Hey, how about that. IBM is in third place this year.
</strong><hr></blockquote>
I'm not an economics expert, nor care to be, but your referencing market share for UNIX, right? I'm under the impression, maybe mistakenly, that IBM is concentrating on LINUX.
IBM is still maintains the largest market share for servers, right??
from the article [quote]"In the overall server market, IBM gained 1.3 percent for a total of 29.3 percent, Gartner reported, with Sun running a close second at 23.6 percent."<hr></blockquote>
A more telling quote from the article might be;
[quote]"IBM also secured the top spot in the overall Linux server market, gaining a whopping 19 percent to reach 34.4 percent of the $236 million market, which Gartner said grew 78.9 percent despite shrinkage in the overall server market."<hr></blockquote>
Let's see the LINUX market grew 78.9% despite an overall shrink in the server market. And the UNIX market is shrinking according to the article. [quote]"According to Gartner, the UNIX server market accounted for 40 percent of the US$4.3 billion in total server sales in the United States, making it the largest single segment despite a sales decline of 5.3 percent, from $1.8 billion to $1.7 billion."<hr></blockquote>
Please correct me if I'm wrong, high probability I am, but IBM is focusing on a growing market, LINUX, capturing the largest market share. IBM is maintaining a solid market share in a declining UNIX market and begining to show serious interest in the low end server market.
Oh great googly moogly, after all this I forgot what your point was, could you clarify what your arguement against IBM is?
Methinks you like to argue for the sake of arguing.
Back on topic
One part of the article I have trouble with is this quote [quote]"... whereby Motorola may not be able to push the G4 above 1.3 Ghz, which would be a paltry 50 Mhz speed bump with the upcoming 7457 G4.<hr></blockquote>
Wouldn't a die shrink to 0.13µnm in and of itself provide a significantly higher increase in Mhz than a "paltry 50 Mhz?? I was under the impression a die shrink to 0.13µm would provide up to a 30% increase in speed??
<strong>Wouldn't a die shrink to 0.13µnm in and of itself provide a significantly higher increase in Mhz than a "paltry 50 Mhz?? I was under the impression a die shrink to 0.13µm would provide up to a 30% increase in speed??</strong><hr></blockquote>
I'm not sure what their deal is, but we should be seeing 1.2GHz to 1.6GHz from the die shrink.
<strong>"....all we want is an insanely powerful killer computer in a killer designed unit that is so fast we have to find ways of slowing it down just to be able to use it."
Amen to that!!!
They need to be so fast it'll give you a nosebleed just trying to comprehend the insane speed.
Just wishful thinking.
I know it will never happend. Well be stuck with the G4 until 2004 and then Apple will go chapter eleven. Just my prediction <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> </strong><hr></blockquote>
Firelark, you have the nail well and truly hit on the head there.
The bottomline is that you are are trying to pass opinion from an opinion site as an authority. Without question, you share the opinion expressed in the piece, but that does not make it fact.</strong><hr></blockquote>
That Sun is #1 in the Unix market is fact, not opinion. This information comes directly from Gartner. The opinions I express about this data are my own.
If you could refer to my actual posts and point out where I've inserted opinion over fact, then please do so. It's posts like these that lead topics astray. Since you're obviously just trying to discredit my facts with a bunch of nonsense, you're also harassing me as well. I don't appreciate you wasting my time.
[[[Sure you can. Considering that's not the data I was referring to, it's pretty cloudy. Why don't you just post the stats for the Dallas Cowboys while you're at it? It may prove just as useful.]]]
This is an clear example of "Slothful Induction" and also an example of the fallacy of "Nothing but objections". Look them up for yourselves and see if they don't apply.
[[[I was replying to "consultant boy"--who apparently "charges" people to hit the Google search button and post the first piece of evidence he finds that supports his argument. ]]]
And how do *you* conduct your research?
[[[You want to look for market penetration specs for servers sold by unit.]]]
That's pretty good... How many Itanium servers did Intel actually sell to date?
Oh, and Mr. Me... Regarding your previous post... EXCELLENT. MacLuv is again cherry picking. he claims people of doing what he himself does... He runs to Google and clicks on the first link that seems to support his claim. In this case his source is a joke.... ;-)
I'm not an economics expert, nor care to be, but your referencing market share for UNIX, right? I'm under the impression, maybe mistakenly, that IBM is concentrating on LINUX.
