Are you from Norway btw? I suspect you have been my professsor in a few subjects at NTNU. </strong><hr></blockquote>
Car analogies are fine if you understand cars
hP = Processor speed
rpm = FSB speed
Torque = Bandwidth = (hP x rpm)
it's exactly the same.
you can have a 1250 hP ( 1.25 GHz processor) with 13 ft-lb/s of torque (1.3 GB bandwidth) and you won't be able to move the car or data around the machine.
<strong>[[[Apple needs marketshare to keep developers on board. Im sorry, but you can tell me car analogies all day long, and its still not a perfect matching analogy. In the long run, selling 170K PowerMac units/Q is going to hurt Apple.]]]
Yep. That's so short term and now all the developers have going for them is relying on upgrades, milking people with nothing really new. More on that later...</strong><hr></blockquote>
OK, don't take this the wrong way Ed M. - I like you're train of thoughts, but after 6 page downs I have a suggestion for you.
I've noticed in a couple of your posts your method of quoting, and have trouble reading them. Use the reply button directly above the post and it will quote the original for you, and you can chop it up (like above). I like reading your responses, but due to formating and length I have difficulty finishing them. If that's 'your style', cool and disregard.
( Gateway sell more computers than Apple, so Gateway is probably the 6th biggest computer maker, and Apple the 7th.. or maybe IBM is selling more computers than Apple?? )
<strong>I hate to disagree with you moki, but whilst the programme will request memory in small chunks most of the time, the processor will always request a cache line, or 32 Bytes at a time. The main reason why most programmes, which don't stream memory in large blocks, don't see much gain from DDR is that there is almost no change in the latency. (Okay, it amounts to the same thing in the end anyway, but I don't like to see people discussing just bandwidth as if it were everything, and ignoring latency. Just look at RDRAM)
DDR is great for streaming data, and of little benefit for almost everything else.</strong><hr></blockquote>
Your'e quite right -- I had forgotten about the way the cache makes these matters more complicated.
Regardless, the situation remains that DDR in the real world doesn't offer double the performance over SDRAM, despite what the name implies.
<strong>Holy crap PSCATES, great post! You really really hit the nail on the head...
As far as MWSF, there is no way in H that there will be a brand new Powermac...not this soon after a big revision...
And as far as the post about this being the groundwork for the next proc. (ie "yikesv.2), I would say "I hope not". DDR is where the industry is today, they didn't prepare for the future, they just caught up to the present. DDRII is right around the corner as is serialATA and Hypertransport. Why does apple still include UltraATA66 for gods sake, are they really saving that much by not making both controllers ATA100, or even RAID built in considering the market for this thing. Anyhoo, The next chip will hopefully be on an all new platform w/ some future (soon to be current) tech.
Oh yeah, I'm also very happy with the current releases!</strong><hr></blockquote>
They include the ATA/66 controller because they've got bunches of them to get rid of. This saves them money and allows them to have 2 IDE buses, making people think they're getting extra value. Apple has done this before, although with a very crappy result (Performa 6300 anybody?).
Gee, I thought ATA 66 ran burst at about 50MB/s and sustained at 35MB/s, which is faster than most HD will use. The only advantage to ATA 100/133 is the ability to recognize bigger HD's. I don't want a 160 GB HD, I would partion it anyway.
For big Video manipulaters/creators it probably is warranted, but not for most users.
So why not have some 66 and 100 controllers.When faster HD's come out the old ATA66'ers will be going cheap.
Ed's right on this one. The Hewcrapq merger gave Apple a leg up!
Also, note...it was in the news recently about Apple outselling Gateway for the first time in a while. So Apple are 5 or 6th now.
Gateway are having problems competing...but Apple is getting its act together.
It will be interesting if between the Switcher campaign, opening the retail stores, Jag (removing many of the barriers to owner a Mac from a PC viewpoint...) and the problems PC co's are having in competing with Dell...if this gives Apple some joy-ward momentum.
ie I'd like to see where Apple is in terms of sales and market share a year from now.
And...will they count iPod PC sales...and any iphone (couldn't resist...if they make one...) sales as part of market penetration?
