How is IBM going to get to 2,ghz this year? The 970 caps at 1.8ghz and the 970+ which caps at 2.4ghz won't even be thought of til next year.
I will be floored if we get a 2ghz machine this year, and when i pick myself up, I'll be reaching for my credit card. I'm expecting the 1.8ghz 970 in a phat new Ives design (who btw is getting 'stale' according to UK design community) and as any designer knows, that's modivation.
<strong>How is IBM going to get to 2,ghz this year? The 970 caps at 1.8ghz and the 970+ which caps at 2.4ghz won't even be thought of til next year.</strong><hr></blockquote>
The 970 is scheduled to start with a top speed of 1.8GHz. There's nothing saying that it won't get faster from there.
Then there's the rapid die shrink to .09μ.
If you stretch the "year" to include MWSF '04 the odds are pretty darn good IMO.
[quote]<strong>I'm expecting the 1.8ghz 970 in a phat new Ives design (who btw is getting 'stale' according to UK design community) and as any designer knows, that's modivation.</strong><hr></blockquote>
Yeah, white is so five minutes ago.
It'll be interesting to see what they do. If they stick with a minitower, it'll be damn hard to beat El Capitan at its current level of refinement. But then FW3200 opens up a lot of interesting possibilities, doesn't it?
<strong>How is IBM going to get to 2,ghz this year? The 970 caps at 1.8ghz and the 970+ which caps at 2.4ghz won't even be thought of til next year.</strong><hr></blockquote>
The original G4 'capped' at 500 MHz, the ppc 970 is released with an initial range of 1.2 - 1.8 GHz. Those two statements are different.
'Capped' implies that the designers _know_ that the chip will have issues beyond the cap. Like 'this spot will get hot', or 'this part interferes with that part beyond frequency X'.
'Initial introduction range' doesn't mean that at all. The 1.2 chip and the 1.8 chip are presumably exactly the same process to make - the ones that do very well end up as 1.8's. But what happens when people figure out _why_ not all of the 1.2's pass the 1.8's certification? -> Slight changes to the processes _NOT_ the design -> higher speeds.
Now, it is possible that the 970 is 'capped' at 1.8, but I wouldn't put money on it.
It'll be interesting to see what they do. If they stick with a minitower, it'll be damn hard to beat El Capitan at its current level of refinement. But then FW3200 opens up a lot of interesting possibilities, doesn't it? </strong><hr></blockquote>
Well, everything I've seen says it starts at 1.4ghz and top end is 1.8ghz. That makes sense seeing that the top G4 is currently 1.4ghz. Yea, I'd rather have an aluminum case or something, it's not necessarily the el captitan as much as it is it's outside. Just redo the exterior is typical fashion and I'm sure it will rock.
The original G4 'capped' at 500 MHz, the ppc 970 is released with an initial range of 1.2 - 1.8 GHz. Those two statements are different.
'Capped' implies that the designers _know_ that the chip will have issues beyond the cap. Like 'this spot will get hot', or 'this part interferes with that part beyond frequency X'.
'Initial introduction range' doesn't mean that at all. The 1.2 chip and the 1.8 chip are presumably exactly the same process to make - the ones that do very well end up as 1.8's. But what happens when people figure out _why_ not all of the 1.2's pass the 1.8's certification? -> Slight changes to the processes _NOT_ the design -> higher speeds.
Now, it is possible that the 970 is 'capped' at 1.8, but I wouldn't put money on it.</strong><hr></blockquote>
My bad, I used 'capped' too loosely when i meerly meant released ending speed or something like that
There's nothing comming from IBM to suggest that 970 will just stop at 1.8 GHz. Look att Motorola.. they don't seem to acknowledge 7455 over 1 GHz, but Apple just announced 1.42 GHz parts. They've said +1 GHz though.
Just because IBM shows how a 1.2GHz 970 scores and how much wattage it uses, doesn't mean they ever intend to release one at that speed. Hell if they had a pre-production 970 working at 1.2GHz way back when then that is pretty impressive, especially considering AMD had pre-production Hammers running at 800-900MHz when they are expected to be over 2GHz when they are released officially.
