There are a variety of efforts(different organizations and programs involved at different times) that do just that. Especially in areas moist enough that you can't just do a 'don't leave stagnant water sitting in containers/gutters/etc' campaign to eliminate much of the mosquito breeding area, bed nets are the low-hanging-fruit in terms of reducing the average number of bites per person, especially when you consider how cheap they are and how long they last unless abused(the insecticide-impregnated ones do eventually turn into normal ones; but are still mechanically effective).
I assume that the vaccine efforts are partially a matter of "Well, I'm an immunologist not a field health/education worker, so what am I best suited to do?", partially a matter of protecting people during the time they aren't in bed; and perhaps also the hope of eventually making a sufficient portion of humans resistant and crashing the population of malaria causing protozoa entirely. P. knowlesi unfortunately has an animal reservoir(some non-human primates); but some of the other common plasmodia don't, so if you could increase resistance enough you might be able to hit the point of substantial additional gains 'for free' as the number of infected mosquitos drops and the population crashes.
Aside from immediate considerations, working on a malaria vaccine probably gets some additional interest because of its greater value(both humanitarian and commercial) if climate change should cause the current range of the disease(mostly ghastly tropical pestholes filled with people who can't afford expensive drugs) into wealthier areas of the world. There may also be a basic-research interest: unlike most pathogens, plasmodia are eukaryotic; so I'm sure that the relevant specialists find all sorts of fascinating differences between the biology of the pathogen/host interaction in malaria vs. that in infections by bacteria or viruses. You aren't going to commercialize a drug on basic research alone; but if you want research to happen it certainly doesn't hurt to be novel and interesting.
My pleasure. I'm always glad to see a discussion take a turn for the better rather than just sliding off the rails. Unfortunately, it seems as though the value of known-inaccurate simplified models is often enough poorly understood that some people treat them as "Haha, your model doesn't happen in real life, therefore Economics Refuted!" and others treat them as though their results can actually be trusted when talking about the real-world situations that they are intended to help analyze.
In this case, perfect information is obviously not happening(if nothing else, you'd be crowned God-Emperor of HR for all eternity if you actually found a way of objectively ranking an employee's expertise with enough precision to justify the difference between their salary and the category average down to the last dollar, or even the nearest $10k in a lot of cases); but it does seem like a pretty decent example of how a situation goes from being substantially not-'free-market'(information is both imperfect and asymmetric, with Google knowing all the salaries and each employee knowing only their salary) to one that is markedly closer to 'free market'(Google knows all the salaries, each employee knows at least a fair number of salaries; and is negotiating from a position of much better price information).
I admit that my initial post was pretty snippy; I get annoyed at the cries of "SOCIALISM!!!", especially now that the Cold War is over, all the 'communist' states have either collapsed or turned into crony-capitalist states of various flavors; and the closest thing you can find to 'socialism' is capitalist countries with comparatively cushy social safety nets; and whoever the AC was pushed my buttons.
That specific annoyance aside, though, I'm actually rather fascinated by how useful(across a wide variety of disciplines) models that we know are false can be, despite their falsehood. They are wrong; but by being wrong in well defined ways that are amenable to (relatively) simple analysis they can be such a good jumping off point for examining the real world and figuring out how it must be different in order to produce the results you see.
I find the 'esport' label irksome, mostly because I've never understood the big deal about watching sports; but if the defining characteristic of 'sport-ness' were physical rigor; coal mining and working in a sweatshop would be major athletic events and they'd force golfers to walk the entire course and carry their own bags.
Gaming is obviously pretty low intensity for most muscles, though the rigor of the mental drill is considerable; and the amount of carpal tunnel and similar injuries are actually alarmingly high.