Steve Jobs has set his target at 10%.
That would give Apple alot more leverage with developers.
I half take Ed's point about Dell having '14%'. Viewed in those terms...if Apple can get to 10%...then things look far more healthy and interesting from Apple's point of view.
Well, if you bothered to read it you would see that Apple is into the top 5 manufacturers...
From the article:
"Apple jumped into the top five U.S. makers as a result of HP and Compaq merging."
SoI guess #4 spot was correct. And no, Gateway doesn't sell more machines than Apple.
---
Ed M.</strong><hr></blockquote>
Huh? Your original claim was that Apple is the 4th biggest computer maker <strong>in the world</strong>. That means <strong>global marketshare</strong>, right?
Like I said, Apple is not among the top 5, and the 6th spot is taken by IBM, Sony or Gateway. Its about a decade ago Apple was in the top5 list.
Global computer sales is about 130 million units, and Apple got about 2.6 % marketshare.
I promise, I'll eat the keyboard im typing on now, if Apple sell more than 220K powermac units this Q. Wont happend..
I totally agree, isn't this one of the fundamental principles of capitalism?
Let's say AI Inc. sells 1000 computers. We have our IPO, and the following year we sell 5000. Our stock skyrockets (like the 1999 net economy bubble). One year later recession hits, and the whole industry starts selling fewer and fewer computers. But AI's are so great that people keep buying them. But guess what, we only manage to increase our sales from 5000 to 5500: our stock collapse.
Capitalism, unfortunately, is based on growth. There is just one option: if you don't grow, you sink...
Sheesh...
ZoSo</strong><hr></blockquote>
ZoSo,
I'm afraid you're a little off in your reasoning.
First, you claim that our stock will have appreciated in the former instance and burst in the latter. This is only true if there is a stock market bubble in general, fueled by wrong expectations of future demand of our product. If the market foresaw that we'd sell 5000 in year one and 5500 in year two, the price of our stock would reflect that immediately. You shouldn't take what happened in the last 4 years on the stock market as an indication of what capitalism is all about, but more as a side effect of too much money and too little brains - and that's not a function of capitalism, but of human IQ.
Second, you actually make the argument that our share price would fall if our market share increased. That is, you acknowledge that our share price would fall at 5500 units, but 5500 units would be a higher market share than 5000 if the market itself had reduced in size.
In short (and this is meant as informed advice rather than an insult) the stock market and capitalism are not married to the notion of market share, revenue or customers - they are married to net profits over the long term - and you should ignore the advice of people who say otherwise if you don't want to take unnecessary risk with your investments in the future.
The roads.... I believe that the "roads" in this particular case is the *data*. As of now, most of the data that is exchanged between systems (and platforms) is transparent in that it can be taken from platform to platform\t
]]
</strong>
I'm sorry Ed, but DirectX ain't just data. Nor is "Embrace and Extend" - Micro$ofts raison detre - a mere matter of data. Both are about forcing proprietary standards, and controlling specific markets ...
Just like railroads were Ed ...
<strong>
[[
EXACTLY! And some people care about HOW you get there rather than just getting there. And you are ignoring a lot of other variables. You distill it down too much. What about efficiency, gas mileage, maintenance etc.? shouldn't these also be considered before choosing the appropriate car for the trip.? Similarly, what about the payload and or passenger compliment. A Vett can get from New York to Boston just fine, but what if you needed to take 5 people? So, I don't buy into your interpretation, there are too many other considerations that you conveniently left out. There are other areas that would affect he choice, that's why there is a host of different vehicles offered to suit the many different needs. If it were as simple as traveling from point A to point B then maybe the argument would have *some* merit. BTW, have you noticed that there are TONS of different toothpaste and toothbrushes on the market? Why is that?
]]
</strong>
... because Micro$oft, sorry, Colgate can't enforce proprietary access to teeth like Microsoft can with DirectX and games in Windoze, or with embrace and extend standard breaking policies with IE, Java, etc.
the analogy holds
<strong>
[[
You know... I always chuckle when I hear this... It's funny because companies essentially have "made a deal with the devil" in that Micro$oft has them locked in because of all the interdependencies. They are getting cooked slowly and likely spending more over time supporting a platform that has them locked in.
