That's 3.2GB/sec per processor, for a total bandwidth to the memory of 6.4GB/sec. If the bus was architected not as a ring but with two independent connections to the memory controller, then the total bandwidth would be 12.8GB/sec, which is more than even dual-channel DDR400 could provide. In other words - it doesn't really matter, because no memory can feed that right now. If future generations of the 970 retain the 4:1 DDR FSB, then we'll always have more bus bandwidth than memory bandwidth either way.
On a side note, it's a welcome change to have a faster bus than memory instead of the other way around.
No, you misunderstand what the ring topology would look like. Data would be introduced to the ring at some point and be sent to the next device in the ring. That device would look at it and/or forward it on, and so on until the data arrived back at the place where it started whereupon it would be removed from the ring. This is done so that all devices can see all of the data (i.e. bus snooping) and to avoid having signal lines which are other than point-to-point. In this kind of a topology, therefore, all devices send and receive all data which means you have an effective 3.2 GB/sec data rate which is shared between all the devices.
Based on the published information so far, this is one of the potential bus architectures they might be using in the 970. The others have been discussed before (i.e. one bus per FSB), I just thought it interesting to post another than hadn't been considered so far.
I can't agree I'm afraid...the point of the xserve is reliability not out and out speed.
As the situation stands the xserve is slightly slower than the top PM's.
There's no need to put a hot expensive chip in a server box that needs to be stable and problem free.
I agree. Apple will probably introduce a 970 xserve in the near future after the PMs. The release will probably be close to the PM release but the PMs will be the first to get it.
No, you misunderstand what the ring topology would look like. Data would be introduced to the ring at some point and be sent to the next device in the ring. That device would look at it and/or forward it on, and so on until the data arrived back at the place where it started whereupon it would be removed from the ring. This is done so that all devices can see all of the data (i.e. bus snooping) and to avoid having signal lines which are other than point-to-point. In this kind of a topology, therefore, all devices send and receive all data which means you have an effective 3.2 GB/sec data rate which is shared between all the devices.
Well, IBM has gotten their old token-ring architecture working at a pretty astonishingly high degree of efficiency, so maybe they could get this to work.
I have to say, it's an elegant solution if it doesn't impose too much of a bottleneck.
Apple should use the fastest available chip in the low end single CPU machine, and duals in the other two. PM's are still very expensive and with 970 backed performance they'll better justify their current prices but will still ALL fall in the high-end. 1500 for a machine without a display? Not good, considering that a single low speed ppc970 will not offer any kind of performance revelation over far cheaper x86 boxen, the game is either to use the fastest 970's available or duals.
1.8 single, dual 1.4, and dual 1.8. Same configs as currently used. Or they could possibly use something just a mite slower in the low end and include a superdrive, which at the price ought to be included in all PM's, or they could, gasp, further lower the price of the entry level tower.
i dont think price will be a big issue at leats price per speed issues. I just made the cheapest dell i could while being comparable to a G4 and it wasn't bad
--------- $3,034 ------------
Dell Precision? Workstation 340 Minitower: Intel® Pentium® 4 Processor, 3.06GHz, 512K / 533 Front Side Bus
Memory: 512MB PC800 ECC RDRAM® (2 RIMMS?)
Keyboard: Entry Level Quietkey Keyboard, PS/2, (No Hot Keys)
that is the cheapest PC that is comparable i could make...perhaps i was at the wrong place at dell because i'll admit i dotn know it well but those specs are very close to the G4 (dual 1.42 base system)
perhaps i missed something so please feel free to correct any errors
...with a 970 there will be no doubt about price...i dont even really see a problem with price now, i think u are greatly devaluating OS X and other software
In this kind of a topology, therefore, all devices send and receive all data which means you have an effective 3.2 GB/sec data rate which is shared between all the devices.[/B]
Whoops! I guess I forgot about that (rather important) bit.
