Hi. I'll try and answer some questions. it's not often I get to come back and talk some more so soon. Firstly, I cannot give you a concrete schedule on Apple's release dates. In the next month or so, is too soon but this summer is a likely release date. Unless you work for Apple you really don't have this knowledge (unless you work for an OEM graphics company like ATI or NVIDIA).
The case. As I stated we received the hardware in a G4 case. It is silver and light grey and it contains an equal amount of drive bays. But this is a common practice when releasing such hardware. Sometimes we get then in bare metal boxes that are nothing more than cheap PC cases with no front bezel. And late in development we might get a case that will resemble the finished case very closely.
PCI slots. They look like the slots in the G4 PowerMacs. This is also a PCI-X standard slot.
Firewire ports remain at 2 and are the same as the ones used in a IEEE1394a implimentation. Firewire comes in many different flavors. Apple's modern machines support 100, 200, and 400Mbps speeds while the new implimentation can scale to 1600Mbps over shielded copper. For faster than 1600Mb speeds you will need to transfer over fiber. This is not a design aspect for these machines however. But running a 400MBps device on a 1600Mbps Firewire bus will not result in faster performance. Firewire to IDE bridge chips will need to be redesigned to take advantage. But at 200MBps, even ATA/100 will be hard pressed to max out IEEE1394b.
<strong>Unless you work for Apple you really don't have this knowledge (unless you work for an OEM graphics company like ATI or NVIDIA).</strong><hr></blockquote>
By telling us all these things don't you worry that either ATI or Nvidia will lose alot of money when steve finds out about someone possibiliy from those company leaked some info about the next powermac ?
test 3 rack-mount PPC boxes. sealed. but inside contains some monitoring hardware. Can set/simulate RAM size, manipulate bus and CPU speed, measure temp and power drain.
2 boxes have same graphics, 1 has different. All use DDR. max 1-2GB. CPU heat and power is very very low. benchmark is close to 867-933 G4 on low machine, much faster on top. Can set PPC speed ~60% more than bottom machine. USB is 1.1 spec. One machine use 1Gb ethernet, others don't use. AGP is same speed all machines. One machine test with 4200rpm HDD, other 2 test only with 5400.
Firewire speed is faster setting for 2 and 4 times now using firewire. One machine use 2 firewire, other machine use one only. People says we have slow machine only test, very very fast machine no people can use, very hot, but many extra for PCI, RAM, and CPU.
<strong>...They took 7500 off the roadmap. To me this either means the G5 used by Apple will indeed be a 85XX or that the 7500 (the original name) will be an Apple specific part number that will not be available for outside buyers. Therefore no reason to include it on a roadmap.</strong><hr></blockquote>
Ha Ha Ha!
What kind of logic is that? No need to even mention something on roadmaps, since apple is the client. Apple can wait. They used to do that.
Remember that on Moto's roadmap, the G5/85xx gets an integrated DDR controller, so adopting "local" DDR SDRAM (in place of L3 cache as you point out) really isn't out of line at all.
Plus, what gets more interesting is, how will Apple implement multiple processors in this situation?
On other boards (which will remain nameless for the moment), there's been some speculation about going to a MP design where you'll have (say) multiple daughterboards, each daughterboard with its own DDR . The connection between daughterboard is via RapidIO (a nice switched bus- why make the bus switched if you're never going to need it?). Then the question is, now that you've got loosely coupled CPUs, what sort of a model do you present?
More speculation on this led to a non-uniform memory access model, where each CPU has local memory, but all memory is addressable by all CPUs- it's just that the latency out to non-local memory is higher.
This should avoid wierd cache coherency schemes since all memory is addressable to all CPUs- no need to reserve some memory per CPU.
Remember that on Moto's roadmap, the G5/85xx gets an integrated DDR controller, so adopting "local" DDR SDRAM (in place of L3 cache as you point out) really isn't out of line at all.
Plus, what gets more interesting is, how will Apple implement multiple processors in this situation?
On other boards (which will remain nameless for the moment), there's been some speculation about going to a MP design where you'll have (say) multiple daughterboards, each daughterboard with its own DDR . The connection between daughterboard is via RapidIO (a nice switched bus- why make the bus switched if you're never going to need it?). Then the question is, now that you've got loosely coupled CPUs, what sort of a model do you present?</strong><hr></blockquote>
Well, NeXTStep had some pretty buff clustering capabilities as I recall, and the whole point of clustering is that multiple machines act like one big machine from the point of view of the software...
