IPv6 implements some nice features that aren't aimed at a larger address space. IPv6 provides for priority and quality of service information in the packet, allowing for better priority based routing. It also doesn't permit for fragmenting packets, which makes life easier for both routing and stitching it back together at the destination. And distrobution of the addresses is done more fairly. It's not the US and western Europe (to a lesser extent) grab the address space they'd like and the rest of the world can scrounge for what's left.
NAT does blur the line between Network layer and transport layer somewhat. NAT uses TCP or UDP ports to do routing. Good design would dictate that independant modules of a system should stay indepedant, NAT doesn't do that. Not that it's really a big deal here, there's not much change of a new transport layer protocol grabbing hold anyomre.
the same market that won't buy a console because they think the graphics suck right now. Or that the ~500 lines of resolution on a TV is woefully insufficient to render the 'proper' graphical detail they desire in their games. Or of course, the rich yet clueless. (note: those two are not mutually exclusive...)
Personally my 9600XT is plenty good for my gaming needs, I'd like to be able to run everything at 1280x1024 (native res for my LCD), but I'm not complaing about 1024x768 or even lower, they look just fine to me. Which is why I'm deffinately not the target audiance for SLI.
Right now it's deffinately uneeded, but a year from now we may seem games wher 1280x1024 or 1600x1200 bring a 7800GT / GTX to 'marginally' playable frame rates (say about 30 FPS), you already saw Splinter Cell: Chaos theory was brought under 40 FPS at 1600x1200 AA and AF enabled. It's certainly not unreasonable to expect far more graphically demanding games over the next few years.
but it won't happen. The more 'photorealistic' the engine can make the game world, the more art and design is needed to take advantage of that. You might not spend those dev dollars on stretching every last polygon out of the engine, but you will spent them making those polygons look good. If you have 2GiB of video memory availible for textures and associated maps, you'd better make good use of it. No more repeating the same box image over and over in every level.
Honestly I think graphcis are an adivsary to game play; and probalby will continue to be after we reach a point where more graphics rendering power is of no benifit. (Which I doubt will be anytime soon) The simple fact is there is generally a limited budget to do everything, and right now limited CPU / GPU cycles. If you have to develope tons of super high quality and complex models and textures to take advantage of Really super fast rendering engine(TM) then the AI, the story, ect suffer. If you have to optimize out every last CPU cycle of waste in the Really pretty slow and boring rendering engine(TM) then the AI, story ect suffer.
I think the fact is, graphics sell games, gameplay keeps you playing. Only one of those is useful for a companies bottom line. (Unless you're making The Sims and have 457 expansion packs you need to sell)
They are going to have AGP versions of the X1 600 and X1 300. Just not the X1 800. Just as a side note, before you jump back to Nvidia, there won't be an AGP version of the 7800 series either. (At least one is not planned, some industrious company may build their own using Nvidia's bridge chip, just as some one may do that for ATI, it just won't be anytime soon).
Right now you AGP choices are probably going to end with 6800, or x850 being the fastest. Sorry but us AGP users are pretty much now considered a legacy market. PCIe is the power slot of the future...
only if the subsequent loops are dependant on data from the current loop.
something like for(int i = n-1; i>0; i--){ n = n * i }
obviously the new value of n depends on the value for n calculated by the last loop so that might not be a good candidate to try and parallelize. (actually factorial is something that can be written to take advantage of instruction level parallelism (ILP), I choose not too simply for the example).
