Could someone move at least part of this thread to CURRENT HARDWARE? I think Lemon Bon Bon's displeasure with the CURRENT state of PowerMac offerings shouldn't be in FUTURE HARDWARE for obvious reasons.
We can't split threads up, so instead I'll ask that the topic return to speculation about what Apple will unveil in July, given what we know about the XServe.
Posts about current Apple hardware belong in Current Hardware. Rants about Apple hardware belong in General Discussion.
And really, there only needs to be one of those threads. There is no benefit in hijacking every thread on the boards to mention that you consider current Apple hardware to be inadequate. Even if, for your purposes, it is. <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
It's worth mentioning that The Register's current crop of rumours fit in line with a lot of the stuff some people were hearing long before anybody decided to publish it.
They may not be right about clock speeds but I wouldn't be surprised if they are correct on the chip.
Would there be any benefit on a desktop machine to keeping the 4 ATA/100 busses? I mean lets say the G4 IS still stuck at 1-1.2 Ghz and basically the machine gets an XServe type DDR solution and maybe a minor speed bump (we won't really have good idea one way or the other until maybe a week or so before the show). Could Apple differentiate the new "high-end" machines from the old high-end (and the encroaching consumer Macs) by a built in RAID solution on the top machine? The XServe as I understand it, introduced the ability to be able to boot from a raid. Albeit, this is a software raid but it would still net an appreciable speed boost over a single drive wouldn't it? OS X loves a fast disk, so how much of a speed boost would they realize from a raid? Would that along with (if NMR is right) 10.2 cover for a processor disappointment?
many people aren't going to pay $2999 for a single 1Ghz proc rackmount, when they can purchase a $2299 1.2Ghz with better ddr implementation...
wake up and smell the poo...
the mwny powermacs will be basically identical to the Xserve specs. maybe you can buy a dual 1.2Ghz as the "Ultimate" option, woop-ti-Fricken-doo.
-no 1394b
-no ddr fsb
-no usb 2.0
apple will dress up the powermac with a neat-o tablet device that wows all dorks who fall under the "i-need-a-computer-with-sex-appeal-to-achieve-my-unconscious-sexual-goals" market. not that there is anything wrong with this, just better ways to achieve these goals.
<strong>We can't split threads up, so instead I'll ask that the topic return to speculation about what Apple will unveil in July, given what we know about the XServe.
Posts about current Apple hardware belong in Current Hardware. Rants about Apple hardware belong in General Discussion.
And really, there only needs to be one of those threads. There is no benefit in hijacking every thread on the boards to mention that you consider current Apple hardware to be inadequate. Even if, for your purposes, it is. <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" /> </strong><hr></blockquote>
Nicely put.
BTW: Should Apple implement a faster FSB in July? Yes. Will Apple implment a faster FSB in July? Probably not, cause Apple just seems to be behind in these types of things. Given the Xserve specs, I think we will see the smokescreen DDR Power Macs in July, true REAL DDR Power Macs in January.
BTW: Should Apple implement a faster FSB in July? Yes. Will Apple implment a faster FSB in July? Probably not, cause Apple just seems to be behind in these types of things. Given the Xserve specs, I think we will see the smokescreen DDR Power Macs in July, true REAL DDR Power Macs in January.
Just to be clear: its not up to Apple to implement a DDR FSB. This is a feature of the processor, and that is up to Motorola (or whatever PPC supplier Apple uses).
Here are some of my thoughts/questions on future Macs and PC comparisons to them. First bandwidth to the processor and ram is first and foremost the "performance" gain. If Apple can't put more zero's and ones to the the processor and associated hardware the "2.4" gig that PC are wailing about means bubcus. The relevance of Processor speed stopped when they reached about 800 Mhz. I have found that very very few people in a professional PC user atomosphere can even push a 1 ghz pc processor to what it could do. The thing is even a programmer can't get out of the zero and one counter what it can do if they can not push code to it from the ram and hd. So give us real architecture that will choke a processor and then worry about what Intel is doing that Moto isn't.
Next higher clocks is a product of deeper pipelines in todays market, not of producing a processor that can get things in and out faster. So what happens is, the pipeline gets deeper the code gets split further going in and needs more time to be reassembled coming out. More overhead is lost in the split/recombine area so the percent gain is cut badly.