<strong>Yeah! Dorsal's back! Can Kormac be far behind? </strong><hr></blockquote>
Let that last keynote serve as yet another pin through the eye of that great mass of ignorant believers out there. We were all supposed to be using pocketable super-computers nearly 2 years ago. At this rate Apple will be lucky if it can supply half the performance of a PocketPC in a package 10 times the size and 5 times the price. Soon PalmOS geeks will be gloating about how their palms out-perform Powermacs in 3-D.
But don't worry, I'm sure Stevie-boy will make it all better with a nice farcical price increase.
Now that the iMac is priced to compete head to head with high-end 'Nix workstations, the market share is going to fall into Apple's domain like a landslide.
<strong>And if you can remember what I was talked about, I think we will see good sign soon.</strong><hr></blockquote>
Honestly I don't remember except something about DNA you talked about since like 1999. But since 2001 Apple already changed the strategy from a computer maker to consumer electronics maker, so I am not sure if those things you said in 1999 still holds even you got inside info at 1999.
But I do believe that G5 will be coming before the end of 2002.
I don't see how anything I posted is dumb, maybe a little too dramatic, sure. Anything Kormac has talked about has not come to pass and never will. My analysis is at least as in tune with reality as his.
That price increase is just plain contemptuous of loyal Apple customers. Before, Apple was falling behind by moving slowly on spec and standing pat on price, now they're falling further behind by standing pat on spec while raising prices.
I sure hope this is temporary. I think that it is, but it's bad press nonetheless. In an industry that isn't exactly fair to Apple they're going to get lambasted for this price increase. Margins are certainly important, but you simply cannot compete without lower priced models. Those don't have to be low margin models, but they need to get affordable to larger audience if they are truly serious about market share. Anyone with a long term business investment in a platform needs to be concerned about market share too. Timely updates of software, third party support, peripheral availibility, etc depends on that.
Why isn't there a 500 USD computer in the Apple line-up? Many people simply WILL NOT buy a computer above that price point. If they can't build the CRT iMac down to that price, then they need to re-engineer it so that they can. Consumers, education buyers, and businesses who vote with their pocketbooks, have been voting against Apple's rigid product structures. IF you need to strip down a computers spec a bit to get a cheaper option, then do it. This isn't like selling a high-end car, or a watch. When some-one buys a computer, they begin making a software investment in that platform. Eventually, the prospect of switching just becomes too expensive/inconvenient, even when people would really like to switch. A sale lost today isn't neccessarily a sale you can get back on the next upgrade cycle. It is even more important to bring people into a platform on the entry level.
Apple's software answer has been brilliant. And it's professional hardware answer adequate. But it has no complete consumer hardware answer. Recent moves only take Apple further out of the low-end market, which incidentally is the only market that can grow your market-share. While it isn't catastrophic to raise prices, it is pretty bad. Tied into a recent trend of sloth like hardware advancement, it just looks that much worse.
People don't really mind paying more, but they want to get more when they do. Take the iPod and HD display. No problem with those. They're clearly high-end products, with high, but very competitive prices. iMac is supposed to be a consumer machine, and as such it is approaching the wrong end of the consumer price range. They need to do better.
Comments
The case. As I stated we received the hardware in a G4 case. It is silver and light grey and it contains an equal amount of drive bays. But this is a common practice when releasing such hardware. Sometimes we get then in bare metal boxes that are nothing more than cheap PC cases with no front bezel. And late in development we might get a case that will resemble the finished case very closely.
PCI slots. They look like the slots in the G4 PowerMacs. This is also a PCI-X standard slot.
Firewire ports remain at 2 and are the same as the ones used in a IEEE1394a implimentation. Firewire comes in many different flavors. Apple's modern machines support 100, 200, and 400Mbps speeds while the new implimentation can scale to 1600Mbps over shielded copper. For faster than 1600Mb speeds you will need to transfer over fiber. This is not a design aspect for these machines however. But running a 400MBps device on a 1600Mbps Firewire bus will not result in faster performance. Firewire to IDE bridge chips will need to be redesigned to take advantage. But at 200MBps, even ATA/100 will be hard pressed to max out IEEE1394b.
<strong>Unless you work for Apple you really don't have this knowledge (unless you work for an OEM graphics company like ATI or NVIDIA).</strong><hr></blockquote>
By telling us all these things don't you worry that either ATI or Nvidia will lose alot of money when steve finds out about someone possibiliy from those company leaked some info about the next powermac ?
test 3 rack-mount PPC boxes. sealed. but inside contains some monitoring hardware. Can set/simulate RAM size, manipulate bus and CPU speed, measure temp and power drain.