</strong><hr></blockquote>
First of all, I appreciate that you've done some homework and looked up some facts. But you're taking my main point and trying to relate it to irrelevant data.
Look guys, I really don't have the time and sit down and illustrate this for you. I've given you a lot to go on already.
The whole point of getting into servers and marketshare, etc, was to talk about Apple's market penetration into the server world. Apple would be classified under UNIX, not Linux. Furthermore, the data that everyone is pulling only illustrates why I don't think going IBM is a good idea for Apple--at the very least for Xserve. I'll post my case when I finish it. Until then, this thread has gone so far off topic it needs to be closed.
My source is Gartner research. It has always been Gartner research. Are you suggesting they are a joke? I would think that your appeal to popular opinion does nothing to help you make a case against my arguments. But as you are constantly making a case against me and not my ideas, I would say that you are baiting me for the sake of your own entertainment.
Mr. Ed, the facts speak for themselves. If you do not wish to participate constructively in this thread, don't bother posting, because all you're doing is harrasing me.
This is an clear example of "Slothful Induction" and also an example of the fallacy of "Nothing but objections". Look them up for yourselves and see if they don't apply.
</strong><hr></blockquote>
It's not a matter of slothful induction.
I supported by objection with relevant data supporting my objection.
[quote]
<strong>
[[[I was replying to "consultant boy"--who apparently "charges" people to hit the Google search button and post the first piece of evidence he finds that supports his argument. ]]]
And how do *you* conduct your research?
</strong><hr></blockquote>
The original poster was using his "credentials" to act as a leading authority on the subject. In this case he used old, irrelivant data to support his argument. Obviously he is not a leading authority.
How I conduct research is unimportant. What is important is that the information I use to support my arguments is as factual and accurate as it can be. If you're so big on the art of debate, you will find that you're trying to *turn the table* on me, or rather try to discredit me by questioning my authority. If we were in an actual debate, you would be laughed at, as I have never put forth any claim to be an official authority by telling people I get paid for my services. Like, who cares? It's obvious the guy's not a consultant--unless he consults people for talkin' *trash*--which makes me think you're probably his biggest client.
How would this data be imporant to you for determining Apple's market penetration? Could you elaborate?
[quote]<strong>
Oh, and Mr. Me... Regarding your previous post... EXCELLENT. MacLuv is again cherry picking. he claims people of doing what he himself does... He runs to Google and clicks on the first link that seems to support his claim. In this case his source is a joke.... ;-)
</strong><hr></blockquote>
As previously stated, the facts speak for themselves.
Practice your bad debating styles someplace else, like--in politics. If you get good enough, maybe you can run for the Mayor of Uranus.
When you do post this post you're promising, be sure to say what you mean. "The server market" is a lot different from "the UNIX market" or "the UNIX server market," or "the market that ships with UNIX and UNIX-like OS'." If you take the stance that neither Linux or BSD is a proper UNIX (which they aren't), then the BSDish OS X doesn't qualify either. So if you look at someone measuring "the UNIX market" you have to check what they mean by "UNIX."
Also, restricting the analysis to "the UNIX market" comes with a price: Nobody is locked into neat categories, because it's common for servers to run different OS' than all of the machines around them, and most server OS' overlap heavily in terms of what they are capable of. The high end, SVR4-based UNIX market - SGI's IRIX, Sun's Solaris, IBM's AIX, HPaq's Tru64 and HP/UX - is bleeding share, which is going to Windows, Linux and BSD servers. There is also a long-standing and accelerating move from Windows to Linux and BSD. The free UNIX-like OS' are hard to track, because they are frequently installed after the fact on servers that shipped with another OS, and on machines that weren't considered servers when they were shipped: lots of cast-off desktops get repurposed for file, print, and intranet web serving duties.