]]
</strong>
I totally agree - but the problem is, if MS controls access to the market you need to reach (thru standards and applications), just like a 19th century railroad baron, and as a company, in order to make a buck, you need to reach that market, you're stuck paying their taxes, regardless.
If the rail guage you want to use, doesn't take you to where you want to go, you can't use it, no matter how much you like it.
Now you can pay the expense of building a whole new railway yourself to the guage you like, but how many companies are going to go to the trouble of writing a whole new standard of Java, DirectX, Office (for which now, Microsoft is playing nice) etc etc just to use the app they need to?
It's probably better just to pay the MS tax.
<strong>
[[
What of the long term costs over a given period of time and taking into consideration all the hardware and software and OS updates that were required within the same platform?\t
]]
</strong>
see above ... if you have no alternative, because access to the market is controlled, you're stuck.
\t
<strong>
[[
[[[it's cost and hassle factor - Micro$soft knows this ]]]
I'm sure they do.. once the contract is signed, they have you locked. And if it really is cost and hassle, what's more costly to support than Windows and what's more of a hassle than Windows?\t
]]
</strong>
see above
<strong>
\t[[
[[[B - become so incomparibly cheaper and better than the competition, that it's worth the hassle of unloading.\t]]]
This is probably a myth since the cost of switching would really be an insignificant cost given the cost of support, upgrades, maintenance et. al that M$ has you locked into.
]]
</strong>
see above
<strong>
[[
The frog-in-the-pot-of-water analogy comes to mind.. And I believe Gartner did a study on just this very point. their suggestion was to ditch M$ and IIS and migrate to LINUX on Apache. (I forget the details), but the point was that it was a BETTER solution and it was cheaper, even over time. but the safe confines that M$ supplies... ooooooo.
]]
</strong>
... and hire extra IT department staff who can now handle both MS and Linux, I'm sure in many cases, this still works out, but it's a risk, and I'm also sure, in many cases, it doesn't.
<strong>
[[
[[[... if you no longer become the standard way people choose to get from A to B, the competing system will take over, and you'll never get back in.\t]]]
Wrong. This argument only holds if the you presume the DATA moving between platforms is proprietary and not transportable. It is.
]]
</strong><hr></blockquote>
I disagree for the same reasons stated above, we are stuck in a railroad analogy since data is not so easily transportable - see above - especially in a standards war, and especially if a certain company - microsoft - goes out of it's way - repeatedly, as they have - to make it so.
Comments
<strong>
Another take on the car-analogy:
...
Terje</strong><hr></blockquote>
Car analogies sucks... End of discussion.
Are you from Norway btw? I suspect you have been my professsor in a few subjects at NTNU.
<strong>
Car analogies sucks... End of discussion.
</strong>
<hr></blockquote>
If you say so. But they are fun...
[quote]
<strong>
Are you from Norway btw? I suspect you have been my professsor in a few subjects at NTNU.
Guilty as charged :-)
Terje
<strong>
Car analogies sucks... End of discussion.
Are you from Norway btw? I suspect you have been my professsor in a few subjects at NTNU.
Car analogies are fine if you understand cars
hP = Processor speed
rpm = FSB speed
Torque = Bandwidth = (hP x rpm)
it's exactly the same.
you can have a 1250 hP ( 1.25 GHz processor) with 13 ft-lb/s of torque (1.3 GB bandwidth) and you won't be able to move the car or data around the machine.
<strong>Another take on the car-analogy:</strong><hr></blockquote>
The white car represents Crelm toothpaste with the miracle ingredient, Fraudulin. The non-white car represents another toothpaste...
Screed
<strong>[[[Apple needs marketshare to keep developers on board. Im sorry, but you can tell me car analogies all day long, and its still not a perfect matching analogy. In the long run, selling 170K PowerMac units/Q is going to hurt Apple.]]]