The alternative would obviously be untenable(either forcing athletes to do without medical care 'for their own protection' or just banning every sickie who needs a drug that might be performance enhancing); but a therapeutic use exemption for psychostimulants is going to make the rule more or less a joke(not that I have a problem with that, personally). Getting a diagnosis for which one of the stimulants is the usual treatment is pretty trivial; and they are cheap, have lots of safety data available, and generally don't raise any red flags among doctors. It depends on where you are, of course; but they might actually be among the few drugs that are easier to get legally than illegally.
Scrabble is only a game about words at fairly low levels of play. If you have two otherwise unprepared people stuck in a room with nothing but scrabble for amusement, yes, the one with the better vocabulary likely has better options. Outside of the amateurs, though, memorization of the approved dictionary(starting with words chosen for good point values, the ability to dispose of letters that are usually tricky to get rid of, and other helpful features; but ideally progressing to all of them) supplants knowledge of the language and the remaining challenge is board control and optimizing the conversion of tiles into points over the course of the game.
There would certainly be additional prep time, even for the unusual characters who are really good at this; but the skills that the game demands for high level play should be transferable to any language(or even a nonsense dictionary) that works reasonably well with representation by a relatively small alphabet.
"Perfect information" is the idealized model. As with any idealized model, economics or otherwise(trajectories with respect to a single point mass in absence of friction, ideal gasses, etc.) it sacrifices real-world attainability for substantial convenience in in building and analyzing the model.
Once you have the idealized model, you have something with which to compare real world outcomes and a basis for studying how and why they deviate from the idealized version. Is it barriers to entry? Asymmetric information? Bounded rationality on the part of some or all actors? That's where the economists who grovel through data come in.
In this case, the point isn't that perfect information is expected of real-world economic happenings; perfect information is not possible in practice. However, 'sharing salary data' is an absolutely textbook example of something that would move the situation from 'very imperfect data' to 'closer to perfect data'. It also likely reduces the asymmetry of information(HQ already knew all the salaries, and possibly some at competing companies as well; these employees now have better information and information that is closer to parity with the actor they are negotiating with).
The purpose of idealized models is not to a deliberately unanswerable demand "If it isn't a 'Free Market' the market isn't free!"; but to act as a simplified analytical tool that allows you to focus more clearly on the aspects of the real market that are most or least like their ideal counterparts and tease out how the non-ideal behavior changes the outcomes.
You do realize that 'perfect information' is one of the defining characteristics of the idealized model of 'free market' behavior? You don't have to like it; but calling anything vaguely related to money that displeases you 'socialistic' is dumb beyond words.
We, the fine folks at the interactive advertising bureau, are delighted by the notion of a 'consumer preference client'. Indeed, we are so strongly committed to it that we recommend that it be incorporated at the hardware level, in order to provide additional trust in the 'Trusted Platform' that forms the foundation of the secure online marketplaces of tomorrow. With a suitably immutable GUID baked into every piece of hardware, we can finally ensure that each and every consumer receives exactly the privacy settings and offers most relevant to them!
How could no one have foreseen the potential abuse and pitfalls of a system like this? Without even reading any further than "Giving Doctors Grades..." I immediately conjured images of a bunch of doctors huddled around each other saying, "I don't want that one." "Well I don't want that one either. My feedback is back at 85% and I can't risk another death screwing me over."
Given the enthusiasm for what is more or less exactly the same plan applied to teachers, it's hard to be too surprised that somebody would think that this is a good idea; though it doesn't shed any light on why they would have ignored the obvious pitfalls.
I can only assume that it's yet another instance of the "We want to measure stuff; but what we really want to measure is hard, so we'll settle for something that is easy to measure but not actually helpful" problem that bedevils quantification efforts of all kinds.
I think that he was hoping for a good, old-school, insertion of knife into abdomen and pulling all the way across to demonstrate contrition through horribly painful death; in the fine Japanese tradition.
I would actually be interested to know what the logic is here: the hacker clearly doesn't like AM, or they wouldn't be spoiling their rumored-IPO quite this enthusiastically, they also don't like the users they are threatening to expose; but they also appear to be really bent out of shape about AM's allegedly-dishonest-and-exploitative 'pay to purge the embarrassing traces' feature.
