Uh, no. If I predict that tomorrow it will either be sunny, or cloudy, or cloudy and rainy, or sunny and rainy, and then on the next day it happens to be sunny, that doesn't necessarily mean that my prediction was reliable. The only difference between this example and Dorsal's predictions is that he is extremely well versed in bleeding edge computer tech, so his predictions represent all of the possible outcomes.</strong><hr></blockquote>
I respectfully disagree. I have read for a very long time Dorsal and Dorsal M's posts, and have found a semblence of accuracy beyond what you describe. From my sources (doesn't it seem like we all have sources by now?), Dorsal M's MO seems to be predicting "Sun Version 2" tomorrow, when in fact, it's the same sun we've had for millions of years -- the same sun version 1. Sure, it may have sunspots tomorrow, but that doesn't mean it's a new and different sun.
When the new sun comes along (and the G5 will be bright), we won't be mistaking it for the same old with a few hacks on it.
Well I'll be damned; maybe jcc has a semi-reliable source. I hadn't heard anyting about new TiBooks being released before WWDC, but looking back at his comment a few days ago, he (his "source") called it pretty well compared to most of the half-cocked predictions I read in AI.
Wasn't there some kind of channel-emptying chaos going on with Apple's laptops in the last couple weeks though? I recall reading something about it, so I suppose a guess wouldn't have been that hard.
Hrmm. Maybe if his source could tell us a little bit about Jaguar (things we don't already know) -- things that Steve will present next Monday -- that could go a long way towards shoring up the ole "source credibility factor."
[quote] Well I'll be damned; maybe jcc has a semi-reliable source. I hadn't heard anyting about new TiBooks being released before WWDC, but looking back at his comment a few days ago, he (his "source") called it pretty well compared to most of the half-cocked predictions I read in AI.<hr></blockquote>
Not being a smart ass, but what do those G5 / Book e code snippets above prove, if anything? I still haven't heard of anyone spotting an *actual* 8540 for sale in the embedded channels, much less a 7500/8500. I still think it's vaporware until someone has *seen* the chip or some variant of it, being used in the real world.
You have provide "Some 64 bit PowerPC Book E test code (randomly taken out of test routines):"
Disclaimer: I really don't know what I'm writing about so feel free to correct me.
I don't know what this means, but aren't the IBM Power 4,5,6 64 bit? And don't anyone get their shorts in a bind, I'm not intimating that a Power 4,5,6 will be used in any desktop computers for Apple. But could Apple be setting up for Mac OS X to be ported to IBM's Power Series?
It is also interesting that the Power 4 has:
Open standards support
Java Transaction API (JTA)(isn't Apple pushing Java?)
Apache Web Caching Accelerator
Didn't I read somewhere that OS X server comes with?
<strong>Not being a smart ass, but what do those G5 / Book e code snippets above prove, if anything? I still haven't heard of anyone spotting an *actual* 8540 for sale in the embedded channels, much less a 7500/8500. I still think it's vaporware until someone has *seen* the chip or some variant of it, being used in the real world.</strong><hr></blockquote>
There are compilers for 64 bit code already, they mostly support the PowerPC 620 / 625 / 630 / 64 series as well as the POWER series (and others). Those processors are / were 64 bit PowerPC processors - but they do not all fulfill the Book E criteria.
Some compilers and assemblers have been updated to generate Book E compliant code. The code examples I posted have been generated using such software.
Of course this does not prove that a G5 exists, and I never said that I have seen a PowerPC 7500 so far. As far as I know the only G5 currently available (I'm talking about Motorola G5) is the PowerPC 8540 - not the real hardware, but the software simulation of it. I haven't seen samples yet.
The only thing I said was that I now have Book E / G5 documentation, and that's all I can talk about. Of course the fact that there is a documentation does not prove that there is a real hardware too.
[quote]I don't know what this means, but aren't the IBM Power 4,5,6 64 bit?<hr></blockquote>
Correct, they are (like their precessors). The Power4 has been available for some time now, the successors are scheduled for 2004 and 2006.
[quote]And don't anyone get their shorts in a bind, I'm not intimating that a Power 4,5,6 will be used in any desktop computers for Apple.<hr></blockquote>
Apple could use the Power4, but you have to keep in mind that this processor is very expensive, produces a lot of heat and does not support AltiVec.
[quote]But could Apple be setting up for Mac OS X to be ported to IBM's Power Series?<hr></blockquote>
Of course they could, but who would buy the machine? Typical MacOS software wouldn't take much profit of the Power4 design.