For some reason making the memory controller party to cache transfers between processors seems kind of stupid to me though. There's just something... wrong about this.
Go ahead, get smug about it, I'm not the one losing market-share on a quaterly basis. Has Apple ever strung together more than 2 quaters of growth (as a percentage of marketshare) in the last 5 years?
I wasn't intending to be 'smug'. For work, I'd like a box that you'd regard as 'extremely overpriced'. I'll pay the Apple-tax for the performance, it's cheaper than an RS6000 by a wee bit. For home I'd like the exact same thing (I think) you would. (A pizzabox with maximum bang, minimum glitz). I'll probably get another iMac though.
I think you're also correct about the upgradeability. One slot that can be either a PCI slot or an AGP slot (depending on what the user wanted most) would please a lot of people. But the very first step to having a viable product in that spot is putting out something with a heck of a lot more performance than a Dual 1.4 G4+. I could see the return of something in this niche, maybe 6mo after the 970's intro - _if_ Apple was doing quite well on the tower sales. But Apple has become a niche player - they have to please the niche's they've got good entrenchment in _before_ they dabble in new niches. The gamer/hotrodder/casemodder/upgrader/whatnot niche is a niche where Apple's _got_ to have a lot less than their global marketshare in that niche.
No, I don't think Apple has put together 2 quarters of marketshare growth in 5 years. But I do think they put together 3 sequential _years_ of unit-sales-growth. I don't think many of the PC-makers have had more than a couple quarters of "marketshare growth" either. Dell has, of course, but name another.
That's 3.2GB/sec per processor, for a total bandwidth to the memory of 6.4GB/sec. If the bus was architected not as a ring but with two independent connections to the memory controller, then the total bandwidth would be 12.8GB/sec, which is more than even dual-channel DDR400 could provide.
Well I guess we'll just have to have quad channel then
1500 for a machine without a display? Not good, considering that a single low speed ppc970 will not offer any kind of performance revelation over far cheaper x86 boxen
Says who? Based on what factual speed data? Are you speculating? I wasn't aware that the PPC 970 was benchmarked. I'm betting that even the CURRENT G4s would keep up with the Current Intel/Athlon systems if they had the 970s BANDWIDTH. Given the fact that the 970 will already outperform the current G4s by a factor of '2', it would be interesting to see what a dual config would do. In any event, I suspect that you are only guessing.
Matsu writes:
Quote:
when Apple provides more machines that do that, I will own more Apples, if they don't, I will own more windows, and if Apple drags its feet too much, I will own only windows. What Steve needs to understand is that macs are nice, but they simply aren't that nice.
I asked you once before and you never provided us with an answer that we could refer back to in the future... How much of a premium would you *allow* Apple to charge over a comparable Windows PC? Oh, and do not include the bargain-basement models or the I-built-it-myselfers... These ARE NOT high-end systems and they aren't representative of what serious computing firms are buying. So, again, how much of a premium are you going to allow over truly similar brand-configured PCs? I want you to also factor in all the other stuff Apple does (touches?) with the whole package that other OEMs simply do not do.
Matsu writes:
Quote:
Go ahead, get smug about it, I'm not the one losing market-share on a quaterly basis
Oh yeah, that'd be Gateway... Errr they're just plain *losing*.
There's no use argueing with a cheapskate. People who put price as their main motivator generally will try to diminish other products to appease their own wants. In Matsu's case here he say OSX isn't "that" much better. Which falls in line with the excuses that most people make who hate giving up coins. "Beggars can't be choosers"
All we get is pessimistic and unfounded statements from Matsu. In his mind PC's are sooo cheap and fast and Macs are horribly expensive and slow. Somewhere in the middle is the truth.
hmurchison... I'm not trying to argue with Matsu, I'm trying to pin him down so we can hold him to anything he says in the future regarding price/performance/feature set/pros and cons etc... If he provides his answer as a percentage (%) then even better. Given that Apple does so much more in terms of actual R&D with respect to their product offerings, it would only be reasonable (fair?) to grant them a bit of a premium. On the flip-side, he talks about the price war that's occurring in the PC sector right now. Well, razor-thin profits trying to move gobs of volume in a sluggish economy is a recipe for disaster... Someone else stated something similar and its absolutely true. All I want him to do is answer my question as I have presented it so we can all take note of his answer. On top of that I want him to use brand-name equipment from PC OEMs. No bargain-basement crates and no home-built boxes. Not only wouldn't it be fair, it would be ridiculous to do otherwise.