Now Moto is doing it too to reach the market that, well it and Apple started. They had the fastest a few years ago, anyone remember that? Well it came back to bit them in the arse because they made megahertz the standard. The other guy "cheated" he said ok here's the thing i can get more Mhz and not a whole lot more performance by doing the pipeline thing. Moto and Apple sat looking for a long while at how to scale the processor with the best processing ability to higher clocks, turns out the cheat wasn't so much of a cheat after all it's Really tough to do.
For Macworld NY well i sure hope Apple finds a way to put enough bandwidth through their machine to get all those fps's and quartz affects working. The best would be a 266 fsb and rapid IO. I would consider buying then because i learned when i went from my upgraded Performa 6400 G3/300 to a G4 400 that even though Macbench said i was only getting 100 points more on my new spiffy G4 processor that i was getting more than triple the performance in fps and all. It's because i was getting 230 MB/sec across the bus as compared to 28 MB/sec.
Give me a true 266 fbs and 1.6 Ghz(my deal with my fiance was 4 times the Mhz and i will upgrade and i'll be inline at the apple store.
PS anyone want to clearify DDR for me, it is my understanding that it is just PC 133 running in parallel. Are there twice as many pins to carry the doubled bandwidth or have they been wasting what was already there?
Many of the rumours floating about point to a bumped G4 and DDR memory.
Why not a G4 with Rapid Io set up?
I can't believe it's taken Apple and Motorola this long to implement faster memory.
I didn't really notice that much difference between DDR ram and single data ram. Maybe it's no big deal. And if Moto's primary market isn't DDR? Then are they going to implement it 'just' for Apple?
Given the tech' 'difficulty'(?) of getting the current G4 and mobo to DDR/Mobo...wouldn't it be best to go straight to Rapid Io with a G4 in it?
Like 'Yikes' kinda? (The Book 'E' architecture is the next one from IBM and Moto'. Surely they've not choice but to adopt it when Moto/IBM start making processors that feature 'Io' this year/next year?)
Then drop the G5 in next year when it's ready (if it's not ready now...I still don't believe it's not ready. I'm still hoping Apple will surprise us as they did with the original G4 announcement.)
Didn't Jobs say in the 'Time' iMac story at the start of the year that Apple 'deferred' introducing G5s until later on in the year?
Didn't Jobs say in the 'Time' iMac story at the start of the year that Apple 'deferred' introducing G5s until later on in the year?
Lemon Bon Bon</strong><hr></blockquote>
LOL, Jobs didn't say that. If he had, don't you think it would be quoted in practically every post at AI?
The G5 isn't going to be here at MWNY, no way. We would have heard more rumors about it by now. You can't keep something like that under wraps so close to it's introduction. ...or maybe?
Apple DID manage to keep the iMac secret until it's intro, but that was different in that it was a repackaging of old tech. New tech requires extensive field testing and we would have heard about it by now. And we have. There have been numerous reports of people field testing either G5s or mysterious sealed Mac enclosures that slay AMD Athlons for breakfast. It's very possible that Apple has been field testing the G5 for 8-12 months now. MWNY is certainly a possiblity.
Combined with the migration of all other Apple products to G4s (except the iBook, but the iBook even had a 66 MHz bus when PMs were using 133 MHz buses).
I say the G5 will be here soon. Maybe not MWNY, but soon...and I doubt very much that the Powermacs will be updated with some ultra-lame DDR hack. If the G5 isn't read at MWNY, then we will probably see no Powermac revision at MWNY.
<strong>PS anyone want to clearify DDR for me, it is my understanding that it is just PC 133 running in parallel. Are there twice as many pins to carry the doubled bandwidth or have they been wasting what was already there?</strong><hr></blockquote>
Basically, the individual RAM cells inside a DDR chip (note the difference here!) are about as fast as the ones in a conventional, SDR module (i.e. both are running at 133MHz).
Let's now assume we have a memory chip with 8 data pins.
In the case of SDR, during a data transfer, the content of 8 individual RAM cells (a RAM cell is where one individual bit is stored) would be read from the chip's internal memory array (the cells are organised into 2-dimensional arrays of rows and columns), and then put the 8 bits onto the respective data pins for one clock cycle.
A DDR chip, on the other hand, would not only read 8, but 16 cells at once. Since we still only have 8 data pins, the 16 bits are then time-multiplexed such that half of them are transmitted during the first half-period of the clock signal (i.e. after the rising signal edge), and the other are then transmitted during the second half-period (i.e. after the falling signal edge). That way, we get an effective data transfer speed of twice the clock speed. (Note that row and column address transfers still happen at the original clock speed only!)