So when you're looking at how Apple is doing in this market, you have to consider that the server market is very fluid right now, and realigning itself across categories. Things are happening that nobody would have predicted a few years ago, like a resurgence in sales of IBM's mainframe lines. Mainframes were dismissed as dinosaurs, but they're coming back. VMS, which was nearly cancelled a few years ago, now has a sizable chunk of the uptime-critical market, and is enjoying a resurgence as a database platform.
Also, server customers are conservative. Unproven technology and unestablished companies are treated skeptically, because reliability, uptime, and support are not optional in this space. Apple came into the fastest-growing segment of the server market (the low end, UNIX-like/Windows segment) with hardware that had never seen the light of day before backed up by an enterprise support structure that literally hadn't existed the year before, and stamped with a brand that was actively scorned until very recently. Given that, with a few exceptions (mostly in biotech and the sciences, where Apple has a much better reputation), curious IT people are buying an Xserve or two, setting them to non-critical tasks, and watching them. In other words, current sales reflect early adopters willing to try this new thing out despite all the strikes against it (none of which, notice, have anything to do with the Xserve's potential as a server). If the word over time from these early adopters is that the Xserve is compatible, easy to maintain and stable - and that OS X isn't a damn thing like MacOS in the ways that IT people care about - then you'll see sales spike as the wait-and-see, conservative customers (the bulk of the server market) adopt it in earnest and set it to real work. If not, well, no spike in sales, and Apple goes back to the drawing board.
Those who favor an x86 processor for the Mac have some good arguments, but they usually lack one quality, vision. An AMD or Intel processor may seem like the best choice, looking at how things were in the past, and how they are in the present. But it can be a big mistake thinking things always stay the same. Leaders do not remain leaders forever, and the most popular of anything will usually change at some point. Many of us believe, and for good reason, that the processor scene is changing, and the IBM 970 is the best possible way for Apple to go today.
PPC desktop processors suffered in the last two years, and Macs suffered because of it. Motorola has been interested in the embedded market and has not done much to improve desktop processors for Apple. IBM was busy with things like the Power 4 for their big servers, and did not see the benefits of an SIMD engine either. But things show strong indications of change. Those glued to the past and present may miss it.
IBM is getting in the business of making custom chips. They made one for Nintendo to use in the Game Cube. They may be doing one for Sony, for the Playstation 3, and have stated they intend to pursue this market. Apple is not limited to tagging along with IBM's need for processors. If Apple wants a chip that IBM does not intend to build for themselves, IBM will build Apple a custom chip. But the future appears brighter yet now that IBM has their own need for that same processor that Apple wants. IBM now has that extra motivation to win with the best.
One of the biggest threats to IBM server business is Linux running on x86 processors. Within three years, this type of server will likely be making big inroads in markets that typically use big servers now. IBM could just market their own x86 Linux server and call it good. But if IBM followed this approach, they would be competing head to head with Dell and HP. There would be little to differentiate an IBM Linux server from any other, and IBM could not maintain their lead in the server market. They need to give customers a reason to run Linux on an IBM processor, rather than an Intel or AMD. They need the best processor in this class, and they are out to get it.
So, looking just at the past and present, the x86 looks favorable. With a vision of the future, it is a different story. If anything should be debated, it should be whether IBM is capable of beating Intel and AMD, when IBM is committed to this task. I feel sure of IBM's commitment, because servers are paramount to their business.
<strong>Those who favor an x86 processor for the Mac have some good arguments, but they usually lack one quality, vision. An AMD or Intel processor may seem like the best choice, looking at how things were in the past, and how they are in the present. But it can be a big mistake thinking things always stay the same. Leaders do not remain leaders forever, and the most popular of anything will usually change at some point. Many of us believe, and for good reason, that the processor scene is changing, and the IBM 970 is the best possible way for Apple to go today.
</strong><hr></blockquote>
This is a great assessment of the industry's current direction. I think that the IBM/Apple combination has potential and a lot of it, if for no other reason than timing.
First off, Intel is saying that x86 Xeon is not where it's at on the server side of things. So they've got Itannic and McKinley. As for the desktop, Intel says hey, we think the world will be happy with x86 for a while longer so we will continue to sell Pentium(32 bit processor) for the desktop.
AMD says wrong. We think there is a market for 64bitness on the desktop and so the Hammer family will be backwards compatible with a desktop implementation in mind.