OEM to OEM, feature to feature, price to price, performance to performance. Also keep in mind that 90%+ of the people out there don't require anything more than a 500 MHz. PIII. I suspect that we are going to see the PC sector hemorrhage a bit more. The machines just aren't selling and regardless of his *claims*, the *cheapest* PCs aren't the ones that are selling the best... I suspect that's why Programmer and Amorph and a few others just seem to skip over his posts anymore. When the Mac is on 64-bit mainstream, consumer desktops and the Wintelon world is still trying to figure out which platform is the way to go (i.e., IA-32 (going to be around well into the next decade), IA-64 (no chance to making it to the desktop), X86-64-AMD (no desktop OS and no *official* Windows Server release), X86-64-Intel (Yamhill? Yeah, and where is the desktop OS?) The future is really uncertain for the Windows folk. They are going to be stuck with 32-bit for quite some time -- and the Mac (and Apple) will be moving forward with whatever IBM has planned on their roadmap.
God help us if Apple decides to licensee OS X to run full-tilt on IBM hardware. Talk about a bloodbath with respect to performance comparisons... But that's getting a little outlandish at this point. For now, I just want him to publicly state for all of us, once and for all (in percentage terms) what he thinks is a *fair* premium/tax for Apple products. Anyone have anything else to add?
they don't stoop as low as $500 machines that are competitive...i guess they dont want to or cant. but mid consumer to pro they are perfectly competitive...and will become even more so...period
An interesting thread indeed. The only problems I have with it are why put the 2 slower processors in an MP configuration, as your high end, or low end machine? It makes no since to me to disappoint your pro customers looking for the highest performance level from a PowerMac, and snub them of getting the fastest possible configuration which is what they need. Anyone using Maya, or any 3D package knows you should never skimp on hardware. Speed is an essential, or... more like a necessity. I'm hoping for Maya 5 Unlimited to be released for OS X Panther sometime soon after these machines. Complete just isn't the same, and even then I should be using a PC with it now, because the performance is there, but I am content to wait for these new PowerMacs, and see what IBM has in store for us.
The second problem I have is why do people keep saying the 970 is not much faster than a G4? Isn't the 970 a PPC - Power4 hybrid crossed together with major improvements including an IBM Altivec SIMD? The Power4 @ 1000-MHz is faster than a P4 @ 4GHz in server performance. The Power4 was second only to 1, or 2 processors in the server ranking list I saw that was floating around when IBM announced this processor. So I'm not to sure what has happened since then, but it seems I missed some new perspective benchmarks, or some info has brought the specs of this processor way down.
GIant bloody fallacy number one, Apple spends more on R&D. They simply DON'T. I've broken this down too many times to count, go back and look it up -- Apple's cost versus a dell will never account for more than maybe 50 bucks (talking R&D) and only because they sell less machines, DELL spends MORE than Apple overall.
Apple has to be in the ballpark, they don't have to be cheaper, intangibles can count, but they can't justify enormous price increases.
Apple's highest end machine may be priced right for pro customers, but it doesn't perform right. Fine. I'm never going to buy that machine or one like it, not from Apple, not from DELL. If the performance comes up to match, and sales come up too, then Apple can justifiably argue the price is right for the people that buy those machines. I've no problem with that, read more carefully if you think that I do.