Hope this clarifies it a bit. If not, I have some diagrams I could scan that might help illustrate it a little further.
Bye,
RazzFazz
(EDIT: Had some terms mixed up, should be correct now.)
[quote] ...give us real architecture that will choke a processor and then worry about what Intel is doing that Moto isn't. <hr></blockquote>
Precisely. I think we'll see Apple do just that. As the DWAG says,
[quote] RapidIO is on Motorola's roadmap, so it's a good bet that in place of an MPX double pumped DDR frontside bus, the G4 or G5 will get a RapidIO bus. <hr></blockquote>
That would help explain the mysterious Athlon-pounding performers reputed to have circulated in closed-door Hollywood circles.
---------------------------------
Hush, Apollo's here ?
And he can tell how I careered through tireless beds
<strong>why the hell would apple release server hardware with these specs, if they can release a powermac with 100% performance gains a month later?</strong><hr></blockquote>
Because you can stuff a whole lot more performance into a tower than you can into a 1U rack?
If Apple shipped a desktop that seriously outperformed their 1U server at CPU-bound tasks they'd just be at par with everyone else. The XServe will still have a number of material advantages over the tower that are important for its customers.
Processing power is not a primary consideration for servers. I/O bandwidth is.
<strong>why the hell would apple release server hardware with these specs, if they can release a powermac with 100% performance gains a month later?
They wouldn't! scratch rapid i/o cuz it ain't gonna be at mwny. like it was said before, ddr is just a smokescreen to keep the elderly purchasers.</strong><hr></blockquote>
It is pretty normal operating procedure to use the previous generation processor in a rack mounted server... especially if the newer one isn't quite ready yet. More interesting is the move to DDR, and large amounts of it, in the servers. Apple has to buy memory in large amounts to keep its costs down, so would it start shipping just a rack mounted server with DDR? Or will the new desktops start shipping with it too? We'll find out in the next little while....
Interesting. That. (I'm shocked. I didn't realise Apple went in for repackaging 'old tech' at a premium price... Ol' honest Stevie...selling out of date Rage cards for loads of dosh? He...he wouldn't do that...would he?)
"I say the G5 will be here soon. Maybe not MWNY, but soon...and I doubt very much that the Powermacs will be updated with some ultra-lame DDR hack. If the G5 isn't read at MWNY, then we will probably see no Powermac revision at MWNY."
Perhaps they have other fish to fry at Macworld New York? Some weird iPod style gadjets et al?
I wish I knew. Blind faith tells me that we'll see the G5 within the year. Optimistically, I want to believe Steve Jobs will pull off a sensational New York show and stun us all.
Part of me thinks the G5 might be 'readied' behind the scenes this year, ramp up at October and ship for Macworld early 2003.
Maybe we'll see no Powermacs at New York? What about Seybold? G5s previewed in August and shipping Sept/October? That leaves a 2003 San Fran' show too close?
My blind faith gut feeling says G5s will be here soon. Recall Macuser or was it Macworld UK expecting G5s by the year's end.
Comments
Could someone move at least part of this thread to CURRENT HARDWARE? I think Lemon Bon Bon's displeasure with the CURRENT state of PowerMac offerings shouldn't be in FUTURE HARDWARE for obvious reasons.
Just my two cents.
Posts about current Apple hardware belong in Current Hardware. Rants about Apple hardware belong in General Discussion.
And really, there only needs to be one of those threads. There is no benefit in hijacking every thread on the boards to mention that you consider current Apple hardware to be inadequate. Even if, for your purposes, it is. <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
They may not be right about clock speeds but I wouldn't be surprised if they are correct on the chip.
wake up and smell the poo...
the mwny powermacs will be basically identical to the Xserve specs. maybe you can buy a dual 1.2Ghz as the "Ultimate" option, woop-ti-Fricken-doo.
-no 1394b
-no ddr fsb
-no usb 2.0
apple will dress up the powermac with a neat-o tablet device that wows all dorks who fall under the "i-need-a-computer-with-sex-appeal-to-achieve-my-unconscious-sexual-goals" market. not that there is anything wrong with this, just better ways to achieve these goals.
<strong>We can't split threads up, so instead I'll ask that the topic return to speculation about what Apple will unveil in July, given what we know about the XServe.
Posts about current Apple hardware belong in Current Hardware. Rants about Apple hardware belong in General Discussion.
And really, there only needs to be one of those threads. There is no benefit in hijacking every thread on the boards to mention that you consider current Apple hardware to be inadequate. Even if, for your purposes, it is. <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" /> </strong><hr></blockquote>
Nicely put.