On the other hand the Power5 and Power6 processors will be designed to be much cheaper and to produce much less heat, at the same time they will offer even better performance than the Power4. This could be an alternative for Apple, although - as far as I know - IBM does not plan to implement AltiVec in the POWER series.
"Sorry, but even the old 500 MHz 7400's AltiVec units are easily capable of saturating PC2100. Plus Apple now sells dual CPU machines- if one CPU is capable of saturating a PC133 rated bus, then two are obviously crippled on it.
DDR is long overdue, as is adoption of the faster ATA flavors (either ATA100 or ATA133), hardware support for 5.1 sound (DTS)- a natural complement for DVDs, built-in sound in or a reasonably priced PCI-based sound solution, FireWire 2 (800 mbps instead of 400), USB 2, dual CPUs throughout the "pro" line, and a built-in hardware IDE RAID controller.
There's a lot that needs to be done, and a valid question is why it's taking so long. Most of the clone x86 chipset makers (Via, SiS, ALi, even nVidia) already have much of the above built in, and their research budgets are often considerably less than Apple's.
We pay a premium- if we're buying the "BMW" of the computer world, why doesn't the high-end offer more features and outperform competitors, the way high-end BMW's do?
<strong>"Sorry, but even the old 500 MHz 7400's AltiVec units are easily capable of saturating PC2100. Plus Apple now sells dual CPU machines- if one CPU is capable of saturating a PC133 rated bus, then two are obviously crippled on it.
DDR is long overdue, as is adoption of the faster ATA flavors (either ATA100 or ATA133), hardware support for 5.1 sound (DTS)- a natural complement for DVDs, built-in sound in or a reasonably priced PCI-based sound solution, FireWire 2 (800 mbps instead of 400), USB 2, dual CPUs throughout the "pro" line, and a built-in hardware IDE RAID controller.
There's a lot that needs to be done, and a valid question is why it's taking so long. Most of the clone x86 chipset makers (Via, SiS, ALi, even nVidia) already have much of the above built in, and their research budgets are often considerably less than Apple's.
We pay a premium- if we're buying the "BMW" of the computer world, why doesn't the high-end offer more features and outperform competitors, the way high-end BMW's do?
Typical MacOS software wouldn't take much profit of the Power4 design.
</strong><hr></blockquote>
I was more referring to Mac OS X Server. Since Mac OS X Server is Unix based wouldn't most Unix software run and if not, couldn't they be recompiled to?
I guess my question is, the 64 bit code you refer to might not be targeted at a 64 bit G5 for desktops, but rather a move on Apple's part to move into the server market. Feasable, desirable, I don't know. Hence, my stumbling questions.
Apple does sell Mac OS X Server, how serious are they about this market? If they are serious, wouldn't the ability to run it on IBM machines be an advantage for Apple?
<strong>Well I'll be damned; maybe jcc has a semi-reliable source. I hadn't heard anyting about new TiBooks being released before WWDC, but looking back at his comment a few days ago, he (his "source") called it pretty well compared to most of the half-cocked predictions I read in AI.</strong><hr></blockquote>No, he was totally wrong. The other sites had predicted 800Mhz top-end, same mobo. The only unique features his source predicted, DDR and 867Mhz, turned out to be wrong.
We pay a premium- if we're buying the "BMW" of the computer world, why doesn't the high-end offer more features and outperform competitors, the way high-end BMW's do?
<hr></blockquote>
Because Apple doesn't make the "BMWs" of the computer world. They make the computer equivalent of a Pontiac Trans Am with a 4 cylinder engine, drum brakes, and a 3 speed automatic. Glam on the outside, antiquated components on the inside.
And Apple doesn't have any choice. They must maintain exorbitant margins because their volume is relatively low and they have to fund all of their own R&D. The only way out for Apple is to increase their market share, so they can sell a larger volume and thus lower their profit margins.
[quote] DDR is long overdue, as is adoption of the faster ATA flavors (either ATA100 or ATA133), hardware support for 5.1 sound (DTS)- a natural complement for DVDs, built-in sound in or a reasonably priced PCI-based sound solution, FireWire 2 (800 mbps instead of 400), USB 2, dual CPUs throughout the "pro" line, and a built-in hardware IDE RAID controller. <hr></blockquote>
You know, these are good points- and the responsibility for not implementing a lot of them lie at Apple's feet. Uni-N has been around for several years now- companies like VIA don't seem to have problems bringing along new chipsets every year. I understand that Uni-N in particular has consistently been improved upon, but still, where is its replacement?
