If it's not too far off topic, I'd like to hear some opinions about why Apple selected a 1 GHz G4 with 256K of L2 cache for the iBook, rather than going with the IBM 750 GX, the G3 Gobi that everyone was looking forward to. It is also 1 GHz but with 1 M L2 cache, four times as much, and dissipates less than 10 W. It likely costs less and would have been an easy and good value upgrade for the iBook, which was already a G3. The iBook could have moved to a G4 when the IBM Mojave is here. I'm pretty sure Mojave will be lower cost than a Motorola G4, probably significantly so.
Apple's decision to ignore the Gobe has puzzled me, but evidently no one else. The only comment I've seen are like, "Great, the iBook has a G4." With the larger L2 cache, I would think the G3 would perform as well as the G4 overall. Gobe would be faster on everything except AltiVec accelerated applications. From a value standpoint, it looks to me like Apple jumped to a G4 too soon. Comments?
They might have had the G4 ibook waiting in the wings for a while. This might explain why they stuck to the Moto G4.. There might be other reasons too.. Maybe the Gobe kicks the crap out of the cpu's in the powerbook's.. Without having any realworld benchmarks I can only guess. I think MOTO let Apple down in such a way that it has held them back with the exception of the G5 tower. Like I said before IBM seem to be working overtime to give them the goods...
Alt-Vec has become very important to the applicaitons that Apple wants to deliver. I believe this is the overriding consideration. If Gobi had a vector capability it probally would be in the ibook right now.
From my perspective the ibook lost a bit with the G4. I would have liked to seen battery life as the overwhelming design consideration. I believe it is a markatable a feature as ultimate performance, and frankly is one of the reasons for the iBooks success. But I do understand Apple a bit with respect to this issue. When Alt Vec can be used it is almost always a significant advantage that the G3 series had no way of overcoming.
Thanks
Dave
Quote:
Originally posted by snoopy
If it's not too far off topic, I'd like to hear some opinions about why Apple selected a 1 GHz G4 with 256K of L2 cache for the iBook, rather than going with the IBM 750 GX, the G3 Gobi that everyone was looking forward to. It is also 1 GHz but with 1 M L2 cache, four times as much, and dissipates less than 10 W. It likely costs less and would have been an easy and good value upgrade for the iBook, which was already a G3. The iBook could have moved to a G4 when the IBM Mojave is here. I'm pretty sure Mojave will be lower cost than a Motorola G4, probably significantly so.
Apple's decision to ignore the Gobe has puzzled me, but evidently no one else. The only comment I've seen are like, "Great, the iBook has a G4." With the larger L2 cache, I would think the G3 would perform as well as the G4 overall. Gobe would be faster on everything except AltiVec accelerated applications. From a value standpoint, it looks to me like Apple jumped to a G4 too soon. Comments?
Alt-Vec has become very important to the applicaitons that Apple wants to deliver. I believe this is the overriding consideration. If Gobi had a vector capability it probally would be in the ibook right now. . .
I have a feeling you are right. Except for lack of AltiVec, Gobi would have been a better choice than a Motorola G4 in every way, battery life (as you point out), performance (1 MB L2 cache), power (under 10 Watts) and price, most likely.
I've been wondering about something else regarding a good value, low-end Mac, but I doubt there are many who have an answer to this question. Does HyperTransport add cost to a system? That is, how does its cost compare with the standard way of connecting system components, before the G5? I was wondering whether it made sense to use HyperTransport in low-end products?
HyperTransport is not at all expensive to implement. That's one of its design goals, and they appear to have succeeded.
I have to continue to go against the Dell-with-an-Apple-badge thinking going on here. The more I think about my little appliance, the more I like it.
It can be integrated to a degree that a PC can't, adding "sex appeal" and reducing cost (expansion through PCI adds cost).
Since it's designed for light, general use, it's better at that than the PCs are, and it can come near the advertised cost for the PC box alone, except that you get a whole system.