Where Apple has the greatest problems is in the range of semi affordable machines, these are always handicapped in some way, or otherwise seriously over-priced, and these are precisely the machines that strike the best balance between profitability and salability. Apple just doesn't compete there. iMacs are far too restricted, low end towers are far too expensive, Apple keeps losing share.
Ed.M, you don't fool me, I find your arguments idiotic, and despite your controlled prose you offer nothing but thinly disguised fan-boyism. "Pin me down???" WTF are you on about? The problems with Apple scarcely get a more honest articulation in the mac community than when I make them (repeatedly). You seem to want to say, nothing from Apple would please me and I want to leave the door open so that nothing from them could possibly please me.
Cling to that fantasy if you like. The honest truth is that it would take very little from Apple to please me, nothing impossible, nothing most of the rest of the industry manages quite easily. That's the frustrating part, it wouldn't take much for Apple to recognize what buyers want/need and find ways to supply it at a competitive price, but they don't. They feel they are unique by virtue of superior software, but the simple reality is that it isn't quite so superior. XP works very well, it's interface tries harder than previous windows to get in your way, but it works very reliably, and as the market numbers show, it isn't hard for most people to get used to it.
When market share numbers start to climb, when investors aren't afraid of Apple anymore, then we can talk, but untill then, Apple just keeps prooving me right. And if that time comes, you will find it can come ONLY because Apple has begun to do things that I've been saying they should be doing for the past 4 years.
You (and Steve) want to insist that somehow Apple faces a complicated problem, but it just isn't so, the problem is very simple and doesn't require huge contortions of marketing, branding, and retail excursions. Apple finds it very difficult to amend their philosophy to meet the demands of the market, so they would rather try and fertilize a niche, but it's expensive (in terms of time and mindshare) and ultimately a losing battle (as the numbers attest), and unnecessary too. Compete, listen, give people what they want, they will buy it -- try too hard to force your wants onto consumers; they won't buy it. A line exists in there, you want to attempt to do both simultaneously, strike the right balance. Apple isn't there. They may never be without a lot of pressure from consumers.
Or maybe we should all just buy macs no matter what Apple offers, as if they were some kind of charity for the aesthetic computing.
God help us if Apple decides to licensee OS X to run full-tilt on IBM hardware.
... and the solution to the 'clone' problems is to force it to be Mac OS X _Server_. At that point Apple could care less how low in price IBM prices blades or whatnot - the 'Apple tax' would be paid completely in the software price.
Whoops! I guess I forgot about that (rather important) bit.
For some reason making the memory controller party to cache transfers between processors seems kind of stupid to me though. There's just something... wrong about this.
Consider that the companion chip (which includes the memory controller) might include things like the AGP bus and a shared L3 cache.
... and the solution to the 'clone' problems is to force it to be Mac OS X _Server_. At that point Apple could care less how low in price IBM prices blades or whatnot - the 'Apple tax' would be paid completely in the software price.
Exactly what I was thinking.
There are many shops that would raise their nose at anything with an Apple logo in their racks. However, the same shops are seriously looking at Linux/Unix as a more secure choice for servers (nevermind the new license schemes coming from Redmond).
This is not a technical comment but a practical one in line with what Matsu said. It seems nearly certain that Apple will use the 970 chip now. A 64-bit chip in a consumer computer. That is exciting. But I hope Apple does not overestimate its appeal when they price it. Ive bought two PCs in the last 2 years for my kids. They had to be fairly powerful as they wanted to play games (surprise!). They cost $1200 and $1100 and included monitors. Matsu is right in that respect. A grand gets you a pretty good PC.
I think I read here that the 970 wont be too costly to produce. So please, Apple, dont pull a Cube thing. Sure, make some money. But dont negate all the good press you are likely to get with too high a price. Puhleeeezzze?!
They cost $1200 and $1100 and included monitors. Matsu is right in that respect. A grand gets you a pretty good PC.
$1299 gets you a pretty good Mac with a Flat Panel no less.
