I like my ring-tones to actually sound like some sort of ringing. I also prefer ringers with short bursts rather than constant noise so I have a chance to answer without inconveniencing everyone around through the full duration of the ringing period. Most of my custom ringtones are so short that nobody can usually tell the sound's origin until I answer the phone. I personally find musical tones to rank from 'rather' to 'very' annoying.
The other sad/annoying thing is that many carriers have phone features (like bluetooth ringtone and background image download... or even the phone's built-in ringtone editor/composer) locked out so they can push their own network-based equivalents for profit - or at least force their customers to search and download them over grossly overpriced GPRS.
1) boil water (~100C, ~212F) 2) pour on instant or ground coffee
Who are you going to sue if you spill your home-brewed coffee on yourself? I know I can burn myself with freshly boiled water and there is no reason for me to treat other similarly brewed drinks as any less potentially hazardous. Freshly brewed coffee will also be close to 100C even without temperature-controlled standby heaters... so I am guessing brewed-on-command stuff like espressos are served at near-boiling-point when black.
Boiled water can cause serious burns, coffee is made by pouring boiled water on grains therefore freshly brewed hot coffee of any origin is an equally serious hazard. There is no reason for this simple bit of common sense to be any different at home, work, restaurants, coffee shops or anywhere else.
Common sense has been discontinued some years ago, this is simply another sign of common sense becoming increasingly scarce.
The earliest sign I know of is the well-known McDonald hot coffee case. AFAIK, this is the case that made suing everyone else for one's own stupidity and total lack of common sense into common practice.
There is no giving Intel's Architect benefit of the doubt. The problem is that Intel's Marketing and Engineering departments were not expecting process advancements to still stall under 4GHz. They banked on ever increasing GHz figures that never materialized. Remember that Netburst was born in the pursuit of GHz infinity, the replay engine was probably selected for its simplicity and ability to scale with frequency, not for its operational and electrical (in-)efficiency. This is the result of letting Marketeers dictate engineering objectives to perpetuate the "Megahurtz Myth". Intel's chip architects were simply doing what the marketing department told the bosses to tell the architects to do. I wonder how many watts and cycles the replay queue and associated discarded computations cost under load.
Northwood and Prescott are the two most openly documented/reviewed/inspected/analyzed Intel cores out there, in large part because Willamette shocked so many with its awful clock-to-clock performance disadvantage over PPro/P2/P3 cores. This caused all the major sites to do thorough exploration of subsequent desktop chips like Northwood that set a new pipeline depth records and Prescott that was even more controversial for shattering that record.
If Intel restructured Netburst to get rid of the replay queue, they could still turn Netburst into a winner... by simply adding condition flags to execution queue entries, they could postpone issuing uOPs until dependencies are guaranteed to be resolved just-in-time or earlier. That way, they would not need a replay queue of any sort and no uOPs would ever waste execution slots multiple times to produce discarded invalid results while waiting for dependencies.
- A former Intel fanboy and wanabe chip architect. (I am currently scheduled to start my new job as an ASIC validation specialist next January.)
What are the differences between X2 and Opteron? 1) Athlons only have one usable HT link while Opterons have three 2) Opterons require buffered DIMMs 3) Opterons can use ECC DIMMs
Other than that, the two are fundamentally the same. The first A64s (S940) were Opterons with defective HT links or surplus Opteron production with disabled HT links, just like Celerons are Pentiums with slow/defective cache blocks or surplus production lobotomized to avoid flooding the market with high-end parts that would kill ASPs. Pentiums magically disappear from the retail channel at the $178 mark thanks to this artificial scarcity. This is why Intel loves unified cores so much - and probably part of their motive to divorce their dual-core chips: this also spares them the trouble/embarasment of introducing dual-core Celerons to liquidate their surplus Pentium-D cores.
If you compared Athlon 64 with Opteron (and their dual-core equivalents), I would expect the only observable technical differences to boil down to the two extra HT links and buffered ECC DIMM support circuitry in the Opteron.
HT and Netburst were good ideas... but they were poorly executed.