Anger about that feature would seem to be something more likely in some portion of the users, or among people who identify with the interests of the users; but this interested party displays only contempt for them; rather than viewing AM's attempt to squeeze them as an amusing and justified punishment.
We obviously have no particular reason to trust their statement; but we do have to expect that they have a reason worth the legal exposure for doing this(especially since the dataset they are talking about would probably be worth a decent sum for sale to others looking for really juicy spearphishing targets ) rather than not attempting the hack at all or hacking but then staying quiet about it. My guess would be that it is more about attacking the site operator than about the users specifically; it is pretty common for at least a person or two to end up suitably embittered during the course of business.
Probably not directly. To the degree that Microsoft has any specific plan to limit game adoption on non-Windows platforms it is called 'DirectX'. It is the first-class set of APIs on Windows and any games developed for it, or drivers developed to support it, are obviously resources dedicated to gaming being better on Windows and either unavailable or produced at additional cost for OpenGL elsewhere.
Once you get into how AMD's OpenGL driver does(or doesn't) apply application specific optimizations for different OpenGL games, though, MS doesn't have nearly as much to gain from any specific meddling. The general success of DirectX and Windows gaming is presumably the reason why AMD cares relatively little(along with the fact that people looking to use proprietary drivers on Linux usually go Nvidia, while AMD is regarded as very much the second choice unless you are looking for the vendor more cooperative with FOSS driver development).
I doubt that there is an actual price list. At least on the Windows side, releasing driver optimizations(sometimes including overt cheating if the popular benchmarking programs of the day are detected, though that seems to be rarer in recent years) to support popular programs and games is something that both Nvidia and ATI do reasonably routinely in order to improve their products' perceived competitiveness. You don't get a profile unless you have moved enough units to make it worth the effort, or there is a particularly embarrassing performance discrepancy; but the GPU vendors are sufficiently concerned about appearances that just stalling you is a poor choice. If anything, it appears to be that the GPU guys are the ones pushing for greater prominence in game, hence the obnoxious little videos that roll when you start some games and the Nvidia's enthusiasm for PhysX and related middleware being used.
On Linux, it looks like AMD doesn't care very much(I assume that the Linux gaming numbers don't justify much engineering labor for specific titles; but if you have a 'Half-Life 2' optimization it wouldn't be terribly laborious to have the intern look up other Source-based games and add those to the list), but the same reasons for AMD's apathy would also discourage any one vendor from paying very much to get on 'the list'.
That said, I have no reason to doubt that a vendor would outright refuse you if you demanded to pay your way onto the list(especially if you provided both cash and access to any information they needed about your application or from your developers); but unless it is actively all kinds of horribly broken without special optimizations, in which case you have a problem, simply selling enough copies makes supporting your application part of the job of selling GPUs. Since Intel is just waiting to scoop up everyone who doesn't care very much about graphics cards, neither vendor can really afford to withold optimizations and take a hammering in reputation just to squeeze a little protection money out of a publisher.
Another handy advantage to moving away from 10base5 and 10base2! Much trickier to open and tap a bunch of twisted pairs; and the system will not be happy with having two devices on the same cable unless your spybot is quite unobtrusive in any sniffing and spoofing it does. Vampire tapping coax would be much easier to automate with reasonable simplicity and reliability and the system is intended to work with multiple devices on the same wire.
Would you care to go into more detail about what brilliant innovations we are missing out on? The market in remote access Trojans and spyware remains open, the Boing subsidiary with the drone(who were the ones proposing a drone-mounted malicious wifi device) remains open, suppliers of ever smaller computers and wifi systems remain open. 'Hacking Team' may have been reasonably competent crackers, if not so hot on defense; but their(hopeful) demise would be a "and nothing of value was lost and we were all slightly more secure for it" situation if I've ever seen one.