<strong>Well, everything I've seen says it starts at 1.4ghz and top end is 1.8ghz.</strong><hr></blockquote>
At the outset, yes. I don't think IBM has ever published the absolute maximum clock speed that the design is capable of.
Maybe it won't get too much farther before the die shrink, just because IBM is planning to move it to .09μ tout suite, but I wouldn't be surprised if it hit 2GHz, and the die-shrunk 970 picked up from there.
The 970, unlike the G4, is designed to scale nicely. It'll be hard to get used to, I know.
At the outset, yes. I don't think IBM has ever published the absolute maximum clock speed that the design is capable of.
Maybe it won't get too much farther before the die shrink, just because IBM is planning to move it to .09μ tout suite, but I wouldn't be surprised if it hit 2GHz, and the die-shrunk 970 picked up from there.
The 970, unlike the G4, is designed to scale nicely. It'll be hard to get used to, I know.
<strong>Hard to get used to, but almost as depressing.
The Pentium 4 is expected to break the 10GHz barrier in 2005.
</strong><hr></blockquote>
So they say. It cracks me up that Intel actually padded the P4's pipeline with no-ops to give the electricity time to propagate across the chip.
But that just illustrates how much they're missing the point, and I think that'll be clear enough by the time 2005 rolls around. They're still basically assuming that PCs will be laid out the way they've always been, with one big CPU at the heart of the machine, but with the motherboard tech coming down the pipe in the next year it's just not going to be like that. I think Apple's long expertise as a systems integrator will come to the fore when high-speed interconnects arrive on the desktop.
Meanwhile, nobody's buying the 3GHz Pentiums because they open Word so much faster than their 2GHz Pentium does. Faced with a 10GHz Pentium... well, it could enable a virus to wipe out your entire machine before you could blink, while it was trying to decide whether or not you really deserved to listen to that music file you just double-clicked on, but somehow I don't see this overcoming the public's rather jaded view of the Pentium's clockspeed. Except for a few pro areas, Windows applications have stalled, and further speed increases are essentially pointless. Even games have reached a point where the video card is more of a bottleneck than the CPU.
The Mac does not have this problem. There are, and will be, plenty of applications to keep an iMac earning its keep, courtesy of Apple.
to stray back to nominal topic of this thread. The 7447/7457. What is the difference betwen the two? Is the new one just a die shrink or does it add features like DDR support?
Meanwhile, nobody's buying the 3GHz Pentiums because they open Word so much faster than their 2GHz Pentium does. Faced with a 10GHz Pentium... well, it could enable a virus to wipe out your entire machine before you could blink, while it was trying to decide whether or not you really deserved to listen to that music file you just double-clicked on, but somehow I don't see this overcoming the public's rather jaded view of the Pentium's clockspeed. Except for a few pro areas, Windows applications have stalled, and further speed increases are essentially pointless.
</strong><hr></blockquote>
Please, please, can you make T-Shirts with that on it? Billboards, bumper stickers, maybe a Matsu Flag with that on it, spam all the mac boards with that, some RFT emails to send out to everyone who is a doom sayer and always posts what Intels plans are next year. WTF cares what they are doing as long as you are doing your work. My dual gig suits me fine, that's why I don't bitch about how slow apple computers are. They are fast enough for my market segment and X is more stable then a 5 leg table.
I agree, my Dual gig DA is fast enough for my needs but faster would be better. I planned on a new purchase when speed increased by 50%, which it has with the new DP 1.4 gig. Too bad they don't have any for 2-months, now I'll wait and see the reviews. But then it will be 3-4 months and maybe then I'll just wait til August for the next revision.
Point being, Apple's non-abillity to deliver on their equipment loses them a near-term sale. Again, I assume it is MOTO's fault for not delivering chips. I can only hope that IBM doesn't lapse into this non-development scenario once (and if) they get onboard Stevie's magic carpet ride..