Part of the reason for this is that desktop CPUs mostly run desktop apps and most desktop apps are single-threaded so Intel and AMD could not afford to give up on single-threaded performance. This forced them to add heaps of logic to extract parallelism and Intel made many (IMO dumb) decisions in the process. The SPARC stuff is used for scientific apps which have a long history of multi-threading and distributed computing so Sun does not have to worry about single-threaded performance, allowing for much simpler, leaner and more efficient designs.
Where I think Netburst is particularly bad is the execution engine... when I read Intel's improved hyper-threading patent, I was struck in disbelief: the execution pipelines are wrapped in a replay queue that blindly re-executes uOPs until they successfully execute and retire. Each instruction that fails to retire on the first pass enters the queue and has its execution latency increased by a dozen cycles until its next replay. Once the queue is full, no more uOPs can be issued so the CPU wastes power and cycles re-executing stale uOPs until they retire, causing execution to stall on all threads. Prescott added independant replay queues for each thread so one single thread would never be able to stall the whole CPU by filling the queue... this could have helped Northwood quite a bit but Prescott's extra latency killed any direct gains from it. Intel should roll back to the Northwood pipeline and re-apply the good Prescott stuff like dedicated integer multiplier and barrel shifter, HT2, SSE3 and a few other things, no miracle but it would be much better than the current Prescotts, though it certainly would not help the saturated FSB issue.
With a pure TLP-oriented CPU, there is no need for deep out-of-order execution, no need for branch prediction and no need for speculative execution. Going for TLP throughput allows the CPU to freeze threads whenever there is no nearby code that can execute deterministically instead of doing desperate deep searches, guesses and speculative execution: more likely than not, the other threads will have enough ready-and-able uOPs to fill the gaps and keep all execution units busy producing useful results on nearly every tick. Stick those SPARC chips on a P4-style shared FSB/RAM platform and they would still choke about as bad as P4s do.
The P4's greatest achile's heel is the shared FSB... it was not an issue back when Netburst was running at sub-2GHz speeds but it clearly is not suitable for multi-threading multi-core multi-processor setups. The shared FSB is clearly taking the 'r' out of Netburst. The single-threaded obsession is also costing AMD and Intel a lot of potential performance, complexity and power.
AMD Opterons each have their own local RAM and can access each other's RAM over the HT links to form a a cache-coherent non-uniform memory architecture - ccNUMA.
Multi-core Opterons have a special internal crossbar switch that allow the cores to share the memory controller and HT links, they do not 'piggy back' on the other. This reduces latencies and increases bandwidth for communication between the two cores and gives both cores the equal-opportunity access to the HT ports and CPU's local RAM. With a NUMA-enabled OS, applications will run off the CPU's local RAM whenever possible to minimize bus contention and this allows Opteron servers' overall bandwidth and processing power to scale up almost linearly with the number of CPUs.
As for Intel's dual-cores, the P4 makes sub-optimal use of its very limited available bandwidth. Turning HT on in a quad-core setup where the FSB is already dry on bandwidth naturally only makes things worse by increasing bus contention. Netburst was a good idea but it was poorly executed and the shared FSB very much killed any potential for scalability. If Intel gave the P4 an integrated RAM controller and a true dual-core CPU (two cores connected through a crossbar switch to shared memory and bus controllers like AMD did for the X2s), things would look much better. I'm not buying Intel again until Intel gets this obvious bit of common sense. The CPU is the largest RAM bandwidth consumer in a system, it should have the most direct RAM access possible. Having to fill pipelines and hide latencies with distant RAM wastes many resources and a fair amount of performance - and to make this bad problem worse, Intel is doing this on a shared bus. Things will get a little better with the upcoming dual-bus chipsets with quad-channel FBDIMM but this will still put a hard limit on practical scalability thanks to the non-scalable RAM bandwidth.
On modern high-performance CPUs, shared busses kill scalability. AMD moved towards independant CPU busses with the K7 and integrated RAM controllers with the K8 to swerve around the scalability brick wall Intel was about to crash into many years ago and has kept on ramming ever since. Right now, Intel's future dual-FSB chipset is nothing more than Intel finally catching up with last millenia's dual-processor K7 platforms, only with bigger bandwidth figures.