That's actually a bit surprising. Microsoft sure went to a lot of trouble with their precious 'protected media path', so if their own application is merrily recording a DRMed stream(as I believe Netflix is on all supported platforms); either they've screwed up or Netflix couldn't be bothered to use the feature. I imagine that re-compressed copies of streams aren't terribly high priority; but I would have imagined that they'd be contractually obligated to at least pretend to care.
Unfortunately, you still need to use the correct incantations(and sometimes some annoyingly expensive and specific material components) if you want the magic to work properly.
I assume that the problem is with noticing the fire. A small Li-ion battery can self-ignite and burn fairly enthusiastically; but isn't too dangerous if it is prevented from setting anything else on fire. The smoke is noxious and any one directly exposed the the flame will be burned; but it just isn't a very big fire. If the battery is hiding down in the cargo hold in somebody's suitcase, it has a better chance of recruiting all the nearby luggage and getting a proper fire started; at which point suppression becomes more difficult and release of enough energy to actually damage the aircraft becomes likely.
I'd be interested to know what the current standard for fire detection in the cargo area is; and how difficult and costly it would be to achieve better early warning.
Your options really depend on what sort of 'high performance' you have in mind. When it comes to performance per core, Xeons typically crush Opterons; but the pricing reflects that, especially if you need the 4-8 socket support and RAS features. If what you need is large amounts of RAM with the lowest possible spending on the system around it, Opterons have tepid performance per core; but are likely to be the cheapest option that still supports ECC, more than one socket, buffered DIMMs, and any other niceties you wouldn't get from just desktop. If your application is one that can be made to fit, GPUs are enormously powerful for the range of things that they are capable of doing well.
They also depend on how big you need your system to be and how tightly coupled it has to be. If your intended application handles its own network-level parallelization and doesn't depend on very low latency, blessed are you. Price per core skyrockets if you go above 2 sockets, and GbE is effectively free(at least in the sense that you pretty much can't buy a system or motherboard that doesn't come with at least 2 NICs by default, often more) and relatively cheap to switch. If you need lower latency, this will hurt more and you are looking at myrinet or infiniband. If your application needs a cluster that presents a single system image; especially one that also has genuinely low latency, you probably need to fortify your checkbook and consult an expert. You can get systems with more than 8 sockets and the appropriate custom interconnect; but you won't like paying for one.
Unless you value your time at surprisingly low rates; you probably won't want to build your own systems from parts; but depending on how tightly coupled you need, this may be something that you need to purchase as a system or something you construct from multiple computers you purchase.
Can you use either hardware you have or AWS(or one of their similar competitors) to better characterize what your application actually needs?
Given that LEDs already lean pretty enthusiastically on those elements, without too much comment; and some of the stuff they bake into the silicon is pretty dreadful I imagine that they'll diplomatically ignore the issue and hope really hard that the Indium content is high enough to make the semiconductors that aren't super small and embedded in epoxy economically recyclable.
Getting changes made when there are at least partial alternatives and the material is distributed in substantial quantities throughout the entire product is much less of an uphill fight than getting changes made to the relatively small chunks of the product where much of the value and the profit are; and the producers say they've exhausted the options. They would probably be loathe to push it unless their position were markedly stronger than it appears.
I'm sure that that's true on total sale, less sure that it's true on profit(you obviously drop out of the 'buy license from ARM, tell TMSC to make it, attempt to sell it' business if you can't at least keep the lights on; but I suspect that Allwinner and Mediatek aren't exactly commanding Xeon margins); but either way Intel has less pressure on the high end. If you do need a relatively powerful processor you options are pretty much 'Intel', a substantial chunk of dead space, 'maybe AMD', then another chunk of dead space, and then one of the 64-bit ARM people.
That isn't really so scary. On the low end, though, they have plenty of alternatives; some quite possibly overtly superior and definitely plenty that are cheaper. That is more scary.
