Kind of surprising. More modern computers start failing from electromigration issues in the CPU (and probably the chipsets and so forth as well) after only about 10 years of use. Or at least, so I've been told by someone who ought to know.
I think you're right. Larrabee or some derivative may be good at graphics. It will almost certainly blow away current GPGPU solutions as workstation accelerators.
And there are smokers who lived to be 100 too. If this technology caused 1% or.1% of people to get cancer 20 years later, would you use it? What if it only affected fertility and birth defects? The acceptable safety ranges of technologies that we accept as a society aren't particularly amenable to anecdotal evidence.
Maybe the "cell-phones cause cancer" guys are reactionary crackpots, but maybe they have a point and they only scream because that's the only way to be heard. Myself, I'll wait for more scientific evidence to judge.
Don't you mean it's always the employee's responsibility? Yes, it is. That's the way the world is structured: if you don't like the way the world treats you, you can't expect the it to change to suit you. But that doesn't mean the employee has necessarily done anything wrong. It's not actually guaranteed that the opportunity you want is available: it is only a truism that good opportunities exist. You just take it for granted for the sake of your own sanity.
All in all, this is why contemporary and Christian morality asks that we look out for others a bit and not just ourselves. But it is still not technically someone's responsibility to look out for us.
I thought Mike Huckabee was polling around 1% before he showed up on the Colbert Show. I don't know about contributions, but it wasn't too long after that he became a pretty influential second-string candidate. Correct me if I'm wrong here.
In other words, there is a shortage of people so killed and driven that they provide their own business case as well as value, and yet prefer to work under whatever salary and employer wishes to pay them. Aren't those the guys that typically will rather work for themselves?
The US is not a monolithic entity. That portion of U.S. policy is in fact owned by multi-national corporations. They're quite happy to pull the rug out from under workers (aka labor costs) in *any* country. And it's not just American-owned or American-headquartered companies. That's the down side of the fun we call capitalism. Of course, as long as workers have a local market monopoly on *something*, they can hold their own in the fight.
That's why Democracy is nominally important. If people get screwed enough to actually participate *enough*, then people have a non-violent recourse against that kind of treatment. Which ought to include protectionism... and a dose of pain for the cartels like RIAA that bought the DMCA in the first place.
You probably wouldn't like my thoughts on software patents then, either. I don't think software patents older than maybe 5-10 years really serve society in any way.
Since patents are fundamentally a global restriction (backed by force) on "everyone's" behavior, there ought to be a high bar to justify their existence and application. The software world changes so fast, and the line between obvious and non-obvious innovation is so blurred, that software patents are a really heavy hammer.
I don't agree with RMS's view of things, but I do think FOSS has a powerful role in the marketplace to keep proprietary interests from charging for longer periods of time for minor innovations of a dozen years ago.
Copyrights are similar, but are weaker and provide a more obvious good. Still, I think allowing 20-yr old abandonware into public domain might be a reasonable step, although the Win 3.11 case wouldn't quite fall into that bucket.
Absolutely! But this beta software is part of the Ubuntu standard distribution. My home desktop is not exactly mission critical, but it's important to my experience with linux. And knowing about the problem, I still choose to use it. It's good stuff. But my comment stands.
The only reason I reboot is to clear wine state or if the desktop is too sunk to let me CTRL-ALT_BACKSPACE. I find once wine dies on one of my favorite apps, restarting wine doesn't always let the app run.
The linux kernel... rock solid. But with compiz, gnome, and wine,... sometimes I end up rebooting because I'm not familiar enough with what to kill and restart.
"The change takes time" argument is a valid one, but it's a centrist argument also. For a "center" to exist, you need a more radical vision. For example, Linux would not exist without the more radical vision of GNU. By that token, we must recognize that big ships like the U.S. and China change slowly. But they have a long way to go, and the center won't move without a radical transforming vision.
China's nationalism and capacity for self-deceit ought to be downright scary for anyone with a sense of history, especially combined with the sense of invincibility that comes from such rapidly growing power. Someone needs to see that, hopefully in China, and dampen it a bit. China is deliberately making a lot of clever and subtle moves to undercut the west (especially America) and establish global power: western countries should be careful not to provoke confrontation, but we can't just roll over for it either.
On the flipside, America's arrogance and interventionalism needs to be clipped: we in America have to realize we have neither the moral authority nor even the capability to police the world the way have in the past. And yet, we strangely don't stand up for our own people's interest in the global world: America's take on globalization is still startlingly naive.
I hope Obama is that guy in the U.S. to change some of that. We'll see.
Our friend from Nvidia mocks Intel for thinking it GPUs are easy even as he makes statements like "making a core isn't all that hard". And it's true: making a core isn't all that hard. Unless it has to perform well, be proven compatible to all kinds of bizarre existing legacy, be cheap, be reliable, be on time... and in short satisfy all the other requirements on a CPU design team for a mainstream product. This guy is completely talking out of his ass on that front.