It's not a Mac, so it carries none of the baggage that the Mac carries. It is an Apple, so it has all the cachet that the Apple name has.
Dell can't compete with it. Their whole model is based on the idea that the customer knows going in that they're going to get a bog standard tower and monitor that are identical to anyone else's except for little plastic bits here and there, so ordering them blind is no big deal. But accessories have to be seen first. This is the genius of iPod, and one reason why it's outselling the Dell DJ by about 20:1. Apple has turned their dependence on retail into an asset, and Dell's direct advantage into a liability.
No PC maker can compete with it (not even Sony, because they have to design for Windows CE, or PocketPC, or whatever MS is calling that failure this week). It plays to all of Apple's strengths: Engineering, integration, design, and desirability. It also makes the PCs look huge and clumsy and antique, and completely overdetermined for what's expected of them (which they are).
It would be able to work alone, or in concert with a Mac, so like the iPod, it's a reason to get a Mac without the requirement that you get a Mac. Perhaps Apple can convince some third party to write software to sync with a Windows PC.
Face it: Apple isn't going to out-Dell Dell. Where the market plays to their strengths, they can and do beat the PC makers - soundly, in some cases - so what they have to do is define the low end market in terms favorable to them. This would do that. A Dell wanna-be box would not (and it would cost a whole lot more than $500).
HyperTransport is not at all expensive to implement. That's one of its design goals, and they appear to have succeeded. . .
That is what I was hoping to hear. Thank you.
I reread the post about your little appliance idea. (The middle of page one if someone is looking for it.) I don't know what you have in mind regarding hardware, but my neighbor wanted a good size screen (17 inch CRT) and a standard keyboard. He also needed a hard drive to store the documents he creates and a connection for his printer, to print his word processor output. If your appliance can provide all that, it would be a great product and would suffice for many folks I know. If it is anything less, I'm pretty sure my neighbor would not buy it, nor would I.
Thinking about your idea, if it is an accessory for broadband internet service it does not need a CD-ROM drive. It only needs an Ethernet port and two USB ports. Since it doesn't have a conventional OS, Apple could have a basic set of free applications, on the hard drive and available on line. Obviously, these would only work for the appliance. There needs to be email, an internet browser and a simple word processor minimum. If the hard drive crashes, there must be a built in way to download another copy of the appliance's simple OS from Apple. This appliance might sell with six months free .Mac usage, for backing up documents and bookmarks.
Hypertransport has the potential to lower system costs. I do not believe that potential has been reached with the current 970 / Northbridge combo.
For Apple to obtain maximum savings through Hypertransport, I suspect that they will have to implement the port on the CPU die. You also have the cost of the other Hypertransport based components that Apple uses on the G5 motherboard, I simply don't know what Apple is pay for these. They could be cheaper than the PCI equivalent or not hard to say other than they will be in the future.
Thanks
dave
Quote:
Originally posted by snoopy
I have a feeling you are right. Except for lack of AltiVec, Gobi would have been a better choice than a Motorola G4 in every way, battery life (as you point out), performance (1 MB L2 cache), power (under 10 Watts) and price, most likely.
I've been wondering about something else regarding a good value, low-end Mac, but I doubt there are many who have an answer to this question. Does HyperTransport add cost to a system? That is, how does its cost compare with the standard way of connecting system components, before the G5? I was wondering whether it made sense to use HyperTransport in low-end products?
. . . I have to continue to go against the Dell-with-an-Apple-badge thinking going on here. . .
Many here feel the same and express it in many way. It's often stated that a Mac must meet certain standards, usually involving quality, I/O ports and performance. Anything less is not a Mac, would hurt Apple's reputation and would take too many sales away from better Macs. Performance of a really cheap Windows PC is also cited as a major source of customer dissatisfaction. Such points all have merit, which makes it really difficult to offer an opposing view. I will try, because I believe there are other issues that are equally important, if not more so.