2 boxes have same graphics, 1 has different. All use DDR. max 1-2GB. CPU heat and power is very very low. benchmark is close to 867-933 G4 on low machine, much faster on top. Can set PPC speed ~60% more than bottom machine. USB is 1.1 spec. One machine use 1Gb ethernet, others don't use. AGP is same speed all machines. One machine test with 4200rpm HDD, other 2 test only with 5400.
Firewire speed is faster setting for 2 and 4 times now using firewire. One machine use 2 firewire, other machine use one only. People says we have slow machine only test, very very fast machine no people can use, very hot, but many extra for PCI, RAM, and CPU.
<strong>...They took 7500 off the roadmap. To me this either means the G5 used by Apple will indeed be a 85XX or that the 7500 (the original name) will be an Apple specific part number that will not be available for outside buyers. Therefore no reason to include it on a roadmap.</strong><hr></blockquote>
Ha Ha Ha!
What kind of logic is that? No need to even mention something on roadmaps, since apple is the client. Apple can wait. They used to do that.
<img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
Remember that on Moto's roadmap, the G5/85xx gets an integrated DDR controller, so adopting "local" DDR SDRAM (in place of L3 cache as you point out) really isn't out of line at all.
Plus, what gets more interesting is, how will Apple implement multiple processors in this situation?
On other boards (which will remain nameless for the moment), there's been some speculation about going to a MP design where you'll have (say) multiple daughterboards, each daughterboard with its own DDR . The connection between daughterboard is via RapidIO (a nice switched bus- why make the bus switched if you're never going to need it?). Then the question is, now that you've got loosely coupled CPUs, what sort of a model do you present?
More speculation on this led to a non-uniform memory access model, where each CPU has local memory, but all memory is addressable by all CPUs- it's just that the latency out to non-local memory is higher.
This should avoid wierd cache coherency schemes since all memory is addressable to all CPUs- no need to reserve some memory per CPU.
Interesting, no?
-HOS
<strong>Amorph-
Remember that on Moto's roadmap, the G5/85xx gets an integrated DDR controller, so adopting "local" DDR SDRAM (in place of L3 cache as you point out) really isn't out of line at all.
Plus, what gets more interesting is, how will Apple implement multiple processors in this situation?
On other boards (which will remain nameless for the moment), there's been some speculation about going to a MP design where you'll have (say) multiple daughterboards, each daughterboard with its own DDR . The connection between daughterboard is via RapidIO (a nice switched bus- why make the bus switched if you're never going to need it?). Then the question is, now that you've got loosely coupled CPUs, what sort of a model do you present?</strong><hr></blockquote>
Well, NeXTStep had some pretty buff clustering capabilities as I recall, and the whole point of clustering is that multiple machines act like one big machine from the point of view of the software...
This is all very interesting.
<strong>Yeah! Dorsal's back! Can Kormac be far behind?
Let that last keynote serve as yet another pin through the eye of that great mass of ignorant believers out there. We were all supposed to be using pocketable super-computers nearly 2 years ago. At this rate Apple will be lucky if it can supply half the performance of a PocketPC in a package 10 times the size and 5 times the price. Soon PalmOS geeks will be gloating about how their palms out-perform Powermacs in 3-D.
But don't worry, I'm sure Stevie-boy will make it all better with a nice farcical price increase.
Now that the iMac is priced to compete head to head with high-end 'Nix workstations, the market share is going to fall into Apple's domain like a landslide.
<strong>I'm sorry Matsu, but that's the dumbest post I think I've ever read.</strong><hr></blockquote>
I'm not too sure, he's got hot competition from a host of other morons on this forum (no names... yet).
However, the most recent Matsu dribble is quite worthy of Most Moronic Post, I really can't decide. Choices, choices
i believe that what were are going to get at macworld newyork IS the real G5!
i think it exist and is ready to go.
i dont think the 7500 is the G5 but a super-G4!
it has a lot of the capabilities of the g4 but it wont be able to keep up with the G5.
i dont even think the 7500 exist yet,but i think the 8500 does.
anyone who doubts apple will be humbled come the next macworld!
i still dont think we will see an all-new motherboard with the initial iteration of the G5 power-mac.
but i think the next macworld san francisco will see an all-new motherboard.
im drooling over the prospect of a aqua-enhanching graphics subsystem and optical port.
i dont think in terms of MIPS or wheatstones or whatever,i think in terms of apps,new ways of working,ect....
and so does apple.
Well. Things begin to happen. I am glad to see it.
Next step will be at NAB at Apr 8th.
Let me ask a question!
Why Apple choose 23inch TFT-LCD Monitor as Cinma HD ?
it does not have any input for HDTV reception.
Or will it be ?
It came from LG.Philips LCD Co., Ltd as OEM part.
