As others have pointed out, but nobody seems to get; Sun's primary market isn't 1, 2, or even necessarily 4 processor markets. Their big market is in the E10000, Sun Fire 12000, and Sun Fire 15000 machines--the smallest of which is limited to a maximum of 52 processors. While their midrange pushes up to 12 processors, and even their entry level servers can push as many as 8 processors. Not only that, but Sun machines are known for scaling incredibly well. Quite simply, Apple does not compete in these markets, IBM does, and for this level of scaling, IBM's Power4 is the high end.
If anything, Apple's highest end box is in a similar position as Sun's highest end workstations--supporting role for massive servers acting as computational farms (render and HPC--IBM's target) and high capacity transaction processing (Sun's target). Yes you can use smaller machines for some of these tasks, but when you NEED that capacity, Apple simply does not exist in that market.
To address your comparison with the Blade 2000 workstations--what if I happen to need a 3DLabs Wildcat or Oxygen graphics board for my CAD box? Can Apple support that?
Yes, one day desktop machines will catch up, but today isn't that day yet.
Lessig might be a lawyer, but not everybody necessarily wants to see him hurt. I donno about the other guy though... I'd pay good money to kick him in the nuts--not to mention RIAA board members or the board members of the Big 5.
Even down at the $10 range, they're competing against mediums that'll beat them out 90 to 99% of the time--at least for me. Hell, my first PS2 game as bought because it sold for $9.99 (Kengo). I spend my time looking through games that're priced at $15.99 (GTA3) or $19.99 (Agent Under Fire). I find DVD's in the 12.99 to $15.99 range that were great when I saw them in VHS, but since they're in DVD now, I pick them up.
Not to mention Anime (around $20) and Britcoms (whole series for around $100), which I see as a comparatively better value than music CDs drawing away my entertainment budget.
Even in this area, where fansubbed Anime can be had for free and WAY before US release, I still go out and spend money on the anime DVD's. Even when I feel the fansubbers did a better job than the releasing company--ironically this is true (call it attention to detail and running notes explaining cultural jokes) more often than not.
I just hope to see Hajime no Ippo out in a set of DVD's stateside soon, missing episodes 23 to 34 is just killin' me.
Then why not test if it REALLY IS worth it to them. They get a monopoly starting at zero point, at the 1 year mark, they pay $1, then it doubles every year.
No point making keeping a copyright too easy--although I would provide provisions that later derivative works must also be published with original credit going to authors who came before and how it all started.
Cost of entry's been made prohibative to anybody except the very rich, and as corporations, they can raise enough money to push that cost higher and higher (look at what happened to webcasting).
All that's left for them is to offer obscene amounts of money to anybody they can't price out of the market.
You're right though, it's not so much communistic as an oligarchic system. In either case, customers lose while consumers end up even.
If there's a single 'monster' or 'alien' in the game, it automatically counts as poor single player TO ME.
Maybe you get FPS to shoot at what amounts to a walking target, but I want people who respond as people would. When I'm filling a room with bullets, I full expect a guy down the hall to call for backup and hunker down waiting for backup while holding me off until backup gets there or I kill him. It gets rather dull when it's just a bunch of monsters charging down the hallway at you.
In that case, the government can do something like subsidize the competition, sure it's unfair, but such grants can come and go as needed. Which means the competition will have to float on it's own eventually, all the government would do is enable it to compete against jumbocorp.
When you have a lot of money to throw around, you can create competition at will.
Of course, this assumes no lobbiest or bought out politicians.
Are you sure? All current SIMD implimentations should be performing 32-bit ops on 128-bit registers nowadays, with the option to do double precision in 2 cycles instead of just one.
Vectors would need to be described by a start and an end requiring a full 128-bits to do work on them. Kind of foolish to provide for SIMD/Vector work without providing registers that could handle it on a load-store architecture, don't you think?
How is calling a processor Post-RISC claiming that it is not RISC? IIRC, his statement was that RISC has grown beyond it's original design philosophy and now includes added complexity that increase performance. What's so bad about evolving to increased complexity anyway?
Moreover, how is this claim anti-Apple? Are they the ONLY company in the world using processors built on Post-RISC design philosophies? Doesn't IBM also use PPC750's in workstations supporting their servers?
Well, we expect them to have the latest and greatest out of 3Dlabs, like the $3000 wildcat's that render previews for the CAD workstation user faster then his poor CPU could.
Did she get XP Pro or XP Home. Pro really is just W2K with butt kissing extras (I should know I made the transition from W2K to XP Pro). XP Home though... now that's a piece of work. MS should just retire that the way of WMe and make Pro the standard for ANY single user single access box.