Apple must begin to increase market share significantly, and a low priced entry level computer is a good way, maybe the only way, to do it and do it noticeably so. A definite upward trend in market share builds confidence, and Apple's developers, business partners and potential customers need this confidence. If market share stays where it is, or moves up very weakly and inconsistently, it will begin to erode confidence in Apple. I believe this point alone is reason enough for Apple to offer a low price computer that runs Mac OS X. The exact price is debatable, but the oft stated "under $500 US without monitor" is a good mark in the sand.
Let's consider the argument that a low cost entry level Mac can breed dissatisfaction and hurt Apple's image, as customers try to get by cheap and do not consider future needs. As time goes by they get dissatisfied because their computer cannot do more. Is that so bad? What happens now in the Windows world? Do unhappy customers begin to bash the Windows PC and switch to a Mac? No! Usually they calm down and realize it was their own mistake. When they finally do buy a better computer, it will be one that runs all the new software applications they just bought. If my neighbor ever gets interested in doing more, he will surely buy a much better Windows PC, and pay more than for the eMac he could have purchased a few weeks ago.
Within the last two years, I've know several first time PC buyers who never seriously considered a Mac because of price. Also, two families I know very well switched from Macs to a Windows. Their old Performa 630 and 6400 got replaced with a Dell and HP. Price was the only issue. None of these folks are dissatisfied with their Windows PC. Their standards are not very high. Why shouldn't Apple get this kind of business and increase market share of the Mac OS X platform? Will an entry level model tarnish the Macintosh name? Apple can choose another name. The only real consideration is how to make a profit on entry level computers and not reduce sales too much on better Macs. Do we think it is impossible to achieve such a goal?
Regarding profit, Apple does not need to meet Dell's prices, but just get closer enough so many will consider Apple. If the Apple product is perceived as being better, or happens to be better looking, many will pay a small difference to get it. Only companies like Dell, who try to sell to everyone, need to have rock bottom prices. Now regarding sales of better Macs, it is these better Macs that people will upgrade to when they wish to do more than their little entry box can do. The entry model needs to be limited, by design. You may argue that many will be satisfied with these limitations and will never buy a better Mac. True! However, these folks will not buy an Apple at all if a low priced model is not available. All those mentioned in the paragraph above are satisfied with their limited Windows PCs. Mac was not an option to them simply because of price.
I've only touched on the consumer market here. A low cost computer to run OS X on the desktop is obviously needed for business and schools too. It's difficult for me to understand what is taking Apple so long to provide it. I'm hoping they are simply getting all their other "duck" in a row before adding this one. Technology is now making this "duck" easier to achieve than ever. There are now video chips for $10 and a low cost Gobi, 750GX, from IBM. If it's not a Mac, and it's limited, maybe it does not need a G4. Anyone for iPippin or ePippin?
Ok, I have a crazy thought. What if Apple got out of the low-end market all together. What if they do a HP-iPod deal with someone. This company would produce under a very controlled license discounted mac clones. Apple would to some extent would guarantee compatibility with OS X and the producer would guarantee a certain level of quality. The company would be limited in the feature set they could offer under the terms of the license and Apple would have some input as to appearence.
Ok, I have a crazy thought. What if Apple got out of the low-end market all together. What if they do a HP-iPod deal with someone. This company would produce under a very controlled license discounted mac clones. Apple would to some extent would guarantee compatibility with OS X and the producer would guarantee a certain level of quality. The company would be limited in the feature set they could offer under the terms of the license and Apple would have some input as to appearence.
Would this type of deal work?
Would any company go for it?
Would any one here buy one?
You can make a lot of business arrangements work with good contracts. The question is whether it is worth the effort of keeping it under control. Is there a benefit that out weighs the problems? If I understand your proposal, you suggest that Apple give the entire low-end Mac OS X market to a clone maker, for a fee. Now two companies must make a profit on the cheapest of all models. Hard to do. Strike one. Also, Apple and the clone maker would have conflicting goals. Apple will want to limit low end models so better Macs keep selling well, but a clone maker will want to exploit low end models to make as many sales as possible. Strike two. It could be make to work, so it isn't out, but it would cause endless headaches in my opinion.