<strong>to stray back to nominal topic of this thread. The 7447/7457. What is the difference betwen the two? Is the new one just a die shrink or does it add features like DDR support?
TIA</strong><hr></blockquote>
The 7447/7457 appears to be a die-shrunk 7445/7455, perhaps with double the L2 cache, and able to support a slightly faster (but still SDR and shared) MaxBus.
If that's the case, it's an incremental improvement. Given the way the 7455 has scaled, it should get the G4 up to 2GHz.
If it follows the same pattern as before, the 7447 won't have pins for the L3 cache whereas the 7457 will. The difference between the 7447/7457 and the 7445/7455 is that the 57 has a larger L2 cache (512 vs. 256K), a faster bus (200 Mhz vs. 167), and room for 4MB of L3 cache. The 7457-RM adds DDR support.
<strong>The 7457-RM adds DDR support.</strong><hr></blockquote>And 7457-RM is a as far as anyone here knows a proposed design witch will (or won't) make an appearance in 2004.
The state of the rumors in early January was that the 7457 would debut in the powermacs. When that didn't happen, rumors began to build that bastardized 7457s would appear in the iMac. Of course that didn't happen either.
My bet, for which I have NO EVIDENCE is this: In the latter part of March, we'll get new powermacs with a modestly bumped MPX bus and otherwise drop-in 7457s.
An argument against this guess is that Apple JUST bumped their powermacs...why would we expect them to bump them again after 3 months? My response: Apple WANTS a nice 6 month progression, but their hands are tied by what Motorola can deliver. If the 7457 is available come the end of March, Apple will ship it, 6 month product cycles be damned.
Obviously the 7457 wasn't ready. That did surprise me because I was told a new G4+ in Jan. I'm wondering if that has changed (duh) maybe the reasoning behind "year of the laptop". Also could be that Apple will/has spend it's time on getting a new chip (970) into a new case by the fall. Notice apple didn't do anything but bump, and the tower was bumped just a little. I think that could mean the towers ar done with G4s and the iMacs were'nt bumped much because they will get to 1.25+ (and possibly a design tweak) by summer then the towers get the 970 in the fall.
It either is very good news or bad news that we didn't get the G4+ (7457?)
At the outset, yes. I don't think IBM has ever published the absolute maximum clock speed that the design is capable of.
]</strong><hr></blockquote>
Yes, let' s make a comparison with the Athlon 64.
The athlon 64 is designed to start at 1,8 ghz (but samples are still at 1,4 ghz). The chip is as huge as a ppc 970 and has a shorter pipeline 12 instead of 14.
There is no reason to limit the speed of the PPC 970 at 1,8 ghz.
I think they just meant that the first ppc produced will be clocked between 1,4 to 1,8 ghz.
Comments
I will be floored if we get a 2ghz machine this year, and when i pick myself up, I'll be reaching for my credit card. I'm expecting the 1.8ghz 970 in a phat new Ives design (who btw is getting 'stale' according to UK design community) and as any designer knows, that's modivation.
Can't wait till this fall either way.
<strong>How is IBM going to get to 2,ghz this year? The 970 caps at 1.8ghz and the 970+ which caps at 2.4ghz won't even be thought of til next year.</strong><hr></blockquote>
The 970 is scheduled to start with a top speed of 1.8GHz. There's nothing saying that it won't get faster from there.
Then there's the rapid die shrink to .09μ.
If you stretch the "year" to include MWSF '04 the odds are pretty darn good IMO.
[quote]<strong>I'm expecting the 1.8ghz 970 in a phat new Ives design (who btw is getting 'stale' according to UK design community) and as any designer knows, that's modivation.</strong><hr></blockquote>
Yeah, white is so five minutes ago.
It'll be interesting to see what they do. If they stick with a minitower, it'll be damn hard to beat El Capitan at its current level of refinement. But then FW3200 opens up a lot of interesting possibilities, doesn't it?