Same story as HyperThreading... HT was in the Willamette core but not enabled until the 3.06GHz/533 Northwood and the/800 variants. EMT64 is in all Prescotts but was only enabled in the newer revisions.
What annoys me most about Intel and x86-64 is that they only promote the extra memory capacity and dismiss the advantages of being able to natively manipulate long integers which are becoming increasingly more common. Then again, benchmarks tend to say Intel's 64bits execution units are half-baked when compared to the boost Athlon64s get from 64bits code.
I really am looking forward to Windows Vista for Supercomputers where each node will require a DX9-compliant graphics sub-system.
The only application-level interface supercomputing nodes need is very fast MPI. With many supercomputers running often disk-less ultra-lightweight BSD or Linux kernels on computing nodes to leave as much RAM and CPU resources to the software as possible, it would make no sense to have all of Windows' AV-centric overhead.
How much will it cost? Too much performance-wise, too much hardware-wise and too much license-wise.
Win-CSS makes no sense... unless Bill wants to have an entry in the top-500 and is willing to give big scholarships and subventions to offset the costs of renting and deploying Win-CCS licenses.
And HD-DVD is less likely to contain Sony's patent-pending anti-disk-trading scheme where the disks' unlock code is stored in an erasable block and overwritten by the player after the disk is loaded for the first time, making it impossible to use the disks in any other player (trading, lending, selling used games/movies, etc.) afterwards.
I lost the link to the actual patent but it was applied for in H1-2000, I think it was April. The abstract clearly said this was to prevent piracy and trade of used copies. Will this wonderful technology make it into Blu-Ray and the PS3? We'll see after the PS3 launches.
Conker's BFD gets really frustrating around 70% into the game... most of "It's war" is particularly frustrating for an average gamer, the final stretch with the bazookas in particular. The final boss was also a pain in the A55. Other than that, I do not remember having any significant technical/playability problems with the game.
AFAIK, Conker's BFD was the first politically-incorrect game licensed by Nintendo since the day where Nintendo officially announced its family-oriented focus 15+ years ago, somewhere in the NES years.
AMD's fabs are not that cramped... AMD is still in the process of retooling (upgrading) some of them to 65nm/300mm and it also has some fabs that have been built/upgraded but not up to production yet.
AMD's production is a little tight (probably until the 65nm fabs finish qualification) but deals like Cray's are long-term one-off (more or less) contracts, AMD will not have to churn out those 262k (my guess) Opterons overnight. I would also not be surprised if Cray's order happened to be for those near-future 65nm Opterons.
"Sparc's strength in the early Internet days was always throughput - even under load"
Are you implying that you can have useful throughput under no load? How do you measure this idle throughput advantage?
The Intel/AMD architectures are historically single-threaded desktop-centric where the most important thing usually is to run one thing really fast. Sun, however, was always in the HPC/workstation game where overall throughput matters most, latencies and single-thread performance be damned. These two groups were playing pretty different games up to recently.
But now, Intel/AMD have hit a GHz and complexity brick wall. They are forced to promote multi-threading multi-core at the desktop-level and optimize their future desktop chip designs for multi-threaded application throughput rather than single-threaded performance. Imagine what would happen if AMD and Intel could afford to quit competing on single-threaded performance overnight: goodbye complex deep out-of-order execution, goodbye branch-prediction and speculative execution - those transistors would be much better spent on implementing quad-threading cores to keep every pipeline filled with useful instructions that will retire cleanly on every clock.
Sacrificing single-thread performance for simultaneous multi-threaded throughput in the above-described way has been the name of Sun's game for the last few years.
Obsession with single-threaded performance is what costs current x86 CPUs the most power. Of course, in the P4/HT case, there is the added power and transistor costs of trying to be a jack-of-all-trades who predictably turned out as a master-of-none. (The P4's uOP replay engine is a neat idea... but re-executing the same stupid uOPs until they meet retirement conditions is woefully wasteful, whoever designed and bothered to patent this should be fired.)
Car manufacturers also win... they get to sell more indecently marked-up cars.