however, if you're doing something that is not dependant on previous loops, various forms of loop unrolling can exploit ILP. take for example blending two images for each row x and each column y, x++, y++ imageTarget[x][y] = 1/2 * imageSrc1[x][y] + 1/2 * imageSrc2[x][y]
one pixel does not depend on the result of the previous, there's no reason you can't do 2, 4, 8, 16 ect pixels inside each loop. Some compilers can take advantage of this already in doing loop unrolling to utilize MMX or SSE (or similar SIMD instruction sets) instructions. It seems like Trips is an instruction set designed to aid the processor in finding and exploiting such ILP. The usefullness of such massively parallel designs in general purpose computing is debatable I would say. On the whole there tend to be a lot more instructions with dependancies than those without. (obviously everything has some dependancies, I mean in such a manner that prevents ILP / loop unrolling). Hardware has been moving towards more parallelism with super-scalar and multi-chip processing and more functional SIMD instruction sets, but software has gone only kicking and screaming into a more parallel world. Athlon and Pentium 3, Pentium M can look at up to something like 14 x86 instructions and decode up to 3 of them per clock cycle. More often than not they can't find 3 suitable instructions to decode. I have a hard time believeing something is going to find 32 (16 per core, 2 cores on the prototype) for general purpose software.
generally if it's the result of a 'good faith mistake,' ie a clerical error, or improperly labled addresses, evidance obtained in search warrant is admissible even when they end up searching the wrong house / building. I'd believe the same train of thought would be applicable to this, a good faith effort to tap the right phone and end up with the wrong one would probably still be admissable in court.
One the one hand, yes of course software developers need to be accountable for their work. This isn't bounded by an license or developer. If you release software, you have a responsibility to maintain and support it. I'm not talking about if some one peverts your work into something malicious or if some one uses an unkown exploit. I don't believe developers need be held responsible for damages relating to thos types of situations. But I do belive if you have software out there, you have a reponsibility to your users to patch security and stability (and privacy and others that might arise) issues in a timely fassion. Barring that, if you sit on your lorells and watch as people use a known bug to do harm with your software I don't have a problem holding a developer responsible for damage. Willfull disregard for the damage your program can have should not be tolerated.
But then again, nailing down when, what and for how long it's reasonable to expect proper support for software is pretty much impossible. I don't epxect MS to be supporting win95, I don't expect Linux to put much active support into the 1.0 kernel, I expect ATM software to maintained so long as an active ATM using it. If you ask 100 different people you'll probably get around 150 different answers, picking any one of them would not only be extremely difficult, but probably dangerous as well.
Silent PC Review did a review of the Phaton 500 back in May; and did a far better job of actually putting it through its paces.
This is a typical PSU review, that is to say worthless. The problem is to do a good PSU reivew you actually need quite a bit of hardware, most little online sites lack even the most basic testing tools (a good multimeter and a controllable load). They make no mention of how they measured the voltages (software, or voltmeter, and from where, pigtail, ATX connector, somewhere else), they put a system that probably doesn't draw 125W DC at load to test out a 500W PSU, they have no real PSU temperature or efficency information. Typical of a site who's reviewing expertiese consists soley of swaping out parts, running 3D Mark and reporting the difference.
Silent PC Review does half way decent reviews, and over the last year or so XBit Labs has starting doing very good PSU reviews. Beyond that there aren't too many places that consistantly hit the mark.
For a silent PSU (not sure why this is that big of a deal, I have a TruePower 330W and can't hear it over the HDD, but I guess some people will always pay for that last dB quieter), there's of course the Phantom 300, the SilverStone 'NF' series, a 300 and a 400W version, the Fortron Source Zen 300; recently reviewed on XBitLabs and Silent PC Review, with just rock solid voltages across the spectrum. And of course the SeaSonic S12 line while not fanless is known to be extremely quiet and highly efficient
Just to follow up on that, Is the AI going to be as moddable and customizable as the rest of the game content?
I know Mr. Caudill mentioned an 'AI SDK' for 'experienced programmers' over on the IGN Civ 4 preview to tailor the AI to their desires. But it was mentioned as a seperate entity from the XML unit files and the basic Phython scripts. Is this because the AI is more hard coded (less of it in easily accessible scripts) than say unit stats, or just an attempt to give a helping hand to less experianced modderings in a rather complex enviorment like the AI.
Basically I was hoping you could go into some more detail on what AI and other more complex modding might entail.
The plan is to pick up a second one when the price drops to around $100.