That said, it will be interesting to see whether Larrabee proves to be an innovative new take on how to do graphics, or in fact a CPU architect's warped view of how to solve a graphics problem. Either is quite possible. I think the conjecture that a Larrabee die is going to cost $1000 to make is probably ridiculous... and even if he's talking about a sale price it's pushing it.
I think the observation that CPU and GPU will probably remain pretty separate except in low cost implementations is probably right. At the flop level we're talking about, memory latency and bandwidth as well as sheer transistor count makes it difficult to believe that partitioning the CPU and GPU logic in the same die makes sense. The dominant interactions are the memory-compute bandwidth, not the CPU-GPU logic bandwidth, so the current partitioning makes sense as far as I can see.
The official terminology for lowering prices to around or below cost in order to drive a competitor out of business is "predatory pricing". And yes, its a monopolistic practice. Sony is as dangerous as Microsoft in this regard.
I totally agree that this is disgusting behavior, and Intel must accept responsibility. However, I think it only fair to point out that the message from above in Intel is to play fair in situations like these. I doubt senior management would authorize something like this: it's probably a case of some overzealous salesperson acting out of line. As an Intel employee, they force us to take tons of classes specifically aimed to prevent this kind of stuff, but it happens anyway. I can't believe it was sanctioned, because it's just too darn stupid. I don't have a lot of visibility on this, though, since I'm in design, not sales.
As far as Classmate vs. XO, I think that's a smokescreen for the real issue: how to push PC-like objects with the right features into low price points and high quality so that they are affordable enough to use for mass education projects. Everybody knows Intel can't design PCs, the only technical role they can play is getting cheap, good-performance, low-power SOCs out the door. As a publicly owned company, they can only do this if they have a financial story. And getting locked into *exclusively* supporting an initiative that only uses the competition's processors? Come on! I'm as big of an intelligent critic of capitalism as anyone I personally know, but expecting Intel to back OLPC to the hilt at any cost is just the dumbest thing I've ever heard, and it amazes me how many otherwise intelligent people don't understand that. Let Intel compete to get the right processors in the right solution; that's all Intel knows how to do anyway (and as long as AMD keeps them honest, that's what they do).
And please, don't give me a rant on how everything from the west is too darned expensive anyway, because most of that boils down to issues of ridiculously manipulated currency markets.
Slashdot Mirror
Via might not be for sale. Taiwan might not be thrilled if it were.
Kind of surprising. More modern computers start failing from electromigration issues in the CPU (and probably the chipsets and so forth as well) after only about 10 years of use. Or at least, so I've been told by someone who ought to know.
I think you're right. Larrabee or some derivative may be good at graphics. It will almost certainly blow away current GPGPU solutions as workstation accelerators.
And there are smokers who lived to be 100 too. If this technology caused 1% or .1% of people to get cancer 20 years later, would you use it? What if it only affected fertility and birth defects? The acceptable safety ranges of technologies that we accept as a society aren't particularly amenable to anecdotal evidence.
Maybe the "cell-phones cause cancer" guys are reactionary crackpots, but maybe they have a point and they only scream because that's the only way to be heard. Myself, I'll wait for more scientific evidence to judge.
Don't you mean it's always the employee's responsibility? Yes, it is. That's the way the world is structured: if you don't like the way the world treats you, you can't expect the it to change to suit you. But that doesn't mean the employee has necessarily done anything wrong. It's not actually guaranteed that the opportunity you want is available: it is only a truism that good opportunities exist. You just take it for granted for the sake of your own sanity.
All in all, this is why contemporary and Christian morality asks that we look out for others a bit and not just ourselves. But it is still not technically someone's responsibility to look out for us.
I thought Mike Huckabee was polling around 1% before he showed up on the Colbert Show. I don't know about contributions, but it wasn't too long after that he became a pretty influential second-string candidate. Correct me if I'm wrong here.
In other words, there is a shortage of people so killed and driven that they provide their own business case as well as value, and yet prefer to work under whatever salary and employer wishes to pay them. Aren't those the guys that typically will rather work for themselves?
The US is not a monolithic entity. That portion of U.S. policy is in fact owned by multi-national corporations. They're quite happy to pull the rug out from under workers (aka labor costs) in *any* country. And it's not just American-owned or American-headquartered companies. That's the down side of the fun we call capitalism. Of course, as long as workers have a local market monopoly on *something*, they can hold their own in the fight.
That's why Democracy is nominally important. If people get screwed enough to actually participate *enough*, then people have a non-violent recourse against that kind of treatment. Which ought to include protectionism... and a dose of pain for the cartels like RIAA that bought the DMCA in the first place.