BTW: Should Apple implement a faster FSB in July? Yes. Will Apple implment a faster FSB in July? Probably not, cause Apple just seems to be behind in these types of things. Given the Xserve specs, I think we will see the smokescreen DDR Power Macs in July, true REAL DDR Power Macs in January.
[ 05-27-2002: Message edited by: Bodhi ]</p>
<strong>
Nicely put.
BTW: Should Apple implement a faster FSB in July? Yes. Will Apple implment a faster FSB in July? Probably not, cause Apple just seems to be behind in these types of things. Given the Xserve specs, I think we will see the smokescreen DDR Power Macs in July, true REAL DDR Power Macs in January.
[ 05-27-2002: Message edited by: Bodhi ]</strong><hr></blockquote>
Just to be clear: its not up to Apple to implement a DDR FSB. This is a feature of the processor, and that is up to Motorola (or whatever PPC supplier Apple uses).
Next higher clocks is a product of deeper pipelines in todays market, not of producing a processor that can get things in and out faster. So what happens is, the pipeline gets deeper the code gets split further going in and needs more time to be reassembled coming out. More overhead is lost in the split/recombine area so the percent gain is cut badly.
Now Moto is doing it too to reach the market that, well it and Apple started. They had the fastest a few years ago, anyone remember that? Well it came back to bit them in the arse because they made megahertz the standard. The other guy "cheated" he said ok here's the thing i can get more Mhz and not a whole lot more performance by doing the pipeline thing. Moto and Apple sat looking for a long while at how to scale the processor with the best processing ability to higher clocks, turns out the cheat wasn't so much of a cheat after all it's Really tough to do.
For Macworld NY well i sure hope Apple finds a way to put enough bandwidth through their machine to get all those fps's and quartz affects working. The best would be a 266 fsb and rapid IO. I would consider buying then because i learned when i went from my upgraded Performa 6400 G3/300 to a G4 400 that even though Macbench said i was only getting 100 points more on my new spiffy G4 processor that i was getting more than triple the performance in fps and all. It's because i was getting 230 MB/sec across the bus as compared to 28 MB/sec.
Give me a true 266 fbs and 1.6 Ghz(my deal with my fiance was 4 times the Mhz and i will upgrade
PS anyone want to clearify DDR for me, it is my understanding that it is just PC 133 running in parallel. Are there twice as many pins to carry the doubled bandwidth or have they been wasting what was already there?
Why not a G4 with Rapid Io set up?
I can't believe it's taken Apple and Motorola this long to implement faster memory.
I didn't really notice that much difference between DDR ram and single data ram. Maybe it's no big deal.
Given the tech' 'difficulty'(?) of getting the current G4 and mobo to DDR/Mobo...wouldn't it be best to go straight to Rapid Io with a G4 in it?
Like 'Yikes' kinda? (The Book 'E' architecture is the next one from IBM and Moto'. Surely they've not choice but to adopt it when Moto/IBM start making processors that feature 'Io' this year/next year?)
Then drop the G5 in next year when it's ready (if it's not ready now...I still don't believe it's not ready. I'm still hoping Apple will surprise us as they did with the original G4 announcement.)
Didn't Jobs say in the 'Time' iMac story at the start of the year that Apple 'deferred' introducing G5s until later on in the year?
Lemon Bon Bon
<strong>
Didn't Jobs say in the 'Time' iMac story at the start of the year that Apple 'deferred' introducing G5s until later on in the year?
Lemon Bon Bon</strong><hr></blockquote>
LOL, Jobs didn't say that. If he had, don't you think it would be quoted in practically every post at AI?
The G5 isn't going to be here at MWNY, no way. We would have heard more rumors about it by now. You can't keep something like that under wraps so close to it's introduction. ...or maybe?
Apple DID manage to keep the iMac secret until it's intro, but that was different in that it was a repackaging of old tech. New tech requires extensive field testing and we would have heard about it by now. And we have. There have been numerous reports of people field testing either G5s or mysterious sealed Mac enclosures that slay AMD Athlons for breakfast. It's very possible that Apple has been field testing the G5 for 8-12 months now. MWNY is certainly a possiblity.
Combined with the migration of all other Apple products to G4s (except the iBook, but the iBook even had a 66 MHz bus when PMs were using 133 MHz buses).
I say the G5 will be here soon. Maybe not MWNY, but soon...and I doubt very much that the Powermacs will be updated with some ultra-lame DDR hack. If the G5 isn't read at MWNY, then we will probably see no Powermac revision at MWNY.