Slashdot Mirror
There are a variety of efforts(different organizations and programs involved at different times) that do just that. Especially in areas moist enough that you can't just do a 'don't leave stagnant water sitting in containers/gutters/etc' campaign to eliminate much of the mosquito breeding area, bed nets are the low-hanging-fruit in terms of reducing the average number of bites per person, especially when you consider how cheap they are and how long they last unless abused(the insecticide-impregnated ones do eventually turn into normal ones; but are still mechanically effective).
I assume that the vaccine efforts are partially a matter of "Well, I'm an immunologist not a field health/education worker, so what am I best suited to do?", partially a matter of protecting people during the time they aren't in bed; and perhaps also the hope of eventually making a sufficient portion of humans resistant and crashing the population of malaria causing protozoa entirely. P. knowlesi unfortunately has an animal reservoir(some non-human primates); but some of the other common plasmodia don't, so if you could increase resistance enough you might be able to hit the point of substantial additional gains 'for free' as the number of infected mosquitos drops and the population crashes.
Aside from immediate considerations, working on a malaria vaccine probably gets some additional interest because of its greater value(both humanitarian and commercial) if climate change should cause the current range of the disease(mostly ghastly tropical pestholes filled with people who can't afford expensive drugs) into wealthier areas of the world. There may also be a basic-research interest: unlike most pathogens, plasmodia are eukaryotic; so I'm sure that the relevant specialists find all sorts of fascinating differences between the biology of the pathogen/host interaction in malaria vs. that in infections by bacteria or viruses. You aren't going to commercialize a drug on basic research alone; but if you want research to happen it certainly doesn't hurt to be novel and interesting.
My pleasure. I'm always glad to see a discussion take a turn for the better rather than just sliding off the rails. Unfortunately, it seems as though the value of known-inaccurate simplified models is often enough poorly understood that some people treat them as "Haha, your model doesn't happen in real life, therefore Economics Refuted!" and others treat them as though their results can actually be trusted when talking about the real-world situations that they are intended to help analyze.
In this case, perfect information is obviously not happening(if nothing else, you'd be crowned God-Emperor of HR for all eternity if you actually found a way of objectively ranking an employee's expertise with enough precision to justify the difference between their salary and the category average down to the last dollar, or even the nearest $10k in a lot of cases); but it does seem like a pretty decent example of how a situation goes from being substantially not-'free-market'(information is both imperfect and asymmetric, with Google knowing all the salaries and each employee knowing only their salary) to one that is markedly closer to 'free market'(Google knows all the salaries, each employee knows at least a fair number of salaries; and is negotiating from a position of much better price information).
I admit that my initial post was pretty snippy; I get annoyed at the cries of "SOCIALISM!!!", especially now that the Cold War is over, all the 'communist' states have either collapsed or turned into crony-capitalist states of various flavors; and the closest thing you can find to 'socialism' is capitalist countries with comparatively cushy social safety nets; and whoever the AC was pushed my buttons.
That specific annoyance aside, though, I'm actually rather fascinated by how useful(across a wide variety of disciplines) models that we know are false can be, despite their falsehood. They are wrong; but by being wrong in well defined ways that are amenable to (relatively) simple analysis they can be such a good jumping off point for examining the real world and figuring out how it must be different in order to produce the results you see.
I find the 'esport' label irksome, mostly because I've never understood the big deal about watching sports; but if the defining characteristic of 'sport-ness' were physical rigor; coal mining and working in a sweatshop would be major athletic events and they'd force golfers to walk the entire course and carry their own bags.
Gaming is obviously pretty low intensity for most muscles, though the rigor of the mental drill is considerable; and the amount of carpal tunnel and similar injuries are actually alarmingly high.
The alternative would obviously be untenable(either forcing athletes to do without medical care 'for their own protection' or just banning every sickie who needs a drug that might be performance enhancing); but a therapeutic use exemption for psychostimulants is going to make the rule more or less a joke(not that I have a problem with that, personally). Getting a diagnosis for which one of the stimulants is the usual treatment is pretty trivial; and they are cheap, have lots of safety data available, and generally don't raise any red flags among doctors. It depends on where you are, of course; but they might actually be among the few drugs that are easier to get legally than illegally.