IBM interestingly agrees with AMD and says yes the world is ready for a 64 bit desktop processor.
As I intimated in my last post, Linux is only run on Pentiums because it's what most geeks can afford. However, I'd bet that anyone that truly appreciates Linux probably still has wet dreams about the legendary Alpha. I don't think any of us can deny that Unix and RISC processors is just the way of things. Hence why IBM is betting on the 970 and Linux. Assuming IBM builds and affordable systems around the 970 with Linux, I'm guessing many geeks might begin to pony up.
Obviously, if Unix on RISC is the natural order, OSX/BSD/Darwin all make sense on the 970.
But wait, there's one more thing. I admit that I could be wrong on this, but I believe that Intel and AMD are just getting into 64 bit market. This is not the case for IBM. The Power4 is a 64 bit chip. Thus, Apple is partnering with a company that has quite a bit of experience in the field. After all, the 970 is a slimmed down version of the only other chip that can be spoken of in the same phrase as the Alpha.
As far as the 64bit desktop market, for the next two years or so we are talking IBM/Apple and AMD.
What Intel was thinking when they left this gaping hole in there processor strategy is anyone's guys. When considering AMD's marriage of the 32 and 64 bit ISAs in the Hammer family so the customer can move and migrate at their choosing vs. Intels all or nothing approach, you have to wonder if Intel was thinking at all.
I can imagine that Intel might ultimately regret leaving the 64bit desktop market to IBM/Apple for any length of time.
so you're saying that Apple is entering the PC server market?</strong><hr></blockquote>
Duh, no. Low end server market. As in cheaper than $10.000 servers. Anything from IBM to x86 based servers go here, but those are not PCs. If Amorph would mean PC Server market, he would say "PC Server market".
Duh, no. Low end server market. As in cheaper than $10.000 servers. Anything from IBM to x86 based servers go here, but those are not PCs. If Amorph would mean PC Server market, he would say "PC Server market".</strong><hr></blockquote>
i didn't realize you were amorph's personal spokesperson.
I don't even know what a low-end server market is. Is that the entry level market that everyone with a clue talks about? Because that includes servers ~ $100,000.
Intel himself think that the X86 series will not last forever. The Prescott seems to be the lattest developpement of the Pentium and the end of the developpement of the pentium : but it give them three years of developpement in front of them. After this they have to move to the 64 bit world. Even if 64 bits chips are not much better than a 32 bits chip, everydesktop chip will be 64 bits.
Intel know that but did not wanted to make a 64 bit X86 32 bit compatible chip at the contrary of AMD and his opteron, he wanted to produce a radically new design with more room for progress in the future : the itanium.
But the itanium did not reaches the result expected and was a partial failure. Since this they produced the itanium 2.
As every geek here, you know that the current X86 chips are more RISC than CISC : rughly they are RISC chip with a X86 decoder in the front end. And this X86 decoder is complex and leads to a waste of performance and transistors.
You will notice also that hardware X86 chips decoder are much more performant than software decoder (emulator) : that's why the itanium was not really good.
Intel at the contrary of Apple (with his transition from 68K to PPC)canno't force the market to move from the X86 code to the Itanium one : he is not alone, AMD is here. And AMD has a magic card : is opteron is backward compatible. The opteron waste power and transistors to be backward compatible, but the current software will run great on him at the cost of minimal code optimisation.
In short the sofware world is not ready for the Itanium reign. Windows must make disapear entirely the DOS (and not only hide him) and make several improvements. So Intel is obliged to wait. But he is still active and due to his fantastic team of engineers his latest pentium 4 rocks.
The PPC 970 chips have a great future in front of them, with many rooms to improve them : better core (power 5 variant), multicore ...
If the X86 are currently better it's because of the great teams of engineers who develop them, not because the X86 code is inherently better : it's just the contrary.
Let's see in 3 years who will have the best chip : IBM or INTEL with his itanium 3 or with his Pentium 5 ? .
Comments
<strong>This is what happens when I'm bored.
<a href="http://www.newsfactor.com/perl/story/17758.html"; target="_blank">*** EL LINKO ***</a>
Hey, how about that. IBM is in third place this year.