This is sorf of the flaw in a lot of people's SLI plans. (Because Crossfire can use 1 non-crossfire compatible card ATI's solution might not be quite as bad. A 6800 GT is still about $275-300 right now, the 6800 Ultra is $350 and up. You can't hardly buy a 5900/U/5950 anymore.
If you don't mind going to ebay or second hand cards for your second one you can probably save a decent amount of money on the second card, of course you already spent a bit of money up front for an SLI motherboard and a PSU to handle a second card. It might end up being about cost effective if you can scrounge up a second card on the cheap in a year or two, but the 'I'll just pick up a second card when they're cheap' really isn't going to materialize unless Nvidia and ATI change their attidude towards older generations, keeping old chips like the 9800 or 5700 in production and at value prices has been the exception, not the rule. Especially for Nvidia.
The 90nm process intel is using has a very similar thing, only not done near to the extent their talking about here.
Intel has two sets of transistors for 90nm, high voltage threshold and low voltage threshold. High VT are fairly power efficient as it is, about 40nA/um leakage, Ion about 31 times greater than Ioff (NMOS) Low VT (which were used extensivly in Prescott to get it to scale to the 4-5ghz range it was intended for), which are horribly inefficient, with a leakage of about 400nm/um, Ion around 3.5 times Ioff (NMOS)
Seems like this is largely a really really high VT transistor, with a few tweaks to the oxide thickness for good measure. In any case, it should help out the ultra low power devices to an extent, but won't effect any of intel's 65nm desktop/laptop chips. (save maybe a chipset, but I doubt we'd see a 65nm chipset).
If you had RFTA (wait this is/., nevermind), the voting comes from a shortlist of a few candidates, much like how Fox/American Idle choose a handful of decent candidates (as well as some pathetic morons for amusemnet) from the thousands that apply, so will the Malaysian space program will choose a few qualified candidates and the people will vote on them.
And of course in the great tradition of misleading, or at least sensationalized headlines, the article only says 'the Malaysian space boses will factor the votes into their final decision.' Ie it's a gimic to get people interested in space, but probably won't have any real effect on the process.
how many new materials, technologies ect have been largely ancillary to the ultimate goal (or even developed without a goal in mind), NASA, DARPA, and similar organizations don't exist to be financialy viable in and of themselves, they exist to reach beyond what we are capable of now, and make it possible.
A bit over $8 billion dollars a year for the next 12 years is pocket change for what benifits we can reap. Not in the next 10 years, but in the next 50, or 100 years.
Mars in ten years?
Oh, that was just a political tool to move the news cycle from whatever massive screwup the white house was involved in that particular week to grandious dreams of unfunded potential futures.
A very minor update to 2000 to convince people to shell out another $100 for a better looking interface, a couple of moderately usefull features little else?
Isn't that why most of the corperate and even many home users (like myself) of 2000 opted NOT to upgrade at all?
The article was sketchy, maybe smaller expense, smaller expectations make some sense. Less cost (to MS and the consumer I would think) per upgrade, less benifit, decide to upgrade every few years, but MS has part of the user base upgrading all the time, not just in the year or so after a big software release.
The report doesn't consider expenses related to design or marketing, or the fact that high-end chips can sell for more because fewer off the production line can actually run at top speed,
said the article.
Design, research, and marketing can run well over a billion dollars for a single core. And who knows if that includes things like the upstart cost for new production plants, with Intel building 4 65nm fabs, they could easily be sinking $10-12 billion dollars into the ability to just produce chips to being with.
Materials / labor ect to turn sand into a CPU isn't a trivial cost, but it's not representative of the cost to bring a CPU to market either.
Well x86 chips have pretty well developed methods of dealing the lack of registers. Register renaming eliminates, or at least minimizes most of the problems with a small register set. (Athlon64 has something like 72 integer registers and 122 90 bit FP registers (two of these are combined to make an XMM register for SSE vectors), almost all of which are availible in 32 bit mode).