One would think that especially in the pro line, it's in Apple's best interest to keep throwing in additional reasons to spend the money on the almost $3k dual GHz machine- even if profit margin is squeezed, in absolute terms, they make the most money on that model. Why not throw in an integrated hardware IDE RAID, just like the x86 world? IMNSHO, I think that full 5.1 sound support in hardware would be a killer reason to upgrade from my beige G3- why can't I use my PowerMac as the center of my home theater?
Perhaps we need to send in letters to convince S. Jobs and Co. that this is where Apple needs to be pushing the Pro line along...
Comments
<strong>
Uh, no. If I predict that tomorrow it will either be sunny, or cloudy, or cloudy and rainy, or sunny and rainy, and then on the next day it happens to be sunny, that doesn't necessarily mean that my prediction was reliable. The only difference between this example and Dorsal's predictions is that he is extremely well versed in bleeding edge computer tech, so his predictions represent all of the possible outcomes.</strong><hr></blockquote>
I respectfully disagree. I have read for a very long time Dorsal and Dorsal M's posts, and have found a semblence of accuracy beyond what you describe. From my sources (doesn't it seem like we all have sources by now?), Dorsal M's MO seems to be predicting "Sun Version 2" tomorrow, when in fact, it's the same sun we've had for millions of years -- the same sun version 1. Sure, it may have sunspots tomorrow, but that doesn't mean it's a new and different sun.
When the new sun comes along (and the G5 will be bright), we won't be mistaking it for the same old with a few hacks on it.
edit -- typed a ")" when I meant a """.
[ 04-27-2002: Message edited by: Chrys Robyn ]</p>
<strong>The source I reference does indeed infer that the processor is a newer G4, capable of native DDR support.</strong><hr></blockquote>
it better be a newer version right? so that it runs cooler that past 800 mHz G4s.
(this was posted by someone who don't know alot about prossesors so sorry if its stupid)
-Owl
Wasn't there some kind of channel-emptying chaos going on with Apple's laptops in the last couple weeks though? I recall reading something about it, so I suppose a guess wouldn't have been that hard.
Hrmm. Maybe if his source could tell us a little bit about Jaguar (things we don't already know) -- things that Steve will present next Monday -- that could go a long way towards shoring up the ole "source credibility factor."
There is a subset of Book E instructions that are considered restricted only to 64-
bit Book E processing. A 32-bit Book E implementation need not implement any of
the following instructions. Likewise, neither should 32-bit Book E applications
utilize any of these instructions. All other Book E instructions shall either be supported
directly by the implementation, or sufficient infrastructure will be provided
to enable software emulation of the instructions.
64-bit integer arithmetic, compare, shift and rotate instructions
adde64[o], addme64[o], addze64[o],
subfe64[o], subfme64[o], subfze64[o],
mulhd, mulhdu, mulld[o], divd, divdu, extsw,
cmp (L=1), cmpi (L=1), cmpl (L=1), cmpli (L=1),
rldcl, rldcr, rldic, rldicl, rldicr, rldimi, sld, srad, sradi, srd,
cntlzd, td, tdi
64-bit extended addressing branch instructions
bcctre[l], bce[l][a], bclre[l], be[l][a]
64-bit extended addressing cache management instructions
dcbae, dcbfe, dcbie, dcbste, dcbte, dcbtste, dcbze, icbie, icbte
64-bit extended addressing load instructions
lbze, lbzue, lbzxe, lbzxue, ldarxe, lde, ldue, ldxe, ldxue, lfde, lfdue, lfdxe,
lfdxue, lfse, lfsue, lfsxe, lfsxue, lhae, lhaue, lhaxe, lhaxue, lhbrxe, lhze,
lhzue, lhzxe, lhzxue, lwarxe, lwbrxe, lwze, lwzue, lwzxe, lwzxue
64-bit extended addressing store instructions
stbe, stbue, stbxe, stbxue, stdcxe., stde, stdue, stdxe, stdxue, stfde,
stfdue, stfdxe, stfdxue, stfiwxe, stfse, stfsue, stfsxe, stfsxue, sthbrxe,
sthe, sthue, sthxe, sthxue, stwbrxe, stwcxe., stwe, stwue, stwxe, stwxue
+ #as: -a64 -mppc64 -mbooke64
+ #objdump: -Dr -Mbooke64
+ #name: xcoff64 BookE tests
+
+ .*: file format aixcoff64-rs6000
+
+ Disassembly of section .text:
+
+ 0000000000000000 <.text>:
+ 0
+ 4
+ 8
+ c
+ 10
+ \t\t\t10: R_BA\t.text
+ 14
+ \t\t\t14: R_BA\t.text
+ 18
+ 1c
+ 20
+ 24
+ 28
+ 2c
+ 30
+ # Motorola PowerPC BookE tests
+ #as: -a64 -mppc64 -mbooke64
+ \t.machine\t"ppc64"
+ \t.csect .text[PR]
+ \t.csect main[DS]
+ main:
+ \t.csect .text[PR]
+ .main:
+ \ttlbre 1, 2, 7
+ \ttlbwe 5, 30, 3
+ \tbce\t1, 5, branch_target_1
+ \tbcel\t2, 6, branch_target_2
+ \tbcea\t3, 7, branch_target_3
+ \tbcela\t4, 8, branch_target_4
Hmmm.