</strong><hr></blockquote>
I'm not an economics expert, nor care to be, but your referencing market share for UNIX, right? I'm under the impression, maybe mistakenly, that IBM is concentrating on LINUX.
IBM is still maintains the largest market share for servers, right??
from the article [quote]"In the overall server market, IBM gained 1.3 percent for a total of 29.3 percent, Gartner reported, with Sun running a close second at 23.6 percent."<hr></blockquote>
A more telling quote from the article might be;
[quote]"IBM also secured the top spot in the overall Linux server market, gaining a whopping 19 percent to reach 34.4 percent of the $236 million market, which Gartner said grew 78.9 percent despite shrinkage in the overall server market."<hr></blockquote>
Let's see the LINUX market grew 78.9% despite an overall shrink in the server market. And the UNIX market is shrinking according to the article. [quote]"According to Gartner, the UNIX server market accounted for 40 percent of the US$4.3 billion in total server sales in the United States, making it the largest single segment despite a sales decline of 5.3 percent, from $1.8 billion to $1.7 billion."<hr></blockquote>
Please correct me if I'm wrong, high probability I am, but IBM is focusing on a growing market, LINUX, capturing the largest market share. IBM is maintaining a solid market share in a declining UNIX market and begining to show serious interest in the low end server market.
Oh great googly moogly, after all this I forgot what your point was, could you clarify what your arguement against IBM is?
Methinks you like to argue for the sake of arguing.
Back on topic
One part of the article I have trouble with is this quote [quote]"... whereby Motorola may not be able to push the G4 above 1.3 Ghz, which would be a paltry 50 Mhz speed bump with the upcoming 7457 G4.<hr></blockquote>
Wouldn't a die shrink to 0.13µnm in and of itself provide a significantly higher increase in Mhz than a "paltry 50 Mhz?? I was under the impression a die shrink to 0.13µm would provide up to a 30% increase in speed??
<strong>
Hey, it twitched!</strong><hr></blockquote>
Some life in 'er yet, 'eh!
<strong>Wouldn't a die shrink to 0.13µnm in and of itself provide a significantly higher increase in Mhz than a "paltry 50 Mhz?? I was under the impression a die shrink to 0.13µm would provide up to a 30% increase in speed??</strong><hr></blockquote>
I'm not sure what their deal is, but we should be seeing 1.2GHz to 1.6GHz from the die shrink.
<strong>"....all we want is an insanely powerful killer computer in a killer designed unit that is so fast we have to find ways of slowing it down just to be able to use it."
Amen to that!!!
They need to be so fast it'll give you a nosebleed just trying to comprehend the insane speed.
Just wishful thinking.
I know it will never happend. Well be stuck with the G4 until 2004 and then Apple will go chapter eleven. Just my prediction <img src="graemlins/oyvey.gif" border="0" alt="[No]" /> </strong><hr></blockquote>
Firelark, you have the nail well and truly hit on the head there.
<strong>
The bottomline is that you are are trying to pass opinion from an opinion site as an authority. Without question, you share the opinion expressed in the piece, but that does not make it fact.</strong><hr></blockquote>
That Sun is #1 in the Unix market is fact, not opinion. This information comes directly from Gartner. The opinions I express about this data are my own.
If you could refer to my actual posts and point out where I've inserted opinion over fact, then please do so. It's posts like these that lead topics astray. Since you're obviously just trying to discredit my facts with a bunch of nonsense, you're also harassing me as well. I don't appreciate you wasting my time.
This is an clear example of "Slothful Induction" and also an example of the fallacy of "Nothing but objections". Look them up for yourselves and see if they don't apply.
[[[I was replying to "consultant boy"--who apparently "charges" people to hit the Google search button and post the first piece of evidence he finds that supports his argument. ]]]
And how do *you* conduct your research?
[[[You want to look for market penetration specs for servers sold by unit.]]]
That's pretty good... How many Itanium servers did Intel actually sell to date?
Oh, and Mr. Me... Regarding your previous post... EXCELLENT. MacLuv is again cherry picking. he claims people of doing what he himself does... He runs to Google and clicks on the first link that seems to support his claim. In this case his source is a joke.... ;-)
--
Ed M.