The extra achitectual registers will help with moderate to long term storage (more than a few dozen clock cycles between uses) as the programmer will explicity specify the data remains in the register, where as with current shuffeling it's up to the CPU (and to some extent how the renamed registers are inteded to work) to determine if a write to cache is in order, or not. And really with the longer storage times, you often have the flexibility to write out to L1 and schedual the load so that there's no penalty for the load. (ie issue the move back to the register the 3 clock cycles prior to when you need it that an L1 load usually takes).
The new registers probably won't make all that much difference in the end. But the again, nothing from the move to 64 bit will be a major impact for a while (at leat on the desktop).
While Hitachi did by IBM's HDD wing, we need to be clear. The actual "DeathStar" drives were a very select line. IBM tried to put 5 platters into their high capacity 75GXP line, the norm is 4 for 3.5'' disks. These lead to excessive head crashed (I've heard up to around 30% of the drives met their death this way).
Even before IBM sold the HDD buisness they had gone back to a 4 platter design which effectivley elminated the 'death' part of the deathstar line.
If you like to boycott them based on passed wrongs, that's fine and your call. (Ther are brands I avoid to this day because of past buiness practices). But there are no quality / reliability issues with any of the current Hitachi hard drives.
hmm... Nvidia does have a history of releasing every driver build they can and letting the user toy with them to find what fits best.
But ATI, SoundBlaster, yes I've been using them for about 3 years now. Before that Nvidia and SoundBlaster, only real stability issues I've ever had were the SB Live and 686B southbridge fighting it out...
Most games do well enough to configure their own settings it's been some time since I even made major changes to video / image quality / performance settings.
I install games, they run, I'm happy. It's not as fast as a console, but as for easy, awfully close for me.
Well, sitting at the desktop, playing no games now steam is eating up 20MiB of ram. (not a massive ammount but 5-10 times what the rest of my little background apps take up). I could (and try to) close it when I exit the game, but (to the best of my knowledge) there is no automated way to close steam on exiting HL2/CS so it's an extra hassle to reclaim that memory.
You can't play online if the steam server that holds your account is down (no authentication), which has happened in the past, and per murphy's law will probably happen in the future.
Though I think it's much better now, in the past large updates have been horrendesly slow to come by as the steam servers weren't up to the bandwitdh tasks.
And of course the big one, stop using or switch to a new - and incompatible - authentication software. Like say right after the release of Half life 3. [/tin foil hat] But you are at the mercy of valve to maintain the steam servers or you loose the ability to play your legally bought games unless you know how to crack it.
Steam isn't a bad thing by itself, I kind of like a lot of what it does. Decent anti-cheat program, seamless updating (both of the anti-cheat and of the games), decent content delivery system (though the pricing sucks). The problem with steam is that it takes rights away from the consumer and gives them to the developer. Namely I have to ask permission to play my own game.
It essentially comes with *nothing* in the box, seemingly just so they can say they offer a lower pricepoint. The cost of accessories is unjustified. The memory cards hold, what, 64MB? And they cost $40?! I found a 512MB USB thumbdrive on Froogle for that price. Christ, how about just slapping on a USB port and letting us use any generic thumbdrive?
The bare console (particularly early in the life span) tend to be a break even, or loose money proposistion for the manufacturer (MS, Nintendo, Sony). They rely on the 'optional' extras like games, hard drives, memory cards, wireless or extra controllers, faceplates ect, to make their money in the first few years. If they open such accories up to open competition you'll likely be paying substantially more for the console itself as the manufacturer can't rely on the income from overpriced accessories.
x86-64, Itanium isn't coming to the destkop, even with the software x86 emulator in the latest Win2003 SP, it's no where near fast enough to make a seemless transition. SSE1/2/3 probably 4 (even if it's an even smaller addition than SSE3 just to say they're doing something new).
Slashdot Mirror
Hello God, this Albert Einstien
I'll see you patent court.
Bring a good lawyer.
IPv6 implements some nice features that aren't aimed at a larger address space.