Comments
Originally posted by snoopy
If it's not too far off topic, I'd like to hear some opinions about why Apple selected a 1 GHz G4 with 256K of L2 cache for the iBook, rather than going with the IBM 750 GX, the G3 Gobi that everyone was looking forward to. It is also 1 GHz but with 1 M L2 cache, four times as much, and dissipates less than 10 W. It likely costs less and would have been an easy and good value upgrade for the iBook, which was already a G3. The iBook could have moved to a G4 when the IBM Mojave is here. I'm pretty sure Mojave will be lower cost than a Motorola G4, probably significantly so.
Apple's decision to ignore the Gobe has puzzled me, but evidently no one else. The only comment I've seen are like, "Great, the iBook has a G4." With the larger L2 cache, I would think the G3 would perform as well as the G4 overall. Gobe would be faster on everything except AltiVec accelerated applications. From a value standpoint, it looks to me like Apple jumped to a G4 too soon. Comments?
They might have had the G4 ibook waiting in the wings for a while. This might explain why they stuck to the Moto G4.. There might be other reasons too.. Maybe the Gobe kicks the crap out of the cpu's in the powerbook's.. Without having any realworld benchmarks I can only guess. I think MOTO let Apple down in such a way that it has held them back with the exception of the G5 tower. Like I said before IBM seem to be working overtime to give them the goods...
Sideshow
From my perspective the ibook lost a bit with the G4. I would have liked to seen battery life as the overwhelming design consideration. I believe it is a markatable a feature as ultimate performance, and frankly is one of the reasons for the iBooks success. But I do understand Apple a bit with respect to this issue. When Alt Vec can be used it is almost always a significant advantage that the G3 series had no way of overcoming.
Thanks
Dave
Originally posted by snoopy
If it's not too far off topic, I'd like to hear some opinions about why Apple selected a 1 GHz G4 with 256K of L2 cache for the iBook, rather than going with the IBM 750 GX, the G3 Gobi that everyone was looking forward to. It is also 1 GHz but with 1 M L2 cache, four times as much, and dissipates less than 10 W. It likely costs less and would have been an easy and good value upgrade for the iBook, which was already a G3. The iBook could have moved to a G4 when the IBM Mojave is here. I'm pretty sure Mojave will be lower cost than a Motorola G4, probably significantly so.
Apple's decision to ignore the Gobe has puzzled me, but evidently no one else. The only comment I've seen are like, "Great, the iBook has a G4." With the larger L2 cache, I would think the G3 would perform as well as the G4 overall. Gobe would be faster on everything except AltiVec accelerated applications. From a value standpoint, it looks to me like Apple jumped to a G4 too soon. Comments?
Originally posted by wizard69
Alt-Vec has become very important to the applicaitons that Apple wants to deliver. I believe this is the overriding consideration. If Gobi had a vector capability it probally would be in the ibook right now. . .
I have a feeling you are right. Except for lack of AltiVec, Gobi would have been a better choice than a Motorola G4 in every way, battery life (as you point out), performance (1 MB L2 cache), power (under 10 Watts) and price, most likely.
I've been wondering about something else regarding a good value, low-end Mac, but I doubt there are many who have an answer to this question. Does HyperTransport add cost to a system? That is, how does its cost compare with the standard way of connecting system components, before the G5? I was wondering whether it made sense to use HyperTransport in low-end products?
I have to continue to go against the Dell-with-an-Apple-badge thinking going on here. The more I think about my little appliance, the more I like it.
- It can be integrated to a degree that a PC can't, adding "sex appeal" and reducing cost (expansion through PCI adds cost).
- Since it's designed for light, general use, it's better at that than the PCs are, and it can come near the advertised cost for the PC box alone, except that you get a whole system.