Matsu's problem is not that he's asking for a cheap Mac. They "do" exist. It's that he's asking for Powermacs to occupy this space when iMacs fit in just fine. Having 970s in Powermacs should allow Apple to boost the low end accordingly. Funny how with Macs you pay more for the Big Box which is in contrast to PC's where you pay more for "not" choosing the Big Box.
Comments
Originally posted by Anonymous Karma
That's 3.2GB/sec per processor, for a total bandwidth to the memory of 6.4GB/sec. If the bus was architected not as a ring but with two independent connections to the memory controller, then the total bandwidth would be 12.8GB/sec, which is more than even dual-channel DDR400 could provide. In other words - it doesn't really matter, because no memory can feed that right now. If future generations of the 970 retain the 4:1 DDR FSB, then we'll always have more bus bandwidth than memory bandwidth either way.
On a side note, it's a welcome change to have a faster bus than memory instead of the other way around.
No, you misunderstand what the ring topology would look like. Data would be introduced to the ring at some point and be sent to the next device in the ring. That device would look at it and/or forward it on, and so on until the data arrived back at the place where it started whereupon it would be removed from the ring. This is done so that all devices can see all of the data (i.e. bus snooping) and to avoid having signal lines which are other than point-to-point. In this kind of a topology, therefore, all devices send and receive all data which means you have an effective 3.2 GB/sec data rate which is shared between all the devices.
Based on the published information so far, this is one of the potential bus architectures they might be using in the 970. The others have been discussed before (i.e. one bus per FSB), I just thought it interesting to post another than hadn't been considered so far.
Originally posted by robster
I can't agree I'm afraid...the point of the xserve is reliability not out and out speed.
As the situation stands the xserve is slightly slower than the top PM's.
There's no need to put a hot expensive chip in a server box that needs to be stable and problem free.
I agree. Apple will probably introduce a 970 xserve in the near future after the PMs. The release will probably be close to the PM release but the PMs will be the first to get it.
Originally posted by Programmer
No, you misunderstand what the ring topology would look like. Data would be introduced to the ring at some point and be sent to the next device in the ring. That device would look at it and/or forward it on, and so on until the data arrived back at the place where it started whereupon it would be removed from the ring. This is done so that all devices can see all of the data (i.e. bus snooping) and to avoid having signal lines which are other than point-to-point. In this kind of a topology, therefore, all devices send and receive all data which means you have an effective 3.2 GB/sec data rate which is shared between all the devices.
Well, IBM has gotten their old token-ring architecture working at a pretty astonishingly high degree of efficiency, so maybe they could get this to work.
I have to say, it's an elegant solution if it doesn't impose too much of a bottleneck.
Originally posted by Matsu
Apple should use the fastest available chip in the low end single CPU machine, and duals in the other two. PM's are still very expensive and with 970 backed performance they'll better justify their current prices but will still ALL fall in the high-end. 1500 for a machine without a display? Not good, considering that a single low speed ppc970 will not offer any kind of performance revelation over far cheaper x86 boxen, the game is either to use the fastest 970's available or duals.
1.8 single, dual 1.4, and dual 1.8. Same configs as currently used. Or they could possibly use something just a mite slower in the low end and include a superdrive, which at the price ought to be included in all PM's, or they could, gasp, further lower the price of the entry level tower.
i dont think price will be a big issue at leats price per speed issues. I just made the cheapest dell i could while being comparable to a G4 and it wasn't bad
--------- $3,034 ------------
Dell Precision? Workstation 340 Minitower: Intel® Pentium® 4 Processor, 3.06GHz, 512K / 533 Front Side Bus
Memory: 512MB PC800 ECC RDRAM® (2 RIMMS?)
Keyboard: Entry Level Quietkey Keyboard, PS/2, (No Hot Keys)
Monitor: No Monitor Option
Graphics Card: nVidia, Quadro4 700XGL, 64MB, VGA/DVI (dual monitor capable)
Hard Drive: 120GB 7200RPM IDE Hard Drive with DataBurst Cache?