IPv6 provides for priority and quality of service information in the packet, allowing for better priority based routing.
It also doesn't permit for fragmenting packets, which makes life easier for both routing and stitching it back together at the destination.
And distrobution of the addresses is done more fairly. It's not the US and western Europe (to a lesser extent) grab the address space they'd like and the rest of the world can scrounge for what's left.
NAT does blur the line between Network layer and transport layer somewhat. NAT uses TCP or UDP ports to do routing. Good design would dictate that independant modules of a system should stay indepedant, NAT doesn't do that. Not that it's really a big deal here, there's not much change of a new transport layer protocol grabbing hold anyomre.
the same market that won't buy a console because they think the graphics suck right now. Or that the ~500 lines of resolution on a TV is woefully insufficient to render the 'proper' graphical detail they desire in their games.
Or of course, the rich yet clueless. (note: those two are not mutually exclusive...)
Personally my 9600XT is plenty good for my gaming needs, I'd like to be able to run everything at 1280x1024 (native res for my LCD), but I'm not complaing about 1024x768 or even lower, they look just fine to me. Which is why I'm deffinately not the target audiance for SLI.
Right now it's deffinately uneeded, but a year from now we may seem games wher 1280x1024 or 1600x1200 bring a 7800GT / GTX to 'marginally' playable frame rates (say about 30 FPS), you already saw Splinter Cell: Chaos theory was brought under 40 FPS at 1600x1200 AA and AF enabled. It's certainly not unreasonable to expect far more graphically demanding games over the next few years.
the board doesn't have an SLI link connector, so no. 2 GPU max still.
but it won't happen.
The more 'photorealistic' the engine can make the game world, the more art and design is needed to take advantage of that. You might not spend those dev dollars on stretching every last polygon out of the engine, but you will spent them making those polygons look good. If you have 2GiB of video memory availible for textures and associated maps, you'd better make good use of it. No more repeating the same box image over and over in every level.
Honestly I think graphcis are an adivsary to game play; and probalby will continue to be after we reach a point where more graphics rendering power is of no benifit. (Which I doubt will be anytime soon) The simple fact is there is generally a limited budget to do everything, and right now limited CPU / GPU cycles. If you have to develope tons of super high quality and complex models and textures to take advantage of Really super fast rendering engine(TM) then the AI, the story, ect suffer. If you have to optimize out every last CPU cycle of waste in the Really pretty slow and boring rendering engine(TM) then the AI, story ect suffer.
I think the fact is, graphics sell games, gameplay keeps you playing. Only one of those is useful for a companies bottom line. (Unless you're making The Sims and have 457 expansion packs you need to sell)
They are going to have AGP versions of the X1 600 and X1 300. Just not the X1 800.
Just as a side note, before you jump back to Nvidia, there won't be an AGP version of the 7800 series either. (At least one is not planned, some industrious company may build their own using Nvidia's bridge chip, just as some one may do that for ATI, it just won't be anytime soon).
Right now you AGP choices are probably going to end with 6800, or x850 being the fastest. Sorry but us AGP users are pretty much now considered a legacy market. PCIe is the power slot of the future...
only if the subsequent loops are dependant on data from the current loop.
something like
for(int i = n-1; i>0; i--){ n = n * i }
obviously the new value of n depends on the value for n calculated by the last loop so that might not be a good candidate to try and parallelize. (actually factorial is something that can be written to take advantage of instruction level parallelism (ILP), I choose not too simply for the example).
however, if you're doing something that is not dependant on previous loops, various forms of loop unrolling can exploit ILP.
take for example blending two images
for each row x and each column y, x++, y++
imageTarget[x][y] = 1/2 * imageSrc1[x][y] + 1/2 * imageSrc2[x][y]
one pixel does not depend on the result of the previous, there's no reason you can't do 2, 4, 8, 16 ect pixels inside each loop.