Scrabble is only a game about words at fairly low levels of play. If you have two otherwise unprepared people stuck in a room with nothing but scrabble for amusement, yes, the one with the better vocabulary likely has better options. Outside of the amateurs, though, memorization of the approved dictionary(starting with words chosen for good point values, the ability to dispose of letters that are usually tricky to get rid of, and other helpful features; but ideally progressing to all of them) supplants knowledge of the language and the remaining challenge is board control and optimizing the conversion of tiles into points over the course of the game.
There would certainly be additional prep time, even for the unusual characters who are really good at this; but the skills that the game demands for high level play should be transferable to any language(or even a nonsense dictionary) that works reasonably well with representation by a relatively small alphabet.
"Perfect information" is the idealized model. As with any idealized model, economics or otherwise(trajectories with respect to a single point mass in absence of friction, ideal gasses, etc.) it sacrifices real-world attainability for substantial convenience in in building and analyzing the model.
Once you have the idealized model, you have something with which to compare real world outcomes and a basis for studying how and why they deviate from the idealized version. Is it barriers to entry? Asymmetric information? Bounded rationality on the part of some or all actors? That's where the economists who grovel through data come in.
In this case, the point isn't that perfect information is expected of real-world economic happenings; perfect information is not possible in practice. However, 'sharing salary data' is an absolutely textbook example of something that would move the situation from 'very imperfect data' to 'closer to perfect data'. It also likely reduces the asymmetry of information(HQ already knew all the salaries, and possibly some at competing companies as well; these employees now have better information and information that is closer to parity with the actor they are negotiating with).
The purpose of idealized models is not to a deliberately unanswerable demand "If it isn't a 'Free Market' the market isn't free!"; but to act as a simplified analytical tool that allows you to focus more clearly on the aspects of the real market that are most or least like their ideal counterparts and tease out how the non-ideal behavior changes the outcomes.
You do realize that 'perfect information' is one of the defining characteristics of the idealized model of 'free market' behavior? You don't have to like it; but calling anything vaguely related to money that displeases you 'socialistic' is dumb beyond words.
Wasn't Google going to index all the world's information, at one point? It appears that they might be making an exception.
We, the fine folks at the interactive advertising bureau, are delighted by the notion of a 'consumer preference client'. Indeed, we are so strongly committed to it that we recommend that it be incorporated at the hardware level, in order to provide additional trust in the 'Trusted Platform' that forms the foundation of the secure online marketplaces of tomorrow. With a suitably immutable GUID baked into every piece of hardware, we can finally ensure that each and every consumer receives exactly the privacy settings and offers most relevant to them!
How could no one have foreseen the potential abuse and pitfalls of a system like this? Without even reading any further than "Giving Doctors Grades..." I immediately conjured images of a bunch of doctors huddled around each other saying, "I don't want that one." "Well I don't want that one either. My feedback is back at 85% and I can't risk another death screwing me over."
Given the enthusiasm for what is more or less exactly the same plan applied to teachers, it's hard to be too surprised that somebody would think that this is a good idea; though it doesn't shed any light on why they would have ignored the obvious pitfalls.
I can only assume that it's yet another instance of the "We want to measure stuff; but what we really want to measure is hard, so we'll settle for something that is easy to measure but not actually helpful" problem that bedevils quantification efforts of all kinds.
I think that he was hoping for a good, old-school, insertion of knife into abdomen and pulling all the way across to demonstrate contrition through horribly painful death; in the fine Japanese tradition.
I would actually be interested to know what the logic is here: the hacker clearly doesn't like AM, or they wouldn't be spoiling their rumored-IPO quite this enthusiastically, they also don't like the users they are threatening to expose; but they also appear to be really bent out of shape about AM's allegedly-dishonest-and-exploitative 'pay to purge the embarrassing traces' feature.