Yep. That's so short term and now all the developers have going for them is relying on upgrades, milking people with nothing really new. More on that later...</strong><hr></blockquote>
OK, don't take this the wrong way Ed M. - I like you're train of thoughts, but after 6 page downs I have a suggestion for you.
I've noticed in a couple of your posts your method of quoting, and have trouble reading them. Use the reply button directly above the post and it will quote the original for you, and you can chop it up (like above). I like reading your responses, but due to formating and length I have difficulty finishing them. If that's 'your style', cool and disregard.
<strong>
Well, Apple *is* the 4th largest computer manufacturer in the world. (shipping volume)
Market share rankings of some familiar PC companies:
Dell: 14.1
Compaq: 11.1 (going to be 0% soon)
HP 7.1
Gateway: 6
--
Ed M.</strong><hr></blockquote>
Nopes.. Appple is like the 7th biggest computer maker. Its like.. about a decade ago Apple was in the top5 list.
<a href="http://news.com.com/2100-1001-944980.html"; target="_blank">http://news.com.com/2100-1001-944980.html</a>;
( Gateway sell more computers than Apple, so Gateway is probably the 6th biggest computer maker, and Apple the 7th.. or maybe IBM is selling more computers than Apple?? )
Anyway, this thread went pretty off topic..
[ 08-19-2002: Message edited by: blabla ]</p>
<strong>Capitalism, unfortunately, is based on growth. There is just one option: if you don't grow, you sink...</strong><hr></blockquote>
Your posts are good, don't discredit yourself with general statements. You know it's more complex than that, at least I hope.
<strong>I hate to disagree with you moki, but whilst the programme will request memory in small chunks most of the time, the processor will always request a cache line, or 32 Bytes at a time. The main reason why most programmes, which don't stream memory in large blocks, don't see much gain from DDR is that there is almost no change in the latency. (Okay, it amounts to the same thing in the end anyway, but I don't like to see people discussing just bandwidth as if it were everything, and ignoring latency. Just look at RDRAM)
DDR is great for streaming data, and of little benefit for almost everything else.</strong><hr></blockquote>
Your'e quite right -- I had forgotten about the way the cache makes these matters more complicated.
Regardless, the situation remains that DDR in the real world doesn't offer double the performance over SDRAM, despite what the name implies.
<a href="http://www6.tomshardware.com/mainboard/00q4/001030/athlon-15.html"; target="_blank">DDR and PIII</a>
<strong>Holy crap PSCATES, great post! You really really hit the nail on the head...
As far as MWSF, there is no way in H that there will be a brand new Powermac...not this soon after a big revision...
And as far as the post about this being the groundwork for the next proc. (ie "yikesv.2), I would say "I hope not". DDR is where the industry is today, they didn't prepare for the future, they just caught up to the present. DDRII is right around the corner as is serialATA and Hypertransport. Why does apple still include UltraATA66 for gods sake, are they really saving that much by not making both controllers ATA100, or even RAID built in considering the market for this thing. Anyhoo, The next chip will hopefully be on an all new platform w/ some future (soon to be current) tech.
Oh yeah, I'm also very happy with the current releases!</strong><hr></blockquote>
They include the ATA/66 controller because they've got bunches of them to get rid of. This saves them money and allows them to have 2 IDE buses, making people think they're getting extra value. Apple has done this before, although with a very crappy result (Performa 6300 anybody?).
For big Video manipulaters/creators it probably is warranted, but not for most users.
So why not have some 66 and 100 controllers.When faster HD's come out the old ATA66'ers will be going cheap.
Remember that link you just posted? The one found here:
<a href="http://news.com.com/2100-1001-944980.html"; target="_blank">http://news.com.com/2100-1001-944980.html</a>;
Well, if you bothered to read it you would see that Apple is into the top 5 manufacturers...
From the article:
"Apple jumped into the top five U.S. makers as a result of HP and Compaq merging."
SoI guess #4 spot was correct. And no, Gateway doesn't sell more machines than Apple.
---
Ed M.
Also, note...it was in the news recently about Apple outselling Gateway for the first time in a while. So Apple are 5 or 6th now.
Gateway are having problems competing...but Apple is getting its act together.