Operating System: Microsoft® Windows® XP Professional, SP1 with Media using NTFS
Mouse: Dell, PS/2 (2-button, no scroll)
CD ROM, DVD, and Read-Write Drives: 4X DVD+RW with Sonic DVDIt! SE -for professional authoring
Other Options: 1394 Controller Card 1394
Hardware Support Services: 3Yr Same Day 4Hr Response Parts + Onsite Labor (M-F 8am-6pm) W3Y5X10 [900-2710] [900-2712]
Installation Services: No Installation NOINSTL [900-9987]
Energy Star?: Energy Star? ES [310-6414]
Security Software - Shipped Separately: Symantec Antivirus CorpEd 8.0 (1 user license) SANTIV [A0059152]
that is the cheapest PC that is comparable i could make...perhaps i was at the wrong place at dell because i'll admit i dotn know it well but those specs are very close to the G4 (dual 1.42 base system)
perhaps i missed something so please feel free to correct any errors
...with a 970 there will be no doubt about price...i dont even really see a problem with price now, i think u are greatly devaluating OS X and other software
Originally posted by Programmer
In this kind of a topology, therefore, all devices send and receive all data which means you have an effective 3.2 GB/sec data rate which is shared between all the devices.[/B]
Whoops! I guess I forgot about that (rather important) bit.
For some reason making the memory controller party to cache transfers between processors seems kind of stupid to me though. There's just something... wrong about this.
Originally posted by Matsu
Go ahead, get smug about it, I'm not the one losing market-share on a quaterly basis. Has Apple ever strung together more than 2 quaters of growth (as a percentage of marketshare) in the last 5 years?
I wasn't intending to be 'smug'. For work, I'd like a box that you'd regard as 'extremely overpriced'. I'll pay the Apple-tax for the performance, it's cheaper than an RS6000 by a wee bit. For home I'd like the exact same thing (I think) you would. (A pizzabox with maximum bang, minimum glitz). I'll probably get another iMac though.
I think you're also correct about the upgradeability. One slot that can be either a PCI slot or an AGP slot (depending on what the user wanted most) would please a lot of people. But the very first step to having a viable product in that spot is putting out something with a heck of a lot more performance than a Dual 1.4 G4+. I could see the return of something in this niche, maybe 6mo after the 970's intro - _if_ Apple was doing quite well on the tower sales. But Apple has become a niche player - they have to please the niche's they've got good entrenchment in _before_ they dabble in new niches. The gamer/hotrodder/casemodder/upgrader/whatnot niche is a niche where Apple's _got_ to have a lot less than their global marketshare in that niche.
No, I don't think Apple has put together 2 quarters of marketshare growth in 5 years. But I do think they put together 3 sequential _years_ of unit-sales-growth. I don't think many of the PC-makers have had more than a couple quarters of "marketshare growth" either. Dell has, of course, but name another.
Originally posted by Anonymous Karma
That's 3.2GB/sec per processor, for a total bandwidth to the memory of 6.4GB/sec. If the bus was architected not as a ring but with two independent connections to the memory controller, then the total bandwidth would be 12.8GB/sec, which is more than even dual-channel DDR400 could provide.
Well I guess we'll just have to have quad channel then
Originally posted by Whisper
Well I guess we'll just have to have quad channel then
yeah, I thought the same
Is there such a thing yet, or planned?
1500 for a machine without a display? Not good, considering that a single low speed ppc970 will not offer any kind of performance revelation over far cheaper x86 boxen
Says who? Based on what factual speed data? Are you speculating? I wasn't aware that the PPC 970 was benchmarked. I'm betting that even the CURRENT G4s would keep up with the Current Intel/Athlon systems if they had the 970s BANDWIDTH. Given the fact that the 970 will already outperform the current G4s by a factor of '2', it would be interesting to see what a dual config would do. In any event, I suspect that you are only guessing.