Anger about that feature would seem to be something more likely in some portion of the users, or among people who identify with the interests of the users; but this interested party displays only contempt for them; rather than viewing AM's attempt to squeeze them as an amusing and justified punishment.
We obviously have no particular reason to trust their statement; but we do have to expect that they have a reason worth the legal exposure for doing this(especially since the dataset they are talking about would probably be worth a decent sum for sale to others looking for really juicy spearphishing targets ) rather than not attempting the hack at all or hacking but then staying quiet about it. My guess would be that it is more about attacking the site operator than about the users specifically; it is pretty common for at least a person or two to end up suitably embittered during the course of business.
Unforeseen consequences.
Probably not directly. To the degree that Microsoft has any specific plan to limit game adoption on non-Windows platforms it is called 'DirectX'. It is the first-class set of APIs on Windows and any games developed for it, or drivers developed to support it, are obviously resources dedicated to gaming being better on Windows and either unavailable or produced at additional cost for OpenGL elsewhere.
Once you get into how AMD's OpenGL driver does(or doesn't) apply application specific optimizations for different OpenGL games, though, MS doesn't have nearly as much to gain from any specific meddling. The general success of DirectX and Windows gaming is presumably the reason why AMD cares relatively little(along with the fact that people looking to use proprietary drivers on Linux usually go Nvidia, while AMD is regarded as very much the second choice unless you are looking for the vendor more cooperative with FOSS driver development).
Give him some credit: you don't get -1 unless your post was at least locally offensive, if not globally.
I doubt that there is an actual price list. At least on the Windows side, releasing driver optimizations(sometimes including overt cheating if the popular benchmarking programs of the day are detected, though that seems to be rarer in recent years) to support popular programs and games is something that both Nvidia and ATI do reasonably routinely in order to improve their products' perceived competitiveness. You don't get a profile unless you have moved enough units to make it worth the effort, or there is a particularly embarrassing performance discrepancy; but the GPU vendors are sufficiently concerned about appearances that just stalling you is a poor choice. If anything, it appears to be that the GPU guys are the ones pushing for greater prominence in game, hence the obnoxious little videos that roll when you start some games and the Nvidia's enthusiasm for PhysX and related middleware being used.
On Linux, it looks like AMD doesn't care very much(I assume that the Linux gaming numbers don't justify much engineering labor for specific titles; but if you have a 'Half-Life 2' optimization it wouldn't be terribly laborious to have the intern look up other Source-based games and add those to the list), but the same reasons for AMD's apathy would also discourage any one vendor from paying very much to get on 'the list'.
That said, I have no reason to doubt that a vendor would outright refuse you if you demanded to pay your way onto the list(especially if you provided both cash and access to any information they needed about your application or from your developers); but unless it is actively all kinds of horribly broken without special optimizations, in which case you have a problem, simply selling enough copies makes supporting your application part of the job of selling GPUs. Since Intel is just waiting to scoop up everyone who doesn't care very much about graphics cards, neither vendor can really afford to withold optimizations and take a hammering in reputation just to squeeze a little protection money out of a publisher.
Another handy advantage to moving away from 10base5 and 10base2! Much trickier to open and tap a bunch of twisted pairs; and the system will not be happy with having two devices on the same cable unless your spybot is quite unobtrusive in any sniffing and spoofing it does. Vampire tapping coax would be much easier to automate with reasonable simplicity and reliability and the system is intended to work with multiple devices on the same wire.
Would you care to go into more detail about what brilliant innovations we are missing out on? The market in remote access Trojans and spyware remains open, the Boing subsidiary with the drone(who were the ones proposing a drone-mounted malicious wifi device) remains open, suppliers of ever smaller computers and wifi systems remain open. 'Hacking Team' may have been reasonably competent crackers, if not so hot on defense; but their(hopeful) demise would be a "and nothing of value was lost and we were all slightly more secure for it" situation if I've ever seen one.