<strong>
I'm not an economics expert, nor care to be, but your referencing market share for UNIX, right? I'm under the impression, maybe mistakenly, that IBM is concentrating on LINUX.
</strong><hr></blockquote>
First of all, I appreciate that you've done some homework and looked up some facts. But you're taking my main point and trying to relate it to irrelevant data.
Look guys, I really don't have the time and sit down and illustrate this for you. I've given you a lot to go on already.
The whole point of getting into servers and marketshare, etc, was to talk about Apple's market penetration into the server world. Apple would be classified under UNIX, not Linux. Furthermore, the data that everyone is pulling only illustrates why I don't think going IBM is a good idea for Apple--at the very least for Xserve. I'll post my case when I finish it. Until then, this thread has gone so far off topic it needs to be closed.
<img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
<strong>
In this case his source is a joke.... ;-)
</strong><hr></blockquote>
My source is Gartner research. It has always been Gartner research. Are you suggesting they are a joke? I would think that your appeal to popular opinion does nothing to help you make a case against my arguments. But as you are constantly making a case against me and not my ideas, I would say that you are baiting me for the sake of your own entertainment.
Mr. Ed, the facts speak for themselves. If you do not wish to participate constructively in this thread, don't bother posting, because all you're doing is harrasing me.
[quote]Originally posted by Ed M.:
<strong>
This is an clear example of "Slothful Induction" and also an example of the fallacy of "Nothing but objections". Look them up for yourselves and see if they don't apply.
</strong><hr></blockquote>
It's not a matter of slothful induction.
I supported by objection with relevant data supporting my objection.
[quote]
<strong>
[[[I was replying to "consultant boy"--who apparently "charges" people to hit the Google search button and post the first piece of evidence he finds that supports his argument. ]]]
And how do *you* conduct your research?
</strong><hr></blockquote>
The original poster was using his "credentials" to act as a leading authority on the subject. In this case he used old, irrelivant data to support his argument. Obviously he is not a leading authority.
How I conduct research is unimportant. What is important is that the information I use to support my arguments is as factual and accurate as it can be. If you're so big on the art of debate, you will find that you're trying to *turn the table* on me, or rather try to discredit me by questioning my authority. If we were in an actual debate, you would be laughed at, as I have never put forth any claim to be an official authority by telling people I get paid for my services. Like, who cares? It's obvious the guy's not a consultant--unless he consults people for talkin' *trash*--which makes me think you're probably his biggest client.
<img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
[quote]<strong>
[[[You want to look for market penetration specs for servers sold by unit.]]]
That's pretty good... How many Itanium servers did Intel actually sell to date?
</strong><hr></blockquote>
<img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />
How would this data be imporant to you for determining Apple's market penetration? Could you elaborate?
[quote]<strong>
Oh, and Mr. Me... Regarding your previous post... EXCELLENT. MacLuv is again cherry picking. he claims people of doing what he himself does... He runs to Google and clicks on the first link that seems to support his claim. In this case his source is a joke.... ;-)
</strong><hr></blockquote>
As previously stated, the facts speak for themselves.
Practice your bad debating styles someplace else, like--in politics. If you get good enough, maybe you can run for the Mayor of Uranus.
<img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
[ 12-07-2002: Message edited by: MacLuv ]</p>
Also, restricting the analysis to "the UNIX market" comes with a price: Nobody is locked into neat categories, because it's common for servers to run different OS' than all of the machines around them, and most server OS' overlap heavily in terms of what they are capable of. The high end, SVR4-based UNIX market - SGI's IRIX, Sun's Solaris, IBM's AIX, HPaq's Tru64 and HP/UX - is bleeding share, which is going to Windows, Linux and BSD servers. There is also a long-standing and accelerating move from Windows to Linux and BSD. The free UNIX-like OS' are hard to track, because they are frequently installed after the fact on servers that shipped with another OS, and on machines that weren't considered servers when they were shipped: lots of cast-off desktops get repurposed for file, print, and intranet web serving duties.