- It's not a Mac, so it carries none of the baggage that the Mac carries. It is an Apple, so it has all the cachet that the Apple name has.
- Dell can't compete with it. Their whole model is based on the idea that the customer knows going in that they're going to get a bog standard tower and monitor that are identical to anyone else's except for little plastic bits here and there, so ordering them blind is no big deal. But accessories have to be seen first. This is the genius of iPod, and one reason why it's outselling the Dell DJ by about 20:1. Apple has turned their dependence on retail into an asset, and Dell's direct advantage into a liability.
- No PC maker can compete with it (not even Sony, because they have to design for Windows CE, or PocketPC, or whatever MS is calling that failure this week). It plays to all of Apple's strengths: Engineering, integration, design, and desirability. It also makes the PCs look huge and clumsy and antique, and completely overdetermined for what's expected of them (which they are).
- It would be able to work alone, or in concert with a Mac, so like the iPod, it's a reason to get a Mac without the requirement that you get a Mac. Perhaps Apple can convince some third party to write software to sync with a Windows PC.
Face it: Apple isn't going to out-Dell Dell. Where the market plays to their strengths, they can and do beat the PC makers - soundly, in some cases - so what they have to do is define the low end market in terms favorable to them. This would do that. A Dell wanna-be box would not (and it would cost a whole lot more than $500).Originally posted by Amorph
HyperTransport is not at all expensive to implement. That's one of its design goals, and they appear to have succeeded. . .
That is what I was hoping to hear. Thank you.
I reread the post about your little appliance idea. (The middle of page one if someone is looking for it.) I don't know what you have in mind regarding hardware, but my neighbor wanted a good size screen (17 inch CRT) and a standard keyboard. He also needed a hard drive to store the documents he creates and a connection for his printer, to print his word processor output. If your appliance can provide all that, it would be a great product and would suffice for many folks I know. If it is anything less, I'm pretty sure my neighbor would not buy it, nor would I.
Thinking about your idea, if it is an accessory for broadband internet service it does not need a CD-ROM drive. It only needs an Ethernet port and two USB ports. Since it doesn't have a conventional OS, Apple could have a basic set of free applications, on the hard drive and available on line. Obviously, these would only work for the appliance. There needs to be email, an internet browser and a simple word processor minimum. If the hard drive crashes, there must be a built in way to download another copy of the appliance's simple OS from Apple. This appliance might sell with six months free .Mac usage, for backing up documents and bookmarks.
Just a thought.
For Apple to obtain maximum savings through Hypertransport, I suspect that they will have to implement the port on the CPU die. You also have the cost of the other Hypertransport based components that Apple uses on the G5 motherboard, I simply don't know what Apple is pay for these. They could be cheaper than the PCI equivalent or not hard to say other than they will be in the future.
Thanks
dave
Originally posted by snoopy
I have a feeling you are right. Except for lack of AltiVec, Gobi would have been a better choice than a Motorola G4 in every way, battery life (as you point out), performance (1 MB L2 cache), power (under 10 Watts) and price, most likely.
I've been wondering about something else regarding a good value, low-end Mac, but I doubt there are many who have an answer to this question. Does HyperTransport add cost to a system? That is, how does its cost compare with the standard way of connecting system components, before the G5? I was wondering whether it made sense to use HyperTransport in low-end products?
Originally posted by Amorph
. . . I have to continue to go against the Dell-with-an-Apple-badge thinking going on here. . .
Many here feel the same and express it in many way. It's often stated that a Mac must meet certain standards, usually involving quality, I/O ports and performance. Anything less is not a Mac, would hurt Apple's reputation and would take too many sales away from better Macs. Performance of a really cheap Windows PC is also cited as a major source of customer dissatisfaction. Such points all have merit, which makes it really difficult to offer an opposing view. I will try, because I believe there are other issues that are equally important, if not more so.