Some compilers can take advantage of this already in doing loop unrolling to utilize MMX or SSE (or similar SIMD instruction sets) instructions. It seems like Trips is an instruction set designed to aid the processor in finding and exploiting such ILP.
The usefullness of such massively parallel designs in general purpose computing is debatable I would say. On the whole there tend to be a lot more instructions with dependancies than those without. (obviously everything has some dependancies, I mean in such a manner that prevents ILP / loop unrolling).
Hardware has been moving towards more parallelism with super-scalar and multi-chip processing and more functional SIMD instruction sets, but software has gone only kicking and screaming into a more parallel world.
Athlon and Pentium 3, Pentium M can look at up to something like 14 x86 instructions and decode up to 3 of them per clock cycle. More often than not they can't find 3 suitable instructions to decode. I have a hard time believeing something is going to find 32 (16 per core, 2 cores on the prototype) for general purpose software.
generally if it's the result of a 'good faith mistake,' ie a clerical error, or improperly labled addresses, evidance obtained in search warrant is admissible even when they end up searching the wrong house / building.
I'd believe the same train of thought would be applicable to this, a good faith effort to tap the right phone and end up with the wrong one would probably still be admissable in court.
One the one hand, yes of course software developers need to be accountable for their work. This isn't bounded by an license or developer. If you release software, you have a responsibility to maintain and support it. I'm not talking about if some one peverts your work into something malicious or if some one uses an unkown exploit. I don't believe developers need be held responsible for damages relating to thos types of situations.
But I do belive if you have software out there, you have a reponsibility to your users to patch security and stability (and privacy and others that might arise) issues in a timely fassion. Barring that, if you sit on your lorells and watch as people use a known bug to do harm with your software I don't have a problem holding a developer responsible for damage. Willfull disregard for the damage your program can have should not be tolerated.
But then again, nailing down when, what and for how long it's reasonable to expect proper support for software is pretty much impossible.
I don't epxect MS to be supporting win95, I don't expect Linux to put much active support into the 1.0 kernel, I expect ATM software to maintained so long as an active ATM using it.
If you ask 100 different people you'll probably get around 150 different answers, picking any one of them would not only be extremely difficult, but probably dangerous as well.
This is a typical PSU review, that is to say worthless. The problem is to do a good PSU reivew you actually need quite a bit of hardware, most little online sites lack even the most basic testing tools (a good multimeter and a controllable load). They make no mention of how they measured the voltages (software, or voltmeter, and from where, pigtail, ATX connector, somewhere else), they put a system that probably doesn't draw 125W DC at load to test out a 500W PSU, they have no real PSU temperature or efficency information. Typical of a site who's reviewing expertiese consists soley of swaping out parts, running 3D Mark and reporting the difference.
Silent PC Review does half way decent reviews, and over the last year or so XBit Labs has starting doing very good PSU reviews. Beyond that there aren't too many places that consistantly hit the mark.
For a silent PSU (not sure why this is that big of a deal, I have a TruePower 330W and can't hear it over the HDD, but I guess some people will always pay for that last dB quieter), there's of course the Phantom 300, the SilverStone 'NF' series, a 300 and a 400W version, the Fortron Source Zen 300; recently reviewed on XBitLabs and Silent PC Review, with just rock solid voltages across the spectrum. And of course the SeaSonic S12 line while not fanless is known to be extremely quiet and highly efficient
Just to follow up on that,
Is the AI going to be as moddable and customizable as the rest of the game content?
I know Mr. Caudill mentioned an 'AI SDK' for 'experienced programmers' over on the IGN Civ 4 preview to tailor the AI to their desires. But it was mentioned as a seperate entity from the XML unit files and the basic Phython scripts.
Is this because the AI is more hard coded (less of it in easily accessible scripts) than say unit stats, or just an attempt to give a helping hand to less experianced modderings in a rather complex enviorment like the AI.
Basically I was hoping you could go into some more detail on what AI and other more complex modding might entail.
The plan is to pick up a second one when the price drops to around $100.