Matsu writes:
when Apple provides more machines that do that, I will own more Apples, if they don't, I will own more windows, and if Apple drags its feet too much, I will own only windows. What Steve needs to understand is that macs are nice, but they simply aren't that nice.
I asked you once before and you never provided us with an answer that we could refer back to in the future... How much of a premium would you *allow* Apple to charge over a comparable Windows PC? Oh, and do not include the bargain-basement models or the I-built-it-myselfers... These ARE NOT high-end systems and they aren't representative of what serious computing firms are buying. So, again, how much of a premium are you going to allow over truly similar brand-configured PCs? I want you to also factor in all the other stuff Apple does (touches?) with the whole package that other OEMs simply do not do.
Matsu writes:
Go ahead, get smug about it, I'm not the one losing market-share on a quaterly basis
Oh yeah, that'd be Gateway... Errr they're just plain *losing*.
--
Ed M.
There's no use argueing with a cheapskate. People who put price as their main motivator generally will try to diminish other products to appease their own wants. In Matsu's case here he say OSX isn't "that" much better. Which falls in line with the excuses that most people make who hate giving up coins. "Beggars can't be choosers"
All we get is pessimistic and unfounded statements from Matsu. In his mind PC's are sooo cheap and fast and Macs are horribly expensive and slow. Somewhere in the middle is the truth.
OEM to OEM, feature to feature, price to price, performance to performance. Also keep in mind that 90%+ of the people out there don't require anything more than a 500 MHz. PIII. I suspect that we are going to see the PC sector hemorrhage a bit more. The machines just aren't selling and regardless of his *claims*, the *cheapest* PCs aren't the ones that are selling the best... I suspect that's why Programmer and Amorph and a few others just seem to skip over his posts anymore. When the Mac is on 64-bit mainstream, consumer desktops and the Wintelon world is still trying to figure out which platform is the way to go (i.e., IA-32 (going to be around well into the next decade), IA-64 (no chance to making it to the desktop), X86-64-AMD (no desktop OS and no *official* Windows Server release), X86-64-Intel (Yamhill? Yeah, and where is the desktop OS?) The future is really uncertain for the Windows folk. They are going to be stuck with 32-bit for quite some time -- and the Mac (and Apple) will be moving forward with whatever IBM has planned on their roadmap.
God help us if Apple decides to licensee OS X to run full-tilt on IBM hardware. Talk about a bloodbath with respect to performance comparisons... But that's getting a little outlandish at this point. For now, I just want him to publicly state for all of us, once and for all (in percentage terms) what he thinks is a *fair* premium/tax for Apple products. Anyone have anything else to add?
--
Ed M.
they don't stoop as low as $500 machines that are competitive...i guess they dont want to or cant. but mid consumer to pro they are perfectly competitive...and will become even more so...period
The second problem I have is why do people keep saying the 970 is not much faster than a G4? Isn't the 970 a PPC - Power4 hybrid crossed together with major improvements including an IBM Altivec SIMD? The Power4 @ 1000-MHz is faster than a P4 @ 4GHz in server performance. The Power4 was second only to 1, or 2 processors in the server ranking list I saw that was floating around when IBM announced this processor. So I'm not to sure what has happened since then, but it seems I missed some new perspective benchmarks, or some info has brought the specs of this processor way down.
Apple has to be in the ballpark, they don't have to be cheaper, intangibles can count, but they can't justify enormous price increases.
Apple's highest end machine may be priced right for pro customers, but it doesn't perform right. Fine. I'm never going to buy that machine or one like it, not from Apple, not from DELL. If the performance comes up to match, and sales come up too, then Apple can justifiably argue the price is right for the people that buy those machines. I've no problem with that, read more carefully if you think that I do.