Just using an example based on my experience working with devices that can have millions of dollars R&D and yet still be sold for pennies
Look at the ICs on any circuit board. Chances are good some of the most expensive parts are sold for about $5 to $10 a piece, the cheapest would run for about a penny per 2 to 8 parts on one die. To produce those parts requires a fabrication plant, which could range from millions for something fairly low end all the way to billions for state of the art--I'm including equipment, by the way. You need to pay for the people to work in these plants. You need to pay for engineers to design, test/verify, and market your part--and its infrastructure. You need to pay for mask layout and pcb layout techs. You also pay environmental fees.
Luckily, your plant can be fabricating lots and lots of different parts all at once. Each can range in terms of R&D from several hundreds of thousands of dollars well into the tens of millions. Now, after all that money's spent how much did those ICs finally go for, again?
Limited run and special case stuff can be expensive, several dollars a piece, but the price quickly drops. Mass production parts are sold at good prices to capture marketshare in hopes of long lifetimes.
That last sentence should have particular meaning in this argument. As your production capability rises, YOUR cost per part DROPS. As production pushes to infinity, YOUR cost per part reaches down towards ZERO.
Now let's apply this knowledge to drugs. Say it's a drug that covers 14 day treatment. How much did they spend on R&D of this one drug? Should we guesstimate $2 to $5 million maybe $10 million--supposing the big capital expenditures are already paid? I wish I could get more exact numbers, if someone in biotech could enlighten me, that would be helpful. So right now we're looking at 14 pills per patient. If we sold it at $1 per pill, we'd need about 750k patients to recoup our expenditures with some profit, that number gets smaller as you push the cost per pill up. What happens if you have MILLIONS of patients PER YEAR? Are you then justified in selling it at $1 or $2 per pill, knowing that your R&D is ALREADY PAID FOR in that first year? Keep in mind you've got what, 7 years to reap your rewards?
I'd have no problems with some of the high cost if the number of patients is small. But don't try to sell me on drugs that go for $10 a pill when you have millions needing them.
To address your last comment. Well, why do researchers at universities go into finding cures for disease? Maybe they have it themselves, maybe they have friends and family who have it. Sounds to me like research would get done if there's profit or not (as in not for profit work vs gratis) involved so long as they can find someone (including gov't) with money to fund them.
In the x86 market, you're lucky to get 15% margin on a machine (10% if you're trying to undercut somebody)--keep in mind I'm talking about LOCAL computer shops selling hardware.
Apple makes A LOT more then this. I wouldn't be surprised if they sold much of their hardware at 50% markup--maybe as high as 75% depending on WHAT it is. Why and how they can do this is easily answered. People may grumble, but they still buy.
Yes software can have 10x the markup (by that I mean 1000%), however, there's a lot more of a fight in that field vs hardware.
It'd be nice to get OS X for x86, and the margins on software are huge, but Apple really does make its money on its systems and Microsoft would find a way to eviscerate them in the software market.
I switched to NiMH only myself--I haven't had to buy new batteries for the past 2 years. Just pick up enough so that you can have a set in the recharger and a set in use at any given time.
I think for the most part the limitation is purely physical-financial.
Since you can impliment any number of things to get at controllers, your limit is how many controllers can you fit onto a die (you get to define this, but larger can mean costlier). Also, without thinking about impedence matching lines, you're also limited in board space for ports.
Some folks don't see them, others are driven insane by them (could just be the work they do results in them commonly seeing the shadows).
The Mitsubishi tubes have gotten very good and that COULD be the trin. I'd choose to replace my hitachi when it dies. As it is, I hope to replace with another shadowmask.
Trins still have a washed out look, to me. Even the Mitsu that I like has it, which results in me lowering the brightness--originally a plus for trins.
Then of course there's the holy grail of CAD displays, the Eizo/Nanao F980--look into it if you want near perfection, but it'll compete against LCD's in pricing.
Slashdot Mirror
As others have pointed out, but nobody seems to get;
Sun's primary market isn't 1, 2, or even necessarily 4 processor markets. Their big market is in the E10000, Sun Fire 12000, and Sun Fire 15000 machines--the smallest of which is limited to a maximum of 52 processors. While their midrange pushes up to 12 processors, and even their entry level servers can push as many as 8 processors. Not only that, but Sun machines are known for scaling incredibly well. Quite simply, Apple does not compete in these markets, IBM does, and for this level of scaling, IBM's Power4 is the high end.