So when you're looking at how Apple is doing in this market, you have to consider that the server market is very fluid right now, and realigning itself across categories. Things are happening that nobody would have predicted a few years ago, like a resurgence in sales of IBM's mainframe lines. Mainframes were dismissed as dinosaurs, but they're coming back. VMS, which was nearly cancelled a few years ago, now has a sizable chunk of the uptime-critical market, and is enjoying a resurgence as a database platform.
Also, server customers are conservative. Unproven technology and unestablished companies are treated skeptically, because reliability, uptime, and support are not optional in this space. Apple came into the fastest-growing segment of the server market (the low end, UNIX-like/Windows segment) with hardware that had never seen the light of day before backed up by an enterprise support structure that literally hadn't existed the year before, and stamped with a brand that was actively scorned until very recently. Given that, with a few exceptions (mostly in biotech and the sciences, where Apple has a much better reputation), curious IT people are buying an Xserve or two, setting them to non-critical tasks, and watching them. In other words, current sales reflect early adopters willing to try this new thing out despite all the strikes against it (none of which, notice, have anything to do with the Xserve's potential as a server). If the word over time from these early adopters is that the Xserve is compatible, easy to maintain and stable - and that OS X isn't a damn thing like MacOS in the ways that IT people care about - then you'll see sales spike as the wait-and-see, conservative customers (the bulk of the server market) adopt it in earnest and set it to real work. If not, well, no spike in sales, and Apple goes back to the drawing board.
[ 12-07-2002: Message edited by: Amorph ]</p>
:cool:
[quote]Apple came into the fastest-growing segment of the server market (the low end, UNIX-like/Windows segment) <hr></blockquote>
so you're saying that Apple is entering the PC server market?
BTW--let's move this to another thread. Then I can concentrate at one thing at a time.
SEGMENTING THE SEVER MARKET.
[ 12-07-2002: Message edited by: MacLuv ]</p>
<strong>oog. What was this thread about again? I've lost track.</strong><hr></blockquote>
Parachute failure.
PPC desktop processors suffered in the last two years, and Macs suffered because of it. Motorola has been interested in the embedded market and has not done much to improve desktop processors for Apple. IBM was busy with things like the Power 4 for their big servers, and did not see the benefits of an SIMD engine either. But things show strong indications of change. Those glued to the past and present may miss it.
IBM is getting in the business of making custom chips. They made one for Nintendo to use in the Game Cube. They may be doing one for Sony, for the Playstation 3, and have stated they intend to pursue this market. Apple is not limited to tagging along with IBM's need for processors. If Apple wants a chip that IBM does not intend to build for themselves, IBM will build Apple a custom chip. But the future appears brighter yet now that IBM has their own need for that same processor that Apple wants. IBM now has that extra motivation to win with the best.
One of the biggest threats to IBM server business is Linux running on x86 processors. Within three years, this type of server will likely be making big inroads in markets that typically use big servers now. IBM could just market their own x86 Linux server and call it good. But if IBM followed this approach, they would be competing head to head with Dell and HP. There would be little to differentiate an IBM Linux server from any other, and IBM could not maintain their lead in the server market. They need to give customers a reason to run Linux on an IBM processor, rather than an Intel or AMD. They need the best processor in this class, and they are out to get it.
So, looking just at the past and present, the x86 looks favorable. With a vision of the future, it is a different story. If anything should be debated, it should be whether IBM is capable of beating Intel and AMD, when IBM is committed to this task. I feel sure of IBM's commitment, because servers are paramount to their business.
[ 12-08-2002: Message edited by: snoopy ]</p>
<strong>
Blah-bla-bla
Look guys, I really don't have the time and sit down and illustrate this for you. I've given you a lot to go on already.
Bla-bla-bla
<img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> </strong><hr></blockquote>
Yes, you've given me a lot to go on...
:cool:
<strong>Those who favor an x86 processor for the Mac have some good arguments, but they usually lack one quality, vision. An AMD or Intel processor may seem like the best choice, looking at how things were in the past, and how they are in the present. But it can be a big mistake thinking things always stay the same. Leaders do not remain leaders forever, and the most popular of anything will usually change at some point. Many of us believe, and for good reason, that the processor scene is changing, and the IBM 970 is the best possible way for Apple to go today.
</strong><hr></blockquote>
This is a great assessment of the industry's current direction. I think that the IBM/Apple combination has potential and a lot of it, if for no other reason than timing.