<a href="http://www.lgphilips-lcd.com:8888/English/product/p.html"; target="_blank">http://www.lgphilips-lcd.com:8888/English/product/p.html</a>;
It is in the montor section.
23" Wide UXGA LM230W1
Feature
-Wide Type
-High Resolution(UXGA)
-Wide Viewing Angle(Super-IPS)
Model Name LM230W1
Active Area[mm] 470.4 x 301.1
Outline Dimension[mm] 550.0 x 362.5
Thickness[mm] 24
Resolution 1920xRGBx1200
Aspect Ratio 16:10
Pixel Pitch[mm] 0.258(98.4)
Number of Colors 16.7M
Luminance[cd/㎡] 235cd/㎡,6lamps
Color Saturation 65
Weight[g] 5,500
Contrast Ratio 300 : 1
Interface TMDS
Viewing Angel[˚,U/D/L/R] 80/80/80/80(Super-IPS)
Color Temperature[K] 6,500
Response Time[ms] 40
MP Schedule Now
23" Wide UXGA LM230W1
What will be next ?
<a href="http://www.lgphilips-lcd.com:8888/English/news/n_cozy.html?idx=401&offset=0&pkinds=movement&pname =news" target="_blank">http://www.lgphilips-lcd.com:8888/English/news/n_cozy.html?idx=401&offset=0&pkinds=movement&pname =news</a>
or
<a href="http://www.samsungelectronics.com/semiconductors/TFT_LCD/product_news/semiadmin_1010104542921_108.html"; target="_blank">http://www.samsungelectronics.com/semiconductors/TFT_LCD/product_news/semiadmin_1010104542921_108.html</a>;
And if you can remember what I was talked about, I think we will see good sign soon.
P.S.: SAMSUNG & LG.Philips LCD Co., Ltd is making TFT-LCD for New iMac.
[ 03-21-2002: Message edited by: kormac77 ]</p>
Suddenly, the "There is no G5 Theory" espoused by y/t looks a little thin.....
Might lose this bet after all.
SdC, SdA, and CMC (Color Me Concerned), aka:
There is no G5
[ 03-21-2002: Message edited by: There is no g5 ]</p>
<strong>And if you can remember what I was talked about, I think we will see good sign soon.</strong><hr></blockquote>
Honestly I don't remember except something about DNA you talked about since like 1999. But since 2001 Apple already changed the strategy from a computer maker to consumer electronics maker, so I am not sure if those things you said in 1999 still holds even you got inside info at 1999.
But I do believe that G5 will be coming before the end of 2002.
Good to see you back old sod. Keep on giving us the inside scoop.
Finally!
<img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
I don't see how anything I posted is dumb, maybe a little too dramatic, sure. Anything Kormac has talked about has not come to pass and never will. My analysis is at least as in tune with reality as his.
That price increase is just plain contemptuous of loyal Apple customers. Before, Apple was falling behind by moving slowly on spec and standing pat on price, now they're falling further behind by standing pat on spec while raising prices.
I sure hope this is temporary. I think that it is, but it's bad press nonetheless. In an industry that isn't exactly fair to Apple they're going to get lambasted for this price increase. Margins are certainly important, but you simply cannot compete without lower priced models. Those don't have to be low margin models, but they need to get affordable to larger audience if they are truly serious about market share. Anyone with a long term business investment in a platform needs to be concerned about market share too. Timely updates of software, third party support, peripheral availibility, etc depends on that.
Why isn't there a 500 USD computer in the Apple line-up? Many people simply WILL NOT buy a computer above that price point. If they can't build the CRT iMac down to that price, then they need to re-engineer it so that they can. Consumers, education buyers, and businesses who vote with their pocketbooks, have been voting against Apple's rigid product structures. IF you need to strip down a computers spec a bit to get a cheaper option, then do it. This isn't like selling a high-end car, or a watch. When some-one buys a computer, they begin making a software investment in that platform. Eventually, the prospect of switching just becomes too expensive/inconvenient, even when people would really like to switch. A sale lost today isn't neccessarily a sale you can get back on the next upgrade cycle. It is even more important to bring people into a platform on the entry level.
Apple's software answer has been brilliant. And it's professional hardware answer adequate. But it has no complete consumer hardware answer. Recent moves only take Apple further out of the low-end market, which incidentally is the only market that can grow your market-share. While it isn't catastrophic to raise prices, it is pretty bad. Tied into a recent trend of sloth like hardware advancement, it just looks that much worse.
People don't really mind paying more, but they want to get more when they do. Take the iPod and HD display. No problem with those. They're clearly high-end products, with high, but very competitive prices. iMac is supposed to be a consumer machine, and as such it is approaching the wrong end of the consumer price range. They need to do better.