Hybrids offer many opportunities for cost-cutting: they can have smaller engines, these can be tuned to operate only within a very narrow high-efficiency range, motor-wheels remove the need for a drive train, etc. Hybrid manufacturers charge over $6000 more for the privilege but their actual production cost overhead is probably under $2000 and it will only go down with volume and improvements.
Now, if hybrids had a $2000 premium, they would be worth considering even in NA. Until then or until gas prices go over $2.50/L, going hybrid makes no economical sense - as TFA said, the extra interests simply gobbles up the savings. I did these calculations once some time ago and concluded that I would have to own the car for ~10 years to recover my initial investment.
The best is single-mode glass (the stuff used for single multi-gigabit 10+ kilometers long stretches with ~2dB/km loss which is highly breakable and fairly expensive, the equally breakable but less expensive multi-mode glass fibers are limited to about 1Gbps and typically less than 5km due to modal dispersion and ~10dB/km losses while bend-tolerant and much less expensive plastic fiber with ~20dB/km losses are limited to only a few hundred meters for applications that need gigabit speeds. For residential 100Mbps service, plastic fiber would be plenty good enough to handle the last km from the nearest back-haul switch to subscriber terminal. (The loss figures are what I remember from a class where they were mentionned nearly 10 years ago.)
And yes, you can bend most plastic fiber on less than a 3cm radius. Sharp bends do increase losses but plastic fibers at least let you back off with little to no permanent damage instead of spontaneously breaking like glass fibers would if you went a little too far.
The Sharp LCD was goggle-less thanks to a special surface finishing and double LCD matrix.
This IBM '3D' display is the same old alternate-frame display trick we have had for years... look at the older nVidia cards, the 'Deluxe' models had an extra port for 3D goggles. This is the same technology, only with an IBM spin. Most modern mid/high-end CRTs could already handle alternate-frame at ~100fps up to at least 1280x960 but video card manufacturers have apparently given up on the 3D goggles and people have forgotten them.
Sony applied for a patent in 2000 on a method for one-time readable data on Blu-Ray disks (basically, the playback device reads an ID sector then overwrites it) which means that PS3 games may be made not be transferable/reselable. This could also imply that if your PS3 goes poof, you are stuck having to re-buy your favourite games.
I'd like to see them (or anyone else) try that out and get publicly executed for it.
Tell me how you can rear-end vehicles from on-coming traffic and I will tell you.
When following a vehicle, you are supposed to keep enough distance to avoid stuff that may fall from it. This is pretty hard to apply to on-coming traffic or to over-taking.
For trucks, I either follow at 2-3X my normal distance whenever I can. As for overtaking them, this is generally unnecessary on highways around here because most local truck drivers are lunatics who typically drive ~10km/h faster than everyone else who is already driving 10-20km/h over the limit.
If you are behind (I did specify the rear-ender case) you can avoid gravel (or rear-ending) by keeping a safe distance appropriate to the circumstances - enough distance for falling pieces to settle on the ground. If you are passing by the on-coming truck, it is no longer a rear-ender type situation but as you said, the truck is already too far for positive identification by the time you hear/see the hit, making it pretty hard to file claims against the driver/company who operated/owned the vehicle.
This flat-bed truck example is right along the lines of rear-ender guilty by default: if you stayed sufficiently far behind the next vehicle, it wouldn't have happened. Same thing for gravel, if you kept your distances, falling pieces would reach the ground instead of your car. Good luck trying to win such a case, all it says to the insurance companies is that you have most likely misjudged your safe distance or were not keeping it.
Live off patents, get killed by patents. It has become one of those "the cure is worse than the disease" situations. The RIAA/MPAA/etc. are headed down the same roads with their lovely uninteroperable hardened DRM wet dreams, analog hole butt-plug and lock-in policies fantasms.
Slashdot Mirror
I like my ring-tones to actually sound like some sort of ringing. I also prefer ringers with short bursts rather than constant noise so I have a chance to answer without inconveniencing everyone around through the full duration of the ringing period. Most of my custom ringtones are so short that nobody can usually tell the sound's origin until I answer the phone. I personally find musical tones to rank from 'rather' to 'very' annoying.