This is sorf of the flaw in a lot of people's SLI plans. (Because Crossfire can use 1 non-crossfire compatible card ATI's solution might not be quite as bad.
A 6800 GT is still about $275-300 right now, the 6800 Ultra is $350 and up.
You can't hardly buy a 5900/U/5950 anymore.
If you don't mind going to ebay or second hand cards for your second one you can probably save a decent amount of money on the second card, of course you already spent a bit of money up front for an SLI motherboard and a PSU to handle a second card. It might end up being about cost effective if you can scrounge up a second card on the cheap in a year or two,
but the 'I'll just pick up a second card when they're cheap' really isn't going to materialize unless Nvidia and ATI change their attidude towards older generations, keeping old chips like the 9800 or 5700 in production and at value prices has been the exception, not the rule. Especially for Nvidia.
The 90nm process intel is using has a very similar thing, only not done near to the extent their talking about here.
Intel has two sets of transistors for 90nm, high voltage threshold and low voltage threshold.
High VT are fairly power efficient as it is, about 40nA/um leakage, Ion about 31 times greater than Ioff (NMOS)
Low VT (which were used extensivly in Prescott to get it to scale to the 4-5ghz range it was intended for), which are horribly inefficient, with a leakage of about 400nm/um, Ion around 3.5 times Ioff (NMOS)
Seems like this is largely a really really high VT transistor, with a few tweaks to the oxide thickness for good measure.
In any case, it should help out the ultra low power devices to an extent, but won't effect any of intel's 65nm desktop/laptop chips. (save maybe a chipset, but I doubt we'd see a 65nm chipset).
If you had RFTA (wait this is /., nevermind),
the voting comes from a shortlist of a few candidates, much like how Fox/American Idle choose a handful of decent candidates (as well as some pathetic morons for amusemnet) from the thousands that apply, so will the Malaysian space program will choose a few qualified candidates and the people will vote on them.
And of course in the great tradition of misleading, or at least sensationalized headlines, the article only says 'the Malaysian space boses will factor the votes into their final decision.'
Ie it's a gimic to get people interested in space, but probably won't have any real effect on the process.
how many new materials, technologies ect have been largely ancillary to the ultimate goal (or even developed without a goal in mind), NASA, DARPA, and similar organizations don't exist to be financialy viable in and of themselves, they exist to reach beyond what we are capable of now, and make it possible. A bit over $8 billion dollars a year for the next 12 years is pocket change for what benifits we can reap. Not in the next 10 years, but in the next 50, or 100 years.
Mars in ten years? Oh, that was just a political tool to move the news cycle from whatever massive screwup the white house was involved in that particular week to grandious dreams of unfunded potential futures.
A very minor update to 2000 to convince people to shell out another $100 for a better looking interface, a couple of moderately usefull features little else?
Isn't that why most of the corperate and even many home users (like myself) of 2000 opted NOT to upgrade at all?
The article was sketchy, maybe smaller expense, smaller expectations make some sense. Less cost (to MS and the consumer I would think) per upgrade, less benifit, decide to upgrade every few years, but MS has part of the user base upgrading all the time, not just in the year or so after a big software release.
The report doesn't consider expenses related to design or marketing, or the fact that high-end chips can sell for more because fewer off the production line can actually run at top speed,
said the article.
Design, research, and marketing can run well over a billion dollars for a single core.
And who knows if that includes things like the upstart cost for new production plants, with Intel building 4 65nm fabs, they could easily be sinking $10-12 billion dollars into the ability to just produce chips to being with.
Materials / labor ect to turn sand into a CPU isn't a trivial cost, but it's not representative of the cost to bring a CPU to market either.
Well x86 chips have pretty well developed methods of dealing the lack of registers.
Register renaming eliminates, or at least minimizes most of the problems with a small register set.
(Athlon64 has something like 72 integer registers and 122 90 bit FP registers (two of these are combined to make an XMM register for SSE vectors), almost all of which are availible in 32 bit mode).