You probably wouldn't like my thoughts on software patents then, either. I don't think software patents older than maybe 5-10 years really serve society in any way.
Since patents are fundamentally a global restriction (backed by force) on "everyone's" behavior, there ought to be a high bar to justify their existence and application. The software world changes so fast, and the line between obvious and non-obvious innovation is so blurred, that software patents are a really heavy hammer.
I don't agree with RMS's view of things, but I do think FOSS has a powerful role in the marketplace to keep proprietary interests from charging for longer periods of time for minor innovations of a dozen years ago.
Copyrights are similar, but are weaker and provide a more obvious good. Still, I think allowing 20-yr old abandonware into public domain might be a reasonable step, although the Win 3.11 case wouldn't quite fall into that bucket.
I don't think Atom was done when they started Larrabee. Just a thought.
Absolutely! But this beta software is part of the Ubuntu standard distribution. My home desktop is not exactly mission critical, but it's important to my experience with linux. And knowing about the problem, I still choose to use it. It's good stuff. But my comment stands.
The only reason I reboot is to clear wine state or if the desktop is too sunk to let me CTRL-ALT_BACKSPACE. I find once wine dies on one of my favorite apps, restarting wine doesn't always let the app run.
The linux kernel... rock solid. But with compiz, gnome, and wine, ... sometimes I end up rebooting because I'm not familiar enough with what to kill and restart.
"The change takes time" argument is a valid one, but it's a centrist argument also. For a "center" to exist, you need a more radical vision. For example, Linux would not exist without the more radical vision of GNU. By that token, we must recognize that big ships like the U.S. and China change slowly. But they have a long way to go, and the center won't move without a radical transforming vision.
China's nationalism and capacity for self-deceit ought to be downright scary for anyone with a sense of history, especially combined with the sense of invincibility that comes from such rapidly growing power. Someone needs to see that, hopefully in China, and dampen it a bit. China is deliberately making a lot of clever and subtle moves to undercut the west (especially America) and establish global power: western countries should be careful not to provoke confrontation, but we can't just roll over for it either.
On the flipside, America's arrogance and interventionalism needs to be clipped: we in America have to realize we have neither the moral authority nor even the capability to police the world the way have in the past. And yet, we strangely don't stand up for our own people's interest in the global world: America's take on globalization is still startlingly naive.
I hope Obama is that guy in the U.S. to change some of that. We'll see.
Our friend from Nvidia mocks Intel for thinking it GPUs are easy even as he makes statements like "making a core isn't all that hard". And it's true: making a core isn't all that hard. Unless it has to perform well, be proven compatible to all kinds of bizarre existing legacy, be cheap, be reliable, be on time... and in short satisfy all the other requirements on a CPU design team for a mainstream product. This guy is completely talking out of his ass on that front. That said, it will be interesting to see whether Larrabee proves to be an innovative new take on how to do graphics, or in fact a CPU architect's warped view of how to solve a graphics problem. Either is quite possible. I think the conjecture that a Larrabee die is going to cost $1000 to make is probably ridiculous... and even if he's talking about a sale price it's pushing it. I think the observation that CPU and GPU will probably remain pretty separate except in low cost implementations is probably right. At the flop level we're talking about, memory latency and bandwidth as well as sheer transistor count makes it difficult to believe that partitioning the CPU and GPU logic in the same die makes sense. The dominant interactions are the memory-compute bandwidth, not the CPU-GPU logic bandwidth, so the current partitioning makes sense as far as I can see.
The official terminology for lowering prices to around or below cost in order to drive a competitor out of business is "predatory pricing". And yes, its a monopolistic practice. Sony is as dangerous as Microsoft in this regard.
Skewed in favor of American-based MNCs, perhaps, but not in favor of America, per se., I think.
As far as Classmate vs. XO, I think that's a smokescreen for the real issue: how to push PC-like objects with the right features into low price points and high quality so that they are affordable enough to use for mass education projects. Everybody knows Intel can't design PCs, the only technical role they can play is getting cheap, good-performance, low-power SOCs out the door. As a publicly owned company, they can only do this if they have a financial story. And getting locked into *exclusively* supporting an initiative that only uses the competition's processors? Come on! I'm as big of an intelligent critic of capitalism as anyone I personally know, but expecting Intel to back OLPC to the hilt at any cost is just the dumbest thing I've ever heard, and it amazes me how many otherwise intelligent people don't understand that. Let Intel compete to get the right processors in the right solution; that's all Intel knows how to do anyway (and as long as AMD keeps them honest, that's what they do).
And please, don't give me a rant on how everything from the west is too darned expensive anyway, because most of that boils down to issues of ridiculously manipulated currency markets.