That's actually a bit surprising. Microsoft sure went to a lot of trouble with their precious 'protected media path', so if their own application is merrily recording a DRMed stream(as I believe Netflix is on all supported platforms); either they've screwed up or Netflix couldn't be bothered to use the feature. I imagine that re-compressed copies of streams aren't terribly high priority; but I would have imagined that they'd be contractually obligated to at least pretend to care.
Unfortunately, you still need to use the correct incantations(and sometimes some annoyingly expensive and specific material components) if you want the magic to work properly.
I assume that the problem is with noticing the fire. A small Li-ion battery can self-ignite and burn fairly enthusiastically; but isn't too dangerous if it is prevented from setting anything else on fire. The smoke is noxious and any one directly exposed the the flame will be burned; but it just isn't a very big fire. If the battery is hiding down in the cargo hold in somebody's suitcase, it has a better chance of recruiting all the nearby luggage and getting a proper fire started; at which point suppression becomes more difficult and release of enough energy to actually damage the aircraft becomes likely.
I'd be interested to know what the current standard for fire detection in the cargo area is; and how difficult and costly it would be to achieve better early warning.
Your options really depend on what sort of 'high performance' you have in mind. When it comes to performance per core, Xeons typically crush Opterons; but the pricing reflects that, especially if you need the 4-8 socket support and RAS features. If what you need is large amounts of RAM with the lowest possible spending on the system around it, Opterons have tepid performance per core; but are likely to be the cheapest option that still supports ECC, more than one socket, buffered DIMMs, and any other niceties you wouldn't get from just desktop. If your application is one that can be made to fit, GPUs are enormously powerful for the range of things that they are capable of doing well.
They also depend on how big you need your system to be and how tightly coupled it has to be. If your intended application handles its own network-level parallelization and doesn't depend on very low latency, blessed are you. Price per core skyrockets if you go above 2 sockets, and GbE is effectively free(at least in the sense that you pretty much can't buy a system or motherboard that doesn't come with at least 2 NICs by default, often more) and relatively cheap to switch. If you need lower latency, this will hurt more and you are looking at myrinet or infiniband. If your application needs a cluster that presents a single system image; especially one that also has genuinely low latency, you probably need to fortify your checkbook and consult an expert. You can get systems with more than 8 sockets and the appropriate custom interconnect; but you won't like paying for one.
Unless you value your time at surprisingly low rates; you probably won't want to build your own systems from parts; but depending on how tightly coupled you need, this may be something that you need to purchase as a system or something you construct from multiple computers you purchase.
Can you use either hardware you have or AWS(or one of their similar competitors) to better characterize what your application actually needs?
But it (used to be) AMD who has zee German fabs!
Given that LEDs already lean pretty enthusiastically on those elements, without too much comment; and some of the stuff they bake into the silicon is pretty dreadful I imagine that they'll diplomatically ignore the issue and hope really hard that the Indium content is high enough to make the semiconductors that aren't super small and embedded in epoxy economically recyclable.
Getting changes made when there are at least partial alternatives and the material is distributed in substantial quantities throughout the entire product is much less of an uphill fight than getting changes made to the relatively small chunks of the product where much of the value and the profit are; and the producers say they've exhausted the options. They would probably be loathe to push it unless their position were markedly stronger than it appears.
I'm sure that that's true on total sale, less sure that it's true on profit(you obviously drop out of the 'buy license from ARM, tell TMSC to make it, attempt to sell it' business if you can't at least keep the lights on; but I suspect that Allwinner and Mediatek aren't exactly commanding Xeon margins); but either way Intel has less pressure on the high end. If you do need a relatively powerful processor you options are pretty much 'Intel', a substantial chunk of dead space, 'maybe AMD', then another chunk of dead space, and then one of the 64-bit ARM people.
That isn't really so scary. On the low end, though, they have plenty of alternatives; some quite possibly overtly superior and definitely plenty that are cheaper. That is more scary.