The other sad/annoying thing is that many carriers have phone features (like bluetooth ringtone and background image download... or even the phone's built-in ringtone editor/composer) locked out so they can push their own network-based equivalents for profit - or at least force their customers to search and download them over grossly overpriced GPRS.
How do you brew coffee at home?
1) boil water (~100C, ~212F)
2) pour on instant or ground coffee
Who are you going to sue if you spill your home-brewed coffee on yourself? I know I can burn myself with freshly boiled water and there is no reason for me to treat other similarly brewed drinks as any less potentially hazardous. Freshly brewed coffee will also be close to 100C even without temperature-controlled standby heaters... so I am guessing brewed-on-command stuff like espressos are served at near-boiling-point when black.
Boiled water can cause serious burns, coffee is made by pouring boiled water on grains therefore freshly brewed hot coffee of any origin is an equally serious hazard. There is no reason for this simple bit of common sense to be any different at home, work, restaurants, coffee shops or anywhere else.
Common sense has been discontinued some years ago, this is simply another sign of common sense becoming increasingly scarce.
The earliest sign I know of is the well-known McDonald hot coffee case. AFAIK, this is the case that made suing everyone else for one's own stupidity and total lack of common sense into common practice.
http://www.amd.com/us-en/Processors/ProductInforma tion/0,,30_118_809_4368,00.html
"Smart MP technology for smarter multiprocessing:
* Dual point-to-point, high-speed system buses
* Innovative bus snooping capability
* Optimized MOESI cache coherency protocol"
The K7 had point-to-point CPU-to-chipset bus and (IIRC) an auxiliary CPU-to-CPU snooping bus.
There is no giving Intel's Architect benefit of the doubt. The problem is that Intel's Marketing and Engineering departments were not expecting process advancements to still stall under 4GHz. They banked on ever increasing GHz figures that never materialized. Remember that Netburst was born in the pursuit of GHz infinity, the replay engine was probably selected for its simplicity and ability to scale with frequency, not for its operational and electrical (in-)efficiency. This is the result of letting Marketeers dictate engineering objectives to perpetuate the "Megahurtz Myth". Intel's chip architects were simply doing what the marketing department told the bosses to tell the architects to do. I wonder how many watts and cycles the replay queue and associated discarded computations cost under load.
Northwood and Prescott are the two most openly documented/reviewed/inspected/analyzed Intel cores out there, in large part because Willamette shocked so many with its awful clock-to-clock performance disadvantage over PPro/P2/P3 cores. This caused all the major sites to do thorough exploration of subsequent desktop chips like Northwood that set a new pipeline depth records and Prescott that was even more controversial for shattering that record.
If Intel restructured Netburst to get rid of the replay queue, they could still turn Netburst into a winner... by simply adding condition flags to execution queue entries, they could postpone issuing uOPs until dependencies are guaranteed to be resolved just-in-time or earlier. That way, they would not need a replay queue of any sort and no uOPs would ever waste execution slots multiple times to produce discarded invalid results while waiting for dependencies.
- A former Intel fanboy and wanabe chip architect.
(I am currently scheduled to start my new job as an ASIC validation specialist next January.)
What are the differences between X2 and Opteron?
1) Athlons only have one usable HT link while Opterons have three
2) Opterons require buffered DIMMs
3) Opterons can use ECC DIMMs
Other than that, the two are fundamentally the same. The first A64s (S940) were Opterons with defective HT links or surplus Opteron production with disabled HT links, just like Celerons are Pentiums with slow/defective cache blocks or surplus production lobotomized to avoid flooding the market with high-end parts that would kill ASPs. Pentiums magically disappear from the retail channel at the $178 mark thanks to this artificial scarcity. This is why Intel loves unified cores so much - and probably part of their motive to divorce their dual-core chips: this also spares them the trouble/embarasment of introducing dual-core Celerons to liquidate their surplus Pentium-D cores.
If you compared Athlon 64 with Opteron (and their dual-core equivalents), I would expect the only observable technical differences to boil down to the two extra HT links and buffered ECC DIMM support circuitry in the Opteron.
HT and Netburst were good ideas... but they were poorly executed.