<strong>How is IBM going to get to 2,ghz this year? The 970 caps at 1.8ghz and the 970+ which caps at 2.4ghz won't even be thought of til next year.</strong><hr></blockquote>
The original G4 'capped' at 500 MHz, the ppc 970 is released with an initial range of 1.2 - 1.8 GHz. Those two statements are different.
'Capped' implies that the designers _know_ that the chip will have issues beyond the cap. Like 'this spot will get hot', or 'this part interferes with that part beyond frequency X'.
'Initial introduction range' doesn't mean that at all. The 1.2 chip and the 1.8 chip are presumably exactly the same process to make - the ones that do very well end up as 1.8's. But what happens when people figure out _why_ not all of the 1.2's pass the 1.8's certification? -> Slight changes to the processes _NOT_ the design -> higher speeds.
Now, it is possible that the 970 is 'capped' at 1.8, but I wouldn't put money on it.
<strong>
Yeah, white is so five minutes ago.
It'll be interesting to see what they do. If they stick with a minitower, it'll be damn hard to beat El Capitan at its current level of refinement. But then FW3200 opens up a lot of interesting possibilities, doesn't it?
Well, everything I've seen says it starts at 1.4ghz and top end is 1.8ghz. That makes sense seeing that the top G4 is currently 1.4ghz. Yea, I'd rather have an aluminum case or something, it's not necessarily the el captitan as much as it is it's outside. Just redo the exterior is typical fashion and I'm sure it will rock.
<strong>
The original G4 'capped' at 500 MHz, the ppc 970 is released with an initial range of 1.2 - 1.8 GHz. Those two statements are different.
'Capped' implies that the designers _know_ that the chip will have issues beyond the cap. Like 'this spot will get hot', or 'this part interferes with that part beyond frequency X'.
'Initial introduction range' doesn't mean that at all. The 1.2 chip and the 1.8 chip are presumably exactly the same process to make - the ones that do very well end up as 1.8's. But what happens when people figure out _why_ not all of the 1.2's pass the 1.8's certification? -> Slight changes to the processes _NOT_ the design -> higher speeds.
Now, it is possible that the 970 is 'capped' at 1.8, but I wouldn't put money on it.</strong><hr></blockquote>
My bad, I used 'capped' too loosely when i meerly meant released ending speed or something like that
[ 01-29-2003: Message edited by: Henriok ]</p>
<strong>Well, everything I've seen says it starts at 1.4ghz and top end is 1.8ghz.</strong><hr></blockquote>
At the outset, yes.
Maybe it won't get too much farther before the die shrink, just because IBM is planning to move it to .09μ tout suite, but I wouldn't be surprised if it hit 2GHz, and the die-shrunk 970 picked up from there.
The 970, unlike the G4, is designed to scale nicely. It'll be hard to get used to, I know.
[ 01-30-2003: Message edited by: Amorph ]</p>
The Pentium 4 is expected to break the 10GHz barrier in 2005.
Barto
<strong>
At the outset, yes.
Maybe it won't get too much farther before the die shrink, just because IBM is planning to move it to .09μ tout suite, but I wouldn't be surprised if it hit 2GHz, and the die-shrunk 970 picked up from there.
The 970, unlike the G4, is designed to scale nicely. It'll be hard to get used to, I know.
[ 01-30-2003: Message edited by: Amorph ]</strong><hr></blockquote>
By the .09 do you mean the 970+? Because that's suppose to scale up to 2.4ghz, starting at 1.8ghz or 2.ghz IIRC (according to NMR
<strong>Hard to get used to, but almost as depressing.
The Pentium 4 is expected to break the 10GHz barrier in 2005.
</strong><hr></blockquote>
So they say. It cracks me up that Intel actually padded the P4's pipeline with no-ops to give the electricity time to propagate across the chip.
But that just illustrates how much they're missing the point, and I think that'll be clear enough by the time 2005 rolls around. They're still basically assuming that PCs will be laid out the way they've always been, with one big CPU at the heart of the machine, but with the motherboard tech coming down the pipe in the next year it's just not going to be like that. I think Apple's long expertise as a systems integrator will come to the fore when high-speed interconnects arrive on the desktop.