First off, Intel is saying that x86 Xeon is not where it's at on the server side of things. So they've got Itannic and McKinley. As for the desktop, Intel says hey, we think the world will be happy with x86 for a while longer so we will continue to sell Pentium(32 bit processor) for the desktop.
AMD says wrong. We think there is a market for 64bitness on the desktop and so the Hammer family will be backwards compatible with a desktop implementation in mind.
IBM interestingly agrees with AMD and says yes the world is ready for a 64 bit desktop processor.
As I intimated in my last post, Linux is only run on Pentiums because it's what most geeks can afford. However, I'd bet that anyone that truly appreciates Linux probably still has wet dreams about the legendary Alpha. I don't think any of us can deny that Unix and RISC processors is just the way of things. Hence why IBM is betting on the 970 and Linux. Assuming IBM builds and affordable systems around the 970 with Linux, I'm guessing many geeks might begin to pony up.
Obviously, if Unix on RISC is the natural order, OSX/BSD/Darwin all make sense on the 970.
But wait, there's one more thing. I admit that I could be wrong on this, but I believe that Intel and AMD are just getting into 64 bit market. This is not the case for IBM. The Power4 is a 64 bit chip. Thus, Apple is partnering with a company that has quite a bit of experience in the field. After all, the 970 is a slimmed down version of the only other chip that can be spoken of in the same phrase as the Alpha.
As far as the 64bit desktop market, for the next two years or so we are talking IBM/Apple and AMD.
What Intel was thinking when they left this gaping hole in there processor strategy is anyone's guys. When considering AMD's marriage of the 32 and 64 bit ISAs in the Hammer family so the customer can move and migrate at their choosing vs. Intels all or nothing approach, you have to wonder if Intel was thinking at all.
I can imagine that Intel might ultimately regret leaving the 64bit desktop market to IBM/Apple for any length of time.
<strong>amorph...
so you're saying that Apple is entering the PC server market?</strong><hr></blockquote>
Duh, no. Low end server market. As in cheaper than $10.000 servers. Anything from IBM to x86 based servers go here, but those are not PCs. If Amorph would mean PC Server market, he would say "PC Server market".
<strong>
Duh, no. Low end server market. As in cheaper than $10.000 servers. Anything from IBM to x86 based servers go here, but those are not PCs. If Amorph would mean PC Server market, he would say "PC Server market".</strong><hr></blockquote>
i didn't realize you were amorph's personal spokesperson.
I don't even know what a low-end server market is. Is that the entry level market that everyone with a clue talks about? Because that includes servers ~ $100,000.
[ 12-08-2002: Message edited by: MacLuv ]</p>
Intel know that but did not wanted to make a 64 bit X86 32 bit compatible chip at the contrary of AMD and his opteron, he wanted to produce a radically new design with more room for progress in the future : the itanium.
But the itanium did not reaches the result expected and was a partial failure. Since this they produced the itanium 2.
As every geek here, you know that the current X86 chips are more RISC than CISC : rughly they are RISC chip with a X86 decoder in the front end. And this X86 decoder is complex and leads to a waste of performance and transistors.
You will notice also that hardware X86 chips decoder are much more performant than software decoder (emulator) : that's why the itanium was not really good.
Intel at the contrary of Apple (with his transition from 68K to PPC)canno't force the market to move from the X86 code to the Itanium one : he is not alone, AMD is here. And AMD has a magic card : is opteron is backward compatible. The opteron waste power and transistors to be backward compatible, but the current software will run great on him at the cost of minimal code optimisation.
In short the sofware world is not ready for the Itanium reign. Windows must make disapear entirely the DOS (and not only hide him) and make several improvements. So Intel is obliged to wait. But he is still active and due to his fantastic team of engineers his latest pentium 4 rocks.
The PPC 970 chips have a great future in front of them, with many rooms to improve them : better core (power 5 variant), multicore ...
If the X86 are currently better it's because of the great teams of engineers who develop them, not because the X86 code is inherently better : it's just the contrary.
Let's see in 3 years who will have the best chip : IBM or INTEL with his itanium 3 or with his Pentium 5 ? .