Part of the reason for this is that desktop CPUs mostly run desktop apps and most desktop apps are single-threaded so Intel and AMD could not afford to give up on single-threaded performance. This forced them to add heaps of logic to extract parallelism and Intel made many (IMO dumb) decisions in the process. The SPARC stuff is used for scientific apps which have a long history of multi-threading and distributed computing so Sun does not have to worry about single-threaded performance, allowing for much simpler, leaner and more efficient designs.
Where I think Netburst is particularly bad is the execution engine... when I read Intel's improved hyper-threading patent, I was struck in disbelief: the execution pipelines are wrapped in a replay queue that blindly re-executes uOPs until they successfully execute and retire. Each instruction that fails to retire on the first pass enters the queue and has its execution latency increased by a dozen cycles until its next replay. Once the queue is full, no more uOPs can be issued so the CPU wastes power and cycles re-executing stale uOPs until they retire, causing execution to stall on all threads. Prescott added independant replay queues for each thread so one single thread would never be able to stall the whole CPU by filling the queue... this could have helped Northwood quite a bit but Prescott's extra latency killed any direct gains from it. Intel should roll back to the Northwood pipeline and re-apply the good Prescott stuff like dedicated integer multiplier and barrel shifter, HT2, SSE3 and a few other things, no miracle but it would be much better than the current Prescotts, though it certainly would not help the saturated FSB issue.
With a pure TLP-oriented CPU, there is no need for deep out-of-order execution, no need for branch prediction and no need for speculative execution. Going for TLP throughput allows the CPU to freeze threads whenever there is no nearby code that can execute deterministically instead of doing desperate deep searches, guesses and speculative execution: more likely than not, the other threads will have enough ready-and-able uOPs to fill the gaps and keep all execution units busy producing useful results on nearly every tick. Stick those SPARC chips on a P4-style shared FSB/RAM platform and they would still choke about as bad as P4s do.
The P4's greatest achile's heel is the shared FSB... it was not an issue back when Netburst was running at sub-2GHz speeds but it clearly is not suitable for multi-threading multi-core multi-processor setups. The shared FSB is clearly taking the 'r' out of Netburst. The single-threaded obsession is also costing AMD and Intel a lot of potential performance, complexity and power.
AMD Opterons each have their own local RAM and can access each other's RAM over the HT links to form a a cache-coherent non-uniform memory architecture - ccNUMA.
Multi-core Opterons have a special internal crossbar switch that allow the cores to share the memory controller and HT links, they do not 'piggy back' on the other. This reduces latencies and increases bandwidth for communication between the two cores and gives both cores the equal-opportunity access to the HT ports and CPU's local RAM. With a NUMA-enabled OS, applications will run off the CPU's local RAM whenever possible to minimize bus contention and this allows Opteron servers' overall bandwidth and processing power to scale up almost linearly with the number of CPUs.
As for Intel's dual-cores, the P4 makes sub-optimal use of its very limited available bandwidth. Turning HT on in a quad-core setup where the FSB is already dry on bandwidth naturally only makes things worse by increasing bus contention. Netburst was a good idea but it was poorly executed and the shared FSB very much killed any potential for scalability. If Intel gave the P4 an integrated RAM controller and a true dual-core CPU (two cores connected through a crossbar switch to shared memory and bus controllers like AMD did for the X2s), things would look much better. I'm not buying Intel again until Intel gets this obvious bit of common sense. The CPU is the largest RAM bandwidth consumer in a system, it should have the most direct RAM access possible. Having to fill pipelines and hide latencies with distant RAM wastes many resources and a fair amount of performance - and to make this bad problem worse, Intel is doing this on a shared bus. Things will get a little better with the upcoming dual-bus chipsets with quad-channel FBDIMM but this will still put a hard limit on practical scalability thanks to the non-scalable RAM bandwidth.
On modern high-performance CPUs, shared busses kill scalability. AMD moved towards independant CPU busses with the K7 and integrated RAM controllers with the K8 to swerve around the scalability brick wall Intel was about to crash into many years ago and has kept on ramming ever since. Right now, Intel's future dual-FSB chipset is nothing more than Intel finally catching up with last millenia's dual-processor K7 platforms, only with bigger bandwidth figures.
