz faster than now, unless they've DDR'd it. And I don't think they have.
From a Register story about the 'launch' on the 7448 dated September 2004 (how pitiful):
"The 90nm 7448 provides 0.9, 1.0 and 1.1V operating voltages, yielding power dissipation of 10W at 1.4GHz, Freescale claims, with compares well with the 7447A's typical and maximum dissipation figures of 21W and 30W, respectively, at 1.4GHz."
And the Freescale release:
http://www.freescale.com/webapp/sps/...elease&tid=FSH
Maybe Apple will use that TSI108 northbridge. There's no need for them to keep making their own northbridges!
http://www.tundra.com/Products/PowerPC/Tsi108/index.cfm
Originally posted by Louzer
October is 2H 05. Right smack in the middle, actually.
Whoops! Mental transposition.
The question is whether Apple would use these if you believe their switching to Intel within 6 months of these things in production.
I think Apple will use these in Powerbooks up to 1.8 GHz, iBooks up to 1.5 GHz and Mac minis up to 1.5 GHz.
Tiger/Intel likely has another 7 to 8 months of debugging and optimizations left to do, so the earliest Mac/Intel machine that can ship is late Q1 06 or early Q2 06. That's enough time for one more PowerPC update of each Mac line.
Originally posted by Louzer
With the powerbook G5's just around the corner (we all know they are!),
DONT SMOKE CRACK.../\\ is why, ok kids?
Originally posted by Booga
The bright side for Apple is that as long as people have a perception of the PowerPC as a good laptop chip, the intermediate "Osborne" effect won't be as bad.
The thing is that to the general public, they don't even know that there's an intel switch, if you go out on the streets and asks people today, the majority won't know that Apple is switching to Intel, go to the Apple webpage, the only place that you'll find information that Apple has switched is if youwatched the keynote, or go into the press release section.
The Osborne effect is overrated
Originally posted by THT
Whoops! Mental transposition.
I think Apple will use these in Powerbooks up to 1.8 GHz, iBooks up to 1.5 GHz and Mac minis up to 1.5 GHz.
Tiger/Intel likely has another 7 to 8 months of debugging and optimizations left to do, so the earliest Mac/Intel machine that can ship is late Q1 06 or early Q2 06. That's enough time for one more PowerPC update of each Mac line.
A good timeframe, let's not forget that Apple will want some of their developers to have products ready to ship when these are released as well, if not it will look bad for Apple, indicating that the "switch" was not as easy as Apple said it was.
I would hope that the Mini tops out at 1.6, a ~200 mhz over their current offerings. 1.5 Ghz is a pittiful upgrade for a 1.42 that they currently offer.
Originally posted by nathan22t
"the embedded market"
case closed
I tend to agree. At this late hour, I highly doubt that this chip will make its way into PowerBooks. Perhaps future iBooks, minis, or emacs, but the race is on against intel. I have no idea who'll win this one, but either way its only a short term win as we know that intel is the big winner in the long run.
Originally posted by Hattig
The G4 isn't a bad chip really. It's just that the infrastructure to support it has lagged way behind.
Consider that the Pentium M is a tweaked Pentium III with a massively updated infrastructure ... if someone at Motorola or Freescale had not just ignored the future and had actually done something similar, there'd have been a reasonably decent processor for laptops still!
Whilst this 90nm G4 will use less power than a Pentium M of the same speed, it could be held back because of the 200MHz bus. That's only 33MHz faster than now, unless they've DDR'd it. And I don't think they have.
The G4 was a great chip in its day. While it's unclear if the "10 Watts" is "typical" or "max", that probably is at least in the ballpark of the Pentium M under typical usage (word processing, web browsing, coding, etc.) And while the Pentium M branched from the Pentium III separately from the Pentium 4, there has been a lot of technology going back and forth. Benchmarks with similar desktop chipsets show the Pentium M 2.13GHz part in the same performance ballpark as the Pentium 4 3.6GHz except for heavy SIMD usage.
The 200MHz FSB definitely hurts the G4. The Pentium M currently has a 533MHz DDR2 bus, and is going to 667MHz with the next revision. Combined with the fact that x86 code is smaller than PowerPC and the Pentium M has 2MB of L2 cache, and any CPU at that bus speed that could have competed with Pentium M would simply starve.
Your point is taken, though. What could have been, had the G4 been developed seriously... It probably could have competed against the Pentium M with a better bus, cache, and a few other minor updates.
Originally posted by THT
Whoops! Mental transposition.
Let me refresh a little your memory
So, the fact that Freescale comes late in technologies like 90nm (and other important improvements), should not prevent us from seeing that the 7447A/B and 7448 chips are produced on schedule. Most probably the 8641 too. Which is worth mentioning given their dark past of the Motorola times.
Originally posted by Xool
I tend to agree. At this late hour, I highly doubt that this chip will make its way into PowerBooks.
And what the Powerbook is supposed to get? It must be updated before the end of the year. The 7448 timing is perfect. And something tells me it will be the last PowerPC chip in the Powerbooks.
"the embedded market"
Stupid question ahead....I know what embedded processors are and I know what discrete processors are.....what I don't know is physically what's the difference between the two processors? Can embedded processors easily be adapted to discrete processors?