The extra achitectual registers will help with moderate to long term storage (more than a few dozen clock cycles between uses) as the programmer will explicity specify the data remains in the register, where as with current shuffeling it's up to the CPU (and to some extent how the renamed registers are inteded to work) to determine if a write to cache is in order, or not.
And really with the longer storage times, you often have the flexibility to write out to L1 and schedual the load so that there's no penalty for the load. (ie issue the move back to the register the 3 clock cycles prior to when you need it that an L1 load usually takes).
The new registers probably won't make all that much difference in the end. But the again, nothing from the move to 64 bit will be a major impact for a while (at leat on the desktop).
wait hang on.
100GB per platter, 500GB drive.
Umm... doing math...
Hmm.. Ok yep. I'm back to being warry about this drive.
While Hitachi did by IBM's HDD wing, we need to be clear.
The actual "DeathStar" drives were a very select line. IBM tried to put 5 platters into their high capacity 75GXP line, the norm is 4 for 3.5'' disks.
These lead to excessive head crashed (I've heard up to around 30% of the drives met their death this way).
Even before IBM sold the HDD buisness they had gone back to a 4 platter design which effectivley elminated the 'death' part of the deathstar line.
If you like to boycott them based on passed wrongs, that's fine and your call. (Ther are brands I avoid to this day because of past buiness practices). But there are no quality / reliability issues with any of the current Hitachi hard drives.
hmm... Nvidia does have a history of releasing every driver build they can and letting the user toy with them to find what fits best.
But ATI, SoundBlaster, yes I've been using them for about 3 years now. Before that Nvidia and SoundBlaster, only real stability issues I've ever had were the SB Live and 686B southbridge fighting it out...
Most games do well enough to configure their own settings it's been some time since I even made major changes to video / image quality / performance settings.
I install games, they run, I'm happy. It's not as fast as a console, but as for easy, awfully close for me.
Well, sitting at the desktop, playing no games now steam is eating up 20MiB of ram. (not a massive ammount but 5-10 times what the rest of my little background apps take up). I could (and try to) close it when I exit the game, but (to the best of my knowledge) there is no automated way to close steam on exiting HL2/CS so it's an extra hassle to reclaim that memory.
You can't play online if the steam server that holds your account is down (no authentication), which has happened in the past, and per murphy's law will probably happen in the future.
Though I think it's much better now, in the past large updates have been horrendesly slow to come by as the steam servers weren't up to the bandwitdh tasks.
And of course the big one, stop using or switch to a new - and incompatible - authentication software. Like say right after the release of Half life 3. [/tin foil hat]
But you are at the mercy of valve to maintain the steam servers or you loose the ability to play your legally bought games unless you know how to crack it.
Steam isn't a bad thing by itself, I kind of like a lot of what it does. Decent anti-cheat program, seamless updating (both of the anti-cheat and of the games), decent content delivery system (though the pricing sucks).
The problem with steam is that it takes rights away from the consumer and gives them to the developer. Namely I have to ask permission to play my own game.
It essentially comes with *nothing* in the box, seemingly just so they can say they offer a lower pricepoint. The cost of accessories is unjustified. The memory cards hold, what, 64MB? And they cost $40?! I found a 512MB USB thumbdrive on Froogle for that price. Christ, how about just slapping on a USB port and letting us use any generic thumbdrive?
The bare console (particularly early in the life span) tend to be a break even, or loose money proposistion for the manufacturer (MS, Nintendo, Sony). They rely on the 'optional' extras like games, hard drives, memory cards, wireless or extra controllers, faceplates ect, to make their money in the first few years.
If they open such accories up to open competition you'll likely be paying substantially more for the console itself as the manufacturer can't rely on the income from overpriced accessories.
x86-64, Itanium isn't coming to the destkop, even with the software x86 emulator in the latest Win2003 SP, it's no where near fast enough to make a seemless transition.
SSE1/2/3 probably 4 (even if it's an even smaller addition than SSE3 just to say they're doing something new).