Same story as HyperThreading... HT was in the Willamette core but not enabled until the 3.06GHz/533 Northwood and the /800 variants. EMT64 is in all Prescotts but was only enabled in the newer revisions.
What annoys me most about Intel and x86-64 is that they only promote the extra memory capacity and dismiss the advantages of being able to natively manipulate long integers which are becoming increasingly more common. Then again, benchmarks tend to say Intel's 64bits execution units are half-baked when compared to the boost Athlon64s get from 64bits code.
Compared to the 12V/60-100A that goes into an electrically-coupled air conditionner, it would be even more negligible.
I really am looking forward to Windows Vista for Supercomputers where each node will require a DX9-compliant graphics sub-system.
The only application-level interface supercomputing nodes need is very fast MPI. With many supercomputers running often disk-less ultra-lightweight BSD or Linux kernels on computing nodes to leave as much RAM and CPU resources to the software as possible, it would make no sense to have all of Windows' AV-centric overhead.
How much will it cost? Too much performance-wise, too much hardware-wise and too much license-wise.
Win-CSS makes no sense... unless Bill wants to have an entry in the top-500 and is willing to give big scholarships and subventions to offset the costs of renting and deploying Win-CCS licenses.
And HD-DVD is less likely to contain Sony's patent-pending anti-disk-trading scheme where the disks' unlock code is stored in an erasable block and overwritten by the player after the disk is loaded for the first time, making it impossible to use the disks in any other player (trading, lending, selling used games/movies, etc.) afterwards.
I lost the link to the actual patent but it was applied for in H1-2000, I think it was April. The abstract clearly said this was to prevent piracy and trade of used copies. Will this wonderful technology make it into Blu-Ray and the PS3? We'll see after the PS3 launches.
Conker's BFD gets really frustrating around 70% into the game... most of "It's war" is particularly frustrating for an average gamer, the final stretch with the bazookas in particular. The final boss was also a pain in the A55. Other than that, I do not remember having any significant technical/playability problems with the game.
AFAIK, Conker's BFD was the first politically-incorrect game licensed by Nintendo since the day where Nintendo officially announced its family-oriented focus 15+ years ago, somewhere in the NES years.
AMD's fabs are not that cramped... AMD is still in the process of retooling (upgrading) some of them to 65nm/300mm and it also has some fabs that have been built/upgraded but not up to production yet.
AMD's production is a little tight (probably until the 65nm fabs finish qualification) but deals like Cray's are long-term one-off (more or less) contracts, AMD will not have to churn out those 262k (my guess) Opterons overnight. I would also not be surprised if Cray's order happened to be for those near-future 65nm Opterons.
Opteron making major inroads into the HPC market is bad for IBM's Power4/5 and lesser players like NEC too.
"Sparc's strength in the early Internet days was always throughput - even under load"
Are you implying that you can have useful throughput under no load? How do you measure this idle throughput advantage?
The Intel/AMD architectures are historically single-threaded desktop-centric where the most important thing usually is to run one thing really fast. Sun, however, was always in the HPC/workstation game where overall throughput matters most, latencies and single-thread performance be damned. These two groups were playing pretty different games up to recently.
But now, Intel/AMD have hit a GHz and complexity brick wall. They are forced to promote multi-threading multi-core at the desktop-level and optimize their future desktop chip designs for multi-threaded application throughput rather than single-threaded performance. Imagine what would happen if AMD and Intel could afford to quit competing on single-threaded performance overnight: goodbye complex deep out-of-order execution, goodbye branch-prediction and speculative execution - those transistors would be much better spent on implementing quad-threading cores to keep every pipeline filled with useful instructions that will retire cleanly on every clock.
Sacrificing single-thread performance for simultaneous multi-threaded throughput in the above-described way has been the name of Sun's game for the last few years.
Obsession with single-threaded performance is what costs current x86 CPUs the most power. Of course, in the P4/HT case, there is the added power and transistor costs of trying to be a jack-of-all-trades who predictably turned out as a master-of-none. (The P4's uOP replay engine is a neat idea... but re-executing the same stupid uOPs until they meet retirement conditions is woefully wasteful, whoever designed and bothered to patent this should be fired.)