[ 05-27-2002: Message edited by: Junkyard Dawg ]</p>
<strong>PS anyone want to clearify DDR for me, it is my understanding that it is just PC 133 running in parallel. Are there twice as many pins to carry the doubled bandwidth or have they been wasting what was already there?</strong><hr></blockquote>
Basically, the individual RAM cells inside a DDR chip (note the difference here!) are about as fast as the ones in a conventional, SDR module (i.e. both are running at 133MHz).
Let's now assume we have a memory chip with 8 data pins.
In the case of SDR, during a data transfer, the content of 8 individual RAM cells (a RAM cell is where one individual bit is stored) would be read from the chip's internal memory array (the cells are organised into 2-dimensional arrays of rows and columns), and then put the 8 bits onto the respective data pins for one clock cycle.
A DDR chip, on the other hand, would not only read 8, but 16 cells at once. Since we still only have 8 data pins, the 16 bits are then time-multiplexed such that half of them are transmitted during the first half-period of the clock signal (i.e. after the rising signal edge), and the other are then transmitted during the second half-period (i.e. after the falling signal edge). That way, we get an effective data transfer speed of twice the clock speed. (Note that row and column address transfers still happen at the original clock speed only!)
Hope this clarifies it a bit. If not, I have some diagrams I could scan that might help illustrate it a little further.
Bye,
RazzFazz
(EDIT: Had some terms mixed up, should be correct now.)
[ 05-27-2002: Message edited by: RazzFazz ]</p>
[quote] ...give us real architecture that will choke a processor and then worry about what Intel is doing that Moto isn't. <hr></blockquote>
Precisely. I think we'll see Apple do just that. As the DWAG says,
[quote] RapidIO is on Motorola's roadmap, so it's a good bet that in place of an MPX double pumped DDR frontside bus, the G4 or G5 will get a RapidIO bus. <hr></blockquote>
That would help explain the mysterious Athlon-pounding performers reputed to have circulated in closed-door Hollywood circles.
---------------------------------
Hush, Apollo's here ?
And he can tell how I careered through tireless beds
Desperate flesh to soul to wed
They wouldn't! scratch rapid i/o cuz it ain't gonna be at mwny. like it was said before, ddr is just a smokescreen to keep the elderly purchasers.
<strong>why the hell would apple release server hardware with these specs, if they can release a powermac with 100% performance gains a month later?</strong><hr></blockquote>
Because you can stuff a whole lot more performance into a tower than you can into a 1U rack?
If Apple shipped a desktop that seriously outperformed their 1U server at CPU-bound tasks they'd just be at par with everyone else. The XServe will still have a number of material advantages over the tower that are important for its customers.
Processing power is not a primary consideration for servers. I/O bandwidth is.
[ 05-27-2002: Message edited by: Amorph ]</p>
<strong>why the hell would apple release server hardware with these specs, if they can release a powermac with 100% performance gains a month later?
They wouldn't! scratch rapid i/o cuz it ain't gonna be at mwny. like it was said before, ddr is just a smokescreen to keep the elderly purchasers.</strong><hr></blockquote>
It is pretty normal operating procedure to use the previous generation processor in a rack mounted server... especially if the newer one isn't quite ready yet. More interesting is the move to DDR, and large amounts of it, in the servers. Apple has to buy memory in large amounts to keep its costs down, so would it start shipping just a rack mounted server with DDR? Or will the new desktops start shipping with it too? We'll find out in the next little while....
Interesting. That. (I'm shocked. I didn't realise Apple went in for repackaging 'old tech' at a premium price... Ol' honest Stevie...selling out of date Rage cards for loads of dosh? He...he wouldn't do that...would he?)
Lemon Bon Bon
Perhaps they have other fish to fry at Macworld New York? Some weird iPod style gadjets et al?
I wish I knew. Blind faith tells me that we'll see the G5 within the year. Optimistically, I want to believe Steve Jobs will pull off a sensational New York show and stun us all.
Part of me thinks the G5 might be 'readied' behind the scenes this year, ramp up at October and ship for Macworld early 2003.
Maybe we'll see no Powermacs at New York? What about Seybold? G5s previewed in August and shipping Sept/October? That leaves a 2003 San Fran' show too close?
My blind faith gut feeling says G5s will be here soon. Recall Macuser or was it Macworld UK expecting G5s by the year's end.
Lemon Bon Bon