It will be interesting if between the Switcher campaign, opening the retail stores, Jag (removing many of the barriers to owner a Mac from a PC viewpoint...) and the problems PC co's are having in competing with Dell...if this gives Apple some joy-ward momentum.
ie I'd like to see where Apple is in terms of sales and market share a year from now.
And...will they count iPod PC sales...and any iphone (couldn't resist...if they make one...) sales as part of market penetration?
Steve Jobs has set his target at 10%.
That would give Apple alot more leverage with developers.
I half take Ed's point about Dell having '14%'. Viewed in those terms...if Apple can get to 10%...then things look far more healthy and interesting from Apple's point of view.
Gives them a stronger base to go after M$ from...
Lemon Bon Bon
<strong>
Remember that link you just posted? The one found here:
<a href="http://news.com.com/2100-1001-944980.html"; target="_blank">http://news.com.com/2100-1001-944980.html</a>;
Well, if you bothered to read it you would see that Apple is into the top 5 manufacturers...
From the article:
"Apple jumped into the top five U.S. makers as a result of HP and Compaq merging."
SoI guess #4 spot was correct. And no, Gateway doesn't sell more machines than Apple.
---
Ed M.</strong><hr></blockquote>
Huh? Your original claim was that Apple is the 4th biggest computer maker <strong>in the world</strong>. That means <strong>global marketshare</strong>, right?
Like I said, Apple is not among the top 5, and the 6th spot is taken by IBM, Sony or Gateway. Its about a decade ago Apple was in the top5 list.
Global computer sales is about 130 million units, and Apple got about 2.6 % marketshare.
I promise, I'll eat the keyboard im typing on now, if Apple sell more than 220K powermac units this Q. Wont happend..
Well, you could have a point about the towers.
And the overpriced obseletism Macs.
They'd sell more Macs if they dropped the price on them eg lack of tower below £1,350
:eek:
Sorry, but when you can get a great x86 tower for between £750 and £1,200? I'd call Apple overpriced.
Still, that's in another thread...
Good things to come at San Fran?
Er.
Bumped iMac.
Or.
Bumped Powerbook.
Or.
iPhone.
Or.
'power'Mac bumped to 1.5 on hackserve.
G5 would be nice though...
Lemon Bon Bon
The 6300 wasn't as bad as the earlier 6200s, as it had a 603e and L2 cache and could give ?100 PowerMacs a run for their money despite the moth-board.
Still had the right 32, left 32 mularkey though.
<strong>
I totally agree, isn't this one of the fundamental principles of capitalism?
Let's say AI Inc. sells 1000 computers. We have our IPO, and the following year we sell 5000. Our stock skyrockets (like the 1999 net economy bubble). One year later recession hits, and the whole industry starts selling fewer and fewer computers. But AI's are so great that people keep buying them. But guess what, we only manage to increase our sales from 5000 to 5500: our stock collapse.
Capitalism, unfortunately, is based on growth. There is just one option: if you don't grow, you sink...
Sheesh...
ZoSo</strong><hr></blockquote>
ZoSo,
I'm afraid you're a little off in your reasoning.
First, you claim that our stock will have appreciated in the former instance and burst in the latter. This is only true if there is a stock market bubble in general, fueled by wrong expectations of future demand of our product. If the market foresaw that we'd sell 5000 in year one and 5500 in year two, the price of our stock would reflect that immediately. You shouldn't take what happened in the last 4 years on the stock market as an indication of what capitalism is all about, but more as a side effect of too much money and too little brains - and that's not a function of capitalism, but of human IQ.
Second, you actually make the argument that our share price would fall if our market share increased. That is, you acknowledge that our share price would fall at 5500 units, but 5500 units would be a higher market share than 5000 if the market itself had reduced in size.
In short (and this is meant as informed advice rather than an insult) the stock market and capitalism are not married to the notion of market share, revenue or customers - they are married to net profits over the long term - and you should ignore the advice of people who say otherwise if you don't want to take unnecessary risk with your investments in the future.