If anything, Apple's highest end box is in a similar position as Sun's highest end workstations--supporting role for massive servers acting as computational farms (render and HPC--IBM's target) and high capacity transaction processing (Sun's target). Yes you can use smaller machines for some of these tasks, but when you NEED that capacity, Apple simply does not exist in that market.
To address your comparison with the Blade 2000 workstations--what if I happen to need a 3DLabs Wildcat or Oxygen graphics board for my CAD box? Can Apple support that?
Yes, one day desktop machines will catch up, but today isn't that day yet.
Lessig might be a lawyer, but not everybody necessarily wants to see him hurt. I donno about the other guy though... I'd pay good money to kick him in the nuts--not to mention RIAA board members or the board members of the Big 5.
Even down at the $10 range, they're competing against mediums that'll beat them out 90 to 99% of the time--at least for me. Hell, my first PS2 game as bought because it sold for $9.99 (Kengo). I spend my time looking through games that're priced at $15.99 (GTA3) or $19.99 (Agent Under Fire). I find DVD's in the 12.99 to $15.99 range that were great when I saw them in VHS, but since they're in DVD now, I pick them up.
Not to mention Anime (around $20) and Britcoms (whole series for around $100), which I see as a comparatively better value than music CDs drawing away my entertainment budget.
Even in this area, where fansubbed Anime can be had for free and WAY before US release, I still go out and spend money on the anime DVD's. Even when I feel the fansubbers did a better job than the releasing company--ironically this is true (call it attention to detail and running notes explaining cultural jokes) more often than not.
I just hope to see Hajime no Ippo out in a set of DVD's stateside soon, missing episodes 23 to 34 is just killin' me.
Too bad it's no longer time-limited.
As Congress has established, if you pay them enough, you can own the rights to be the sole distributor of something forever, simply in smaller chunks.
Patents are only 7 years IIRC, this ensures that new creations are made.
Then why not test if it REALLY IS worth it to them. They get a monopoly starting at zero point, at the 1 year mark, they pay $1, then it doubles every year.
No point making keeping a copyright too easy--although I would provide provisions that later derivative works must also be published with original credit going to authors who came before and how it all started.
3x$30k = 90K
During the year you paid out
10x2500 = 25k
That leaves you with 65k.
He made 30k
He paid out 2500
That leaves him with 27.5k.
So... Who's living paycheck to paycheck and how badly are they living paycheck to paycheck?
Why take with force what you can buy with money?
Cost of entry's been made prohibative to anybody except the very rich, and as corporations, they can raise enough money to push that cost higher and higher (look at what happened to webcasting).
All that's left for them is to offer obscene amounts of money to anybody they can't price out of the market.
You're right though, it's not so much communistic as an oligarchic system. In either case, customers lose while consumers end up even.
If there's a single 'monster' or 'alien' in the game, it automatically counts as poor single player TO ME.
Maybe you get FPS to shoot at what amounts to a walking target, but I want people who respond as people would. When I'm filling a room with bullets, I full expect a guy down the hall to call for backup and hunker down waiting for backup while holding me off until backup gets there or I kill him. It gets rather dull when it's just a bunch of monsters charging down the hallway at you.
In that case, the government can do something like subsidize the competition, sure it's unfair, but such grants can come and go as needed. Which means the competition will have to float on it's own eventually, all the government would do is enable it to compete against jumbocorp.
When you have a lot of money to throw around, you can create competition at will.
Of course, this assumes no lobbiest or bought out politicians.
Maybe because law enforcement can shoot back and everybody knows it?
Are you sure? All current SIMD implimentations should be performing 32-bit ops on 128-bit registers nowadays, with the option to do double precision in 2 cycles instead of just one.
Vectors would need to be described by a start and an end requiring a full 128-bits to do work on them. Kind of foolish to provide for SIMD/Vector work without providing registers that could handle it on a load-store architecture, don't you think?
He's making the claim based on desktop machines, not laptops. Keep it in mind when you're thinking about your power budget.
How is calling a processor Post-RISC claiming that it is not RISC? IIRC, his statement was that RISC has grown beyond it's original design philosophy and now includes added complexity that increase performance. What's so bad about evolving to increased complexity anyway?
Moreover, how is this claim anti-Apple? Are they the ONLY company in the world using processors built on Post-RISC design philosophies? Doesn't IBM also use PPC750's in workstations supporting their servers?
Well, we expect them to have the latest and greatest out of 3Dlabs, like the $3000 wildcat's that render previews for the CAD workstation user faster then his poor CPU could.