Meanwhile, nobody's buying the 3GHz Pentiums because they open Word so much faster than their 2GHz Pentium does. Faced with a 10GHz Pentium... well, it could enable a virus to wipe out your entire machine before you could blink, while it was trying to decide whether or not you really deserved to listen to that music file you just double-clicked on, but somehow I don't see this overcoming the public's rather jaded view of the Pentium's clockspeed. Except for a few pro areas, Windows applications have stalled, and further speed increases are essentially pointless. Even games have reached a point where the video card is more of a bottleneck than the CPU.
The Mac does not have this problem. There are, and will be, plenty of applications to keep an iMac earning its keep, courtesy of Apple.
[ 01-30-2003: Message edited by: Amorph ]</p>
TIA
<strong>
<cut>
Meanwhile, nobody's buying the 3GHz Pentiums because they open Word so much faster than their 2GHz Pentium does. Faced with a 10GHz Pentium... well, it could enable a virus to wipe out your entire machine before you could blink, while it was trying to decide whether or not you really deserved to listen to that music file you just double-clicked on, but somehow I don't see this overcoming the public's rather jaded view of the Pentium's clockspeed. Except for a few pro areas, Windows applications have stalled, and further speed increases are essentially pointless.
</strong><hr></blockquote>
Please, please, can you make T-Shirts with that on it? Billboards, bumper stickers, maybe a Matsu Flag with that on it, spam all the mac boards with that, some RFT emails to send out to everyone who is a doom sayer and always posts what Intels plans are next year. WTF cares what they are doing as long as you are doing your work. My dual gig suits me fine, that's why I don't bitch about how slow apple computers are. They are fast enough for my market segment and X is more stable then a 5 leg table.
[ 01-30-2003: Message edited by: KidRed ]</p>
Point being, Apple's non-abillity to deliver on their equipment loses them a near-term sale. Again, I assume it is MOTO's fault for not delivering chips. I can only hope that IBM doesn't lapse into this non-development scenario once (and if) they get onboard Stevie's magic carpet ride..
<strong>to stray back to nominal topic of this thread. The 7447/7457. What is the difference betwen the two? Is the new one just a die shrink or does it add features like DDR support?
TIA</strong><hr></blockquote>
The 7447/7457 appears to be a die-shrunk 7445/7455, perhaps with double the L2 cache, and able to support a slightly faster (but still SDR and shared) MaxBus.
If that's the case, it's an incremental improvement. Given the way the 7455 has scaled, it should get the G4 up to 2GHz.
<strong>The 7457-RM adds DDR support.</strong><hr></blockquote>And 7457-RM is a as far as anyone here knows a proposed design witch will (or won't) make an appearance in 2004.
What's going on?
We know that Motorola is set to begin pushing the 7457 no later than <a href="http://e-www.motorola.com/webapp/sps/site/overview.jsp?nodeId=02VS0llCc5pzMPsvFLn1b23G5nChPn " target="_blank">March 23, 2002</a>.
My bet, for which I have NO EVIDENCE is this: In the latter part of March, we'll get new powermacs with a modestly bumped MPX bus and otherwise drop-in 7457s.
An argument against this guess is that Apple JUST bumped their powermacs...why would we expect them to bump them again after 3 months? My response: Apple WANTS a nice 6 month progression, but their hands are tied by what Motorola can deliver. If the 7457 is available come the end of March, Apple will ship it, 6 month product cycles be damned.
It either is very good news or bad news that we didn't get the G4+ (7457?)
<strong>
At the outset, yes.
]</strong><hr></blockquote>
Yes, let' s make a comparison with the Athlon 64.
The athlon 64 is designed to start at 1,8 ghz (but samples are still at 1,4 ghz). The chip is as huge as a ppc 970 and has a shorter pipeline 12 instead of 14.
There is no reason to limit the speed of the PPC 970 at 1,8 ghz.
I think they just meant that the first ppc produced will be clocked between 1,4 to 1,8 ghz.