Apple must begin to increase market share significantly, and a low priced entry level computer is a good way, maybe the only way, to do it and do it noticeably so. A definite upward trend in market share builds confidence, and Apple's developers, business partners and potential customers need this confidence. If market share stays where it is, or moves up very weakly and inconsistently, it will begin to erode confidence in Apple. I believe this point alone is reason enough for Apple to offer a low price computer that runs Mac OS X. The exact price is debatable, but the oft stated "under $500 US without monitor" is a good mark in the sand.
Let's consider the argument that a low cost entry level Mac can breed dissatisfaction and hurt Apple's image, as customers try to get by cheap and do not consider future needs. As time goes by they get dissatisfied because their computer cannot do more. Is that so bad? What happens now in the Windows world? Do unhappy customers begin to bash the Windows PC and switch to a Mac? No! Usually they calm down and realize it was their own mistake. When they finally do buy a better computer, it will be one that runs all the new software applications they just bought. If my neighbor ever gets interested in doing more, he will surely buy a much better Windows PC, and pay more than for the eMac he could have purchased a few weeks ago.
Within the last two years, I've know several first time PC buyers who never seriously considered a Mac because of price. Also, two families I know very well switched from Macs to a Windows. Their old Performa 630 and 6400 got replaced with a Dell and HP. Price was the only issue. None of these folks are dissatisfied with their Windows PC. Their standards are not very high. Why shouldn't Apple get this kind of business and increase market share of the Mac OS X platform? Will an entry level model tarnish the Macintosh name? Apple can choose another name. The only real consideration is how to make a profit on entry level computers and not reduce sales too much on better Macs. Do we think it is impossible to achieve such a goal?
Regarding profit, Apple does not need to meet Dell's prices, but just get closer enough so many will consider Apple. If the Apple product is perceived as being better, or happens to be better looking, many will pay a small difference to get it. Only companies like Dell, who try to sell to everyone, need to have rock bottom prices. Now regarding sales of better Macs, it is these better Macs that people will upgrade to when they wish to do more than their little entry box can do. The entry model needs to be limited, by design. You may argue that many will be satisfied with these limitations and will never buy a better Mac. True! However, these folks will not buy an Apple at all if a low priced model is not available. All those mentioned in the paragraph above are satisfied with their limited Windows PCs. Mac was not an option to them simply because of price.
I've only touched on the consumer market here. A low cost computer to run OS X on the desktop is obviously needed for business and schools too. It's difficult for me to understand what is taking Apple so long to provide it. I'm hoping they are simply getting all their other "duck" in a row before adding this one. Technology is now making this "duck" easier to achieve than ever. There are now video chips for $10 and a low cost Gobi, 750GX, from IBM. If it's not a Mac, and it's limited, maybe it does not need a G4. Anyone for iPippin or ePippin?
Would this type of deal work?
Would any company go for it?
Would any one here buy one?
Originally posted by oldmacfan
Ok, I have a crazy thought. What if Apple got out of the low-end market all together. What if they do a HP-iPod deal with someone. This company would produce under a very controlled license discounted mac clones. Apple would to some extent would guarantee compatibility with OS X and the producer would guarantee a certain level of quality. The company would be limited in the feature set they could offer under the terms of the license and Apple would have some input as to appearence.
Would this type of deal work?
Would any company go for it?
Would any one here buy one?
You can make a lot of business arrangements work with good contracts. The question is whether it is worth the effort of keeping it under control. Is there a benefit that out weighs the problems? If I understand your proposal, you suggest that Apple give the entire low-end Mac OS X market to a clone maker, for a fee. Now two companies must make a profit on the cheapest of all models. Hard to do. Strike one. Also, Apple and the clone maker would have conflicting goals. Apple will want to limit low end models so better Macs keep selling well, but a clone maker will want to exploit low end models to make as many sales as possible. Strike two. It could be make to work, so it isn't out, but it would cause endless headaches in my opinion.