Did she get XP Pro or XP Home. Pro really is just W2K with butt kissing extras (I should know I made the transition from W2K to XP Pro). XP Home though... now that's a piece of work. MS should just retire that the way of WMe and make Pro the standard for ANY single user single access box.
Just using an example based on my experience working with devices that can have millions of dollars R&D and yet still be sold for pennies
Look at the ICs on any circuit board. Chances are good some of the most expensive parts are sold for about $5 to $10 a piece, the cheapest would run for about a penny per 2 to 8 parts on one die. To produce those parts requires a fabrication plant, which could range from millions for something fairly low end all the way to billions for state of the art--I'm including equipment, by the way. You need to pay for the people to work in these plants. You need to pay for engineers to design, test/verify, and market your part--and its infrastructure. You need to pay for mask layout and pcb layout techs. You also pay environmental fees.
Luckily, your plant can be fabricating lots and lots of different parts all at once. Each can range in terms of R&D from several hundreds of thousands of dollars well into the tens of millions. Now, after all that money's spent how much did those ICs finally go for, again?
Limited run and special case stuff can be expensive, several dollars a piece, but the price quickly drops. Mass production parts are sold at good prices to capture marketshare in hopes of long lifetimes.
That last sentence should have particular meaning in this argument. As your production capability rises, YOUR cost per part DROPS. As production pushes to infinity, YOUR cost per part reaches down towards ZERO.
Now let's apply this knowledge to drugs. Say it's a drug that covers 14 day treatment. How much did they spend on R&D of this one drug? Should we guesstimate $2 to $5 million maybe $10 million--supposing the big capital expenditures are already paid? I wish I could get more exact numbers, if someone in biotech could enlighten me, that would be helpful. So right now we're looking at 14 pills per patient. If we sold it at $1 per pill, we'd need about 750k patients to recoup our expenditures with some profit, that number gets smaller as you push the cost per pill up. What happens if you have MILLIONS of patients PER YEAR? Are you then justified in selling it at $1 or $2 per pill, knowing that your R&D is ALREADY PAID FOR in that first year? Keep in mind you've got what, 7 years to reap your rewards?
I'd have no problems with some of the high cost if the number of patients is small. But don't try to sell me on drugs that go for $10 a pill when you have millions needing them.
To address your last comment. Well, why do researchers at universities go into finding cures for disease? Maybe they have it themselves, maybe they have friends and family who have it. Sounds to me like research would get done if there's profit or not (as in not for profit work vs gratis) involved so long as they can find someone (including gov't) with money to fund them.
For a hardware company, they write some incredible compilers.
So how much does it cost for me to send a letter and a data cd by fedex ground? USPS costs me about 80 cents, BTW.
You'd be surprised.
In the x86 market, you're lucky to get 15% margin on a machine (10% if you're trying to undercut somebody)--keep in mind I'm talking about LOCAL computer shops selling hardware.
Apple makes A LOT more then this. I wouldn't be surprised if they sold much of their hardware at 50% markup--maybe as high as 75% depending on WHAT it is. Why and how they can do this is easily answered. People may grumble, but they still buy.
Yes software can have 10x the markup (by that I mean 1000%), however, there's a lot more of a fight in that field vs hardware.
It'd be nice to get OS X for x86, and the margins on software are huge, but Apple really does make its money on its systems and Microsoft would find a way to eviscerate them in the software market.
I switched to NiMH only myself--I haven't had to buy new batteries for the past 2 years. Just pick up enough so that you can have a set in the recharger and a set in use at any given time.
I think for the most part the limitation is purely physical-financial.
Since you can impliment any number of things to get at controllers, your limit is how many controllers can you fit onto a die (you get to define this, but larger can mean costlier). Also, without thinking about impedence matching lines, you're also limited in board space for ports.
Depends on you.
Some folks don't see them, others are driven insane by them (could just be the work they do results in them commonly seeing the shadows).
The Mitsubishi tubes have gotten very good and that COULD be the trin. I'd choose to replace my hitachi when it dies. As it is, I hope to replace with another shadowmask.
Trins still have a washed out look, to me. Even the Mitsu that I like has it, which results in me lowering the brightness--originally a plus for trins.
Then of course there's the holy grail of CAD displays, the Eizo/Nanao F980--look into it if you want near perfection, but it'll compete against LCD's in pricing.
Last Windows disk that said it'd work on PowerPC was NT4, IIRC. So you could try booting that. ^_^ You're on your own as far as support goes though.
I really gotta preview what I type a couple of times 'fore submitting.
I meant they were gems for jewelers. Obviously fakes, but they gave off a different color in UV light due to dopent (SP?).
Gotta love lab grown crystals for their physical properties though.