<a href="http://www.powerpage.org/story.lasso?newsID=9272"; target="_blank">O'Grady</a> seems to have beaten him to it and been more detailed and been more accurate.
The specs were out well before my post, and , in fact, I thought the DDR part was the interesting bit. That turned out to be untrue.
My source is/was a friend who posted info on another board, which I noticed.
I'm actually disappointed. That new MoBo and it's faster bus had better show up soon.
Sorry for not getting the specs perfect. Guess both I and my source were hoping for too much.
<strong>JBL is right,
The specs were out well before my post, and , in fact, I thought the DDR part was the interesting bit. That turned out to be untrue.
My source is/was a friend who posted info on another board, which I noticed.
I'm actually disappointed. That new MoBo and it's faster bus had better show up soon.
Sorry for not getting the specs perfect. Guess both I and my source were hoping for too much.</strong><hr></blockquote>
isn't the 1mb cache ddr sram like other Mot chips?
You have provide "Some 64 bit PowerPC Book E test code (randomly taken out of test routines):"
Disclaimer: I really don't know what I'm writing about so feel free to correct me.
I don't know what this means, but aren't the IBM Power 4,5,6 64 bit? And don't anyone get their shorts in a bind, I'm not intimating that a Power 4,5,6 will be used in any desktop computers for Apple. But could Apple be setting up for Mac OS X to be ported to IBM's Power Series?
It is also interesting that the Power 4 has:
Open standards support
Java Transaction API (JTA)(isn't Apple pushing Java?)
Apache Web Caching Accelerator
Didn't I read somewhere that OS X server comes with?
PDF output distribution integrated
Standard API acciss to PDF & e-mail funciton
Isn't Mac OS X using PDF standards?
<strong>Not being a smart ass, but what do those G5 / Book e code snippets above prove, if anything? I still haven't heard of anyone spotting an *actual* 8540 for sale in the embedded channels, much less a 7500/8500. I still think it's vaporware until someone has *seen* the chip or some variant of it, being used in the real world.</strong><hr></blockquote>
There are compilers for 64 bit code already, they mostly support the PowerPC 620 / 625 / 630 / 64 series as well as the POWER series (and others). Those processors are / were 64 bit PowerPC processors - but they do not all fulfill the Book E criteria.
Some compilers and assemblers have been updated to generate Book E compliant code. The code examples I posted have been generated using such software.
Of course this does not prove that a G5 exists, and I never said that I have seen a PowerPC 7500 so far. As far as I know the only G5 currently available (I'm talking about Motorola G5) is the PowerPC 8540 - not the real hardware, but the software simulation of it. I haven't seen samples yet.
The only thing I said was that I now have Book E / G5 documentation, and that's all I can talk about. Of course the fact that there is a documentation does not prove that there is a real hardware too.
Correct, they are (like their precessors). The Power4 has been available for some time now, the successors are scheduled for 2004 and 2006.
[quote]And don't anyone get their shorts in a bind, I'm not intimating that a Power 4,5,6 will be used in any desktop computers for Apple.<hr></blockquote>
Apple could use the Power4, but you have to keep in mind that this processor is very expensive, produces a lot of heat and does not support AltiVec.
[quote]But could Apple be setting up for Mac OS X to be ported to IBM's Power Series?<hr></blockquote>
Of course they could, but who would buy the machine? Typical MacOS software wouldn't take much profit of the Power4 design.
On the other hand the Power5 and Power6 processors will be designed to be much cheaper and to produce much less heat, at the same time they will offer even better performance than the Power4. This could be an alternative for Apple, although - as far as I know - IBM does not plan to implement AltiVec in the POWER series.