Car manufacturers also win... they get to sell more indecently marked-up cars.
Hybrids offer many opportunities for cost-cutting: they can have smaller engines, these can be tuned to operate only within a very narrow high-efficiency range, motor-wheels remove the need for a drive train, etc. Hybrid manufacturers charge over $6000 more for the privilege but their actual production cost overhead is probably under $2000 and it will only go down with volume and improvements.
Now, if hybrids had a $2000 premium, they would be worth considering even in NA. Until then or until gas prices go over $2.50/L, going hybrid makes no economical sense - as TFA said, the extra interests simply gobbles up the savings. I did these calculations once some time ago and concluded that I would have to own the car for ~10 years to recover my initial investment.
There are many types of optical fiber.
The best is single-mode glass (the stuff used for single multi-gigabit 10+ kilometers long stretches with ~2dB/km loss which is highly breakable and fairly expensive, the equally breakable but less expensive multi-mode glass fibers are limited to about 1Gbps and typically less than 5km due to modal dispersion and ~10dB/km losses while bend-tolerant and much less expensive plastic fiber with ~20dB/km losses are limited to only a few hundred meters for applications that need gigabit speeds. For residential 100Mbps service, plastic fiber would be plenty good enough to handle the last km from the nearest back-haul switch to subscriber terminal. (The loss figures are what I remember from a class where they were mentionned nearly 10 years ago.)
And yes, you can bend most plastic fiber on less than a 3cm radius. Sharp bends do increase losses but plastic fibers at least let you back off with little to no permanent damage instead of spontaneously breaking like glass fibers would if you went a little too far.
The Sharp LCD was goggle-less thanks to a special surface finishing and double LCD matrix.
This IBM '3D' display is the same old alternate-frame display trick we have had for years... look at the older nVidia cards, the 'Deluxe' models had an extra port for 3D goggles. This is the same technology, only with an IBM spin. Most modern mid/high-end CRTs could already handle alternate-frame at ~100fps up to at least 1280x960 but video card manufacturers have apparently given up on the 3D goggles and people have forgotten them.
Come the PS3 and console-locked disks.
Sony applied for a patent in 2000 on a method for one-time readable data on Blu-Ray disks (basically, the playback device reads an ID sector then overwrites it) which means that PS3 games may be made not be transferable/reselable. This could also imply that if your PS3 goes poof, you are stuck having to re-buy your favourite games.
I'd like to see them (or anyone else) try that out and get publicly executed for it.
All my incriminating records have 2048bits AES encryption, they should shoot for 90 years instead.
I am guessing 90 days is for cracking NTFS, ZIP and other similar commodity encryption schemes.
Tell me how you can rear-end vehicles from on-coming traffic and I will tell you.
When following a vehicle, you are supposed to keep enough distance to avoid stuff that may fall from it. This is pretty hard to apply to on-coming traffic or to over-taking.
For trucks, I either follow at 2-3X my normal distance whenever I can. As for overtaking them, this is generally unnecessary on highways around here because most local truck drivers are lunatics who typically drive ~10km/h faster than everyone else who is already driving 10-20km/h over the limit.
If you are behind (I did specify the rear-ender case) you can avoid gravel (or rear-ending) by keeping a safe distance appropriate to the circumstances - enough distance for falling pieces to settle on the ground. If you are passing by the on-coming truck, it is no longer a rear-ender type situation but as you said, the truck is already too far for positive identification by the time you hear/see the hit, making it pretty hard to file claims against the driver/company who operated/owned the vehicle.
This flat-bed truck example is right along the lines of rear-ender guilty by default: if you stayed sufficiently far behind the next vehicle, it wouldn't have happened. Same thing for gravel, if you kept your distances, falling pieces would reach the ground instead of your car. Good luck trying to win such a case, all it says to the insurance companies is that you have most likely misjudged your safe distance or were not keeping it.
Gotta love poetic justice.
Live off patents, get killed by patents. It has become one of those "the cure is worse than the disease" situations. The RIAA/MPAA/etc. are headed down the same roads with their lovely uninteroperable hardened DRM wet dreams, analog hole butt-plug and lock-in policies fantasms.