<strong>
[[
The roads.... I believe that the "roads" in this particular case is the *data*. As of now, most of the data that is exchanged between systems (and platforms) is transparent in that it can be taken from platform to platform\t
]]
</strong>
I'm sorry Ed, but DirectX ain't just data. Nor is "Embrace and Extend" - Micro$ofts raison detre - a mere matter of data. Both are about forcing proprietary standards, and controlling specific markets ...
Just like railroads were Ed ...
<strong>
[[
EXACTLY! And some people care about HOW you get there rather than just getting there. And you are ignoring a lot of other variables. You distill it down too much. What about efficiency, gas mileage, maintenance etc.? shouldn't these also be considered before choosing the appropriate car for the trip.? Similarly, what about the payload and or passenger compliment. A Vett can get from New York to Boston just fine, but what if you needed to take 5 people? So, I don't buy into your interpretation, there are too many other considerations that you conveniently left out. There are other areas that would affect he choice, that's why there is a host of different vehicles offered to suit the many different needs. If it were as simple as traveling from point A to point B then maybe the argument would have *some* merit. BTW, have you noticed that there are TONS of different toothpaste and toothbrushes on the market? Why is that?
]]
</strong>
... because Micro$oft, sorry, Colgate can't enforce proprietary access to teeth like Microsoft can with DirectX and games in Windoze, or with embrace and extend standard breaking policies with IE, Java, etc.
the analogy holds
<strong>
[[
You know... I always chuckle when I hear this... It's funny because companies essentially have "made a deal with the devil" in that Micro$oft has them locked in because of all the interdependencies. They are getting cooked slowly and likely spending more over time supporting a platform that has them locked in.
]]
</strong>
I totally agree - but the problem is, if MS controls access to the market you need to reach (thru standards and applications), just like a 19th century railroad baron, and as a company, in order to make a buck, you need to reach that market, you're stuck paying their taxes, regardless.
If the rail guage you want to use, doesn't take you to where you want to go, you can't use it, no matter how much you like it.
Now you can pay the expense of building a whole new railway yourself to the guage you like, but how many companies are going to go to the trouble of writing a whole new standard of Java, DirectX, Office (for which now, Microsoft is playing nice) etc etc just to use the app they need to?
It's probably better just to pay the MS tax.
<strong>
[[
What of the long term costs over a given period of time and taking into consideration all the hardware and software and OS updates that were required within the same platform?\t
]]
</strong>
see above ... if you have no alternative, because access to the market is controlled, you're stuck.
\t
<strong>
[[
[[[it's cost and hassle factor - Micro$soft knows this ]]]
I'm sure they do.. once the contract is signed, they have you locked. And if it really is cost and hassle, what's more costly to support than Windows and what's more of a hassle than Windows?\t
]]
</strong>
see above
<strong>
\t[[
[[[B - become so incomparibly cheaper and better than the competition, that it's worth the hassle of unloading.\t]]]
This is probably a myth since the cost of switching would really be an insignificant cost given the cost of support, upgrades, maintenance et. al that M$ has you locked into.
]]
</strong>
see above
<strong>
[[
The frog-in-the-pot-of-water analogy comes to mind.. And I believe Gartner did a study on just this very point. their suggestion was to ditch M$ and IIS and migrate to LINUX on Apache. (I forget the details), but the point was that it was a BETTER solution and it was cheaper, even over time. but the safe confines that M$ supplies... ooooooo.
]]
</strong>
... and hire extra IT department staff who can now handle both MS and Linux, I'm sure in many cases, this still works out, but it's a risk, and I'm also sure, in many cases, it doesn't.
<strong>
[[
[[[... if you no longer become the standard way people choose to get from A to B, the competing system will take over, and you'll never get back in.\t]]]
Wrong. This argument only holds if the you presume the DATA moving between platforms is proprietary and not transportable. It is.
]]
</strong><hr></blockquote>
I disagree for the same reasons stated above, we are stuck in a railroad analogy since data is not so easily transportable - see above - especially in a standards war, and especially if a certain company - microsoft - goes out of it's way - repeatedly, as they have - to make it so.
[edit - bloody quotes]
[ 08-20-2002: Message edited by: OverToasty ]</p>