DDR is long overdue, as is adoption of the faster ATA flavors (either ATA100 or ATA133), hardware support for 5.1 sound (DTS)- a natural complement for DVDs, built-in sound in or a reasonably priced PCI-based sound solution, FireWire 2 (800 mbps instead of 400), USB 2, dual CPUs throughout the "pro" line, and a built-in hardware IDE RAID controller.
There's a lot that needs to be done, and a valid question is why it's taking so long. Most of the clone x86 chipset makers (Via, SiS, ALi, even nVidia) already have much of the above built in, and their research budgets are often considerably less than Apple's.
We pay a premium- if we're buying the "BMW" of the computer world, why doesn't the high-end offer more features and outperform competitors, the way high-end BMW's do?
My $.02 as always,
-Natebrau "
From Macedition reply to article board.
Lemon Bon Bon <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
Nuff said.
<strong>"Sorry, but even the old 500 MHz 7400's AltiVec units are easily capable of saturating PC2100. Plus Apple now sells dual CPU machines- if one CPU is capable of saturating a PC133 rated bus, then two are obviously crippled on it.
DDR is long overdue, as is adoption of the faster ATA flavors (either ATA100 or ATA133), hardware support for 5.1 sound (DTS)- a natural complement for DVDs, built-in sound in or a reasonably priced PCI-based sound solution, FireWire 2 (800 mbps instead of 400), USB 2, dual CPUs throughout the "pro" line, and a built-in hardware IDE RAID controller.
There's a lot that needs to be done, and a valid question is why it's taking so long. Most of the clone x86 chipset makers (Via, SiS, ALi, even nVidia) already have much of the above built in, and their research budgets are often considerably less than Apple's.
We pay a premium- if we're buying the "BMW" of the computer world, why doesn't the high-end offer more features and outperform competitors, the way high-end BMW's do?
My $.02 as always,
-Natebrau "
From Macedition reply to article board.
Lemon Bon Bon <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
Nuff said.</strong><hr></blockquote>
Do high-end BMW's have usb 2.0, nope they don't need it nor do they need ATA 133.
<strong>
Typical MacOS software wouldn't take much profit of the Power4 design.
</strong><hr></blockquote>
I was more referring to Mac OS X Server. Since Mac OS X Server is Unix based wouldn't most Unix software run and if not, couldn't they be recompiled to?
I guess my question is, the 64 bit code you refer to might not be targeted at a 64 bit G5 for desktops, but rather a move on Apple's part to move into the server market. Feasable, desirable, I don't know. Hence, my stumbling questions.
Apple does sell Mac OS X Server, how serious are they about this market? If they are serious, wouldn't the ability to run it on IBM machines be an advantage for Apple?
<strong>Well I'll be damned; maybe jcc has a semi-reliable source. I hadn't heard anyting about new TiBooks being released before WWDC, but looking back at his comment a few days ago, he (his "source") called it pretty well compared to most of the half-cocked predictions I read in AI.</strong><hr></blockquote>No, he was totally wrong. The other sites had predicted 800Mhz top-end, same mobo. The only unique features his source predicted, DDR and 867Mhz, turned out to be wrong.
We pay a premium- if we're buying the "BMW" of the computer world, why doesn't the high-end offer more features and outperform competitors, the way high-end BMW's do?
<hr></blockquote>
Because Apple doesn't make the "BMWs" of the computer world. They make the computer equivalent of a Pontiac Trans Am with a 4 cylinder engine, drum brakes, and a 3 speed automatic. Glam on the outside, antiquated components on the inside.
And Apple doesn't have any choice. They must maintain exorbitant margins because their volume is relatively low and they have to fund all of their own R&D. The only way out for Apple is to increase their market share, so they can sell a larger volume and thus lower their profit margins.
You know, these are good points- and the responsibility for not implementing a lot of them lie at Apple's feet. Uni-N has been around for several years now- companies like VIA don't seem to have problems bringing along new chipsets every year. I understand that Uni-N in particular has consistently been improved upon, but still, where is its replacement?
One would think that especially in the pro line, it's in Apple's best interest to keep throwing in additional reasons to spend the money on the almost $3k dual GHz machine- even if profit margin is squeezed, in absolute terms, they make the most money on that model. Why not throw in an integrated hardware IDE RAID, just like the x86 world? IMNSHO, I think that full 5.1 sound support in hardware would be a killer reason to upgrade from my beige G3- why can't I use my PowerMac as the center of my home theater?
Perhaps we need to send in letters to convince S. Jobs and Co. that this is where Apple needs to be pushing the Pro line along...
-HOS