Where Apple has the greatest problems is in the range of semi affordable machines, these are always handicapped in some way, or otherwise seriously over-priced, and these are precisely the machines that strike the best balance between profitability and salability. Apple just doesn't compete there. iMacs are far too restricted, low end towers are far too expensive, Apple keeps losing share.
Ed.M, you don't fool me, I find your arguments idiotic, and despite your controlled prose you offer nothing but thinly disguised fan-boyism. "Pin me down???" WTF are you on about? The problems with Apple scarcely get a more honest articulation in the mac community than when I make them (repeatedly). You seem to want to say, nothing from Apple would please me and I want to leave the door open so that nothing from them could possibly please me.
Cling to that fantasy if you like. The honest truth is that it would take very little from Apple to please me, nothing impossible, nothing most of the rest of the industry manages quite easily. That's the frustrating part, it wouldn't take much for Apple to recognize what buyers want/need and find ways to supply it at a competitive price, but they don't. They feel they are unique by virtue of superior software, but the simple reality is that it isn't quite so superior. XP works very well, it's interface tries harder than previous windows to get in your way, but it works very reliably, and as the market numbers show, it isn't hard for most people to get used to it.
When market share numbers start to climb, when investors aren't afraid of Apple anymore, then we can talk, but untill then, Apple just keeps prooving me right. And if that time comes, you will find it can come ONLY because Apple has begun to do things that I've been saying they should be doing for the past 4 years.
You (and Steve) want to insist that somehow Apple faces a complicated problem, but it just isn't so, the problem is very simple and doesn't require huge contortions of marketing, branding, and retail excursions. Apple finds it very difficult to amend their philosophy to meet the demands of the market, so they would rather try and fertilize a niche, but it's expensive (in terms of time and mindshare) and ultimately a losing battle (as the numbers attest), and unnecessary too. Compete, listen, give people what they want, they will buy it -- try too hard to force your wants onto consumers; they won't buy it. A line exists in there, you want to attempt to do both simultaneously, strike the right balance. Apple isn't there. They may never be without a lot of pressure from consumers.
Or maybe we should all just buy macs no matter what Apple offers, as if they were some kind of charity for the aesthetic computing.
Originally posted by Ed M.
God help us if Apple decides to licensee OS X to run full-tilt on IBM hardware.
... and the solution to the 'clone' problems is to force it to be Mac OS X _Server_. At that point Apple could care less how low in price IBM prices blades or whatnot - the 'Apple tax' would be paid completely in the software price.
Originally posted by Anonymous Karma
Whoops! I guess I forgot about that (rather important) bit.
For some reason making the memory controller party to cache transfers between processors seems kind of stupid to me though. There's just something... wrong about this.
Consider that the companion chip (which includes the memory controller) might include things like the AGP bus and a shared L3 cache.
Originally posted by Nevyn
... and the solution to the 'clone' problems is to force it to be Mac OS X _Server_. At that point Apple could care less how low in price IBM prices blades or whatnot - the 'Apple tax' would be paid completely in the software price.
Exactly what I was thinking.
There are many shops that would raise their nose at anything with an Apple logo in their racks. However, the same shops are seriously looking at Linux/Unix as a more secure choice for servers (nevermind the new license schemes coming from Redmond).
Screed
I think I read here that the 970 wont be too costly to produce. So please, Apple, dont pull a Cube thing. Sure, make some money. But dont negate all the good press you are likely to get with too high a price. Puhleeeezzze?!
They cost $1200 and $1100 and included monitors. Matsu is right in that respect. A grand gets you a pretty good PC.
$1299 gets you a pretty good Mac with a Flat Panel no less.
Matsu's problem is not that he's asking for a cheap Mac. They "do" exist. It's that he's asking for Powermacs to occupy this space when iMacs fit in just fine. Having 970s in Powermacs should allow Apple to boost the low end accordingly. Funny how with Macs you pay more for the Big Box which is in contrast to PC's where you pay more for "not" choosing the Big Box.