I'll be interested to see what they're going to do about making it easier to program...
Assuming they did anything interesting on the toolchain side.
Contrary to the summary and your remark, I'm not sure it's Tile64's problem to bring parallel programming to the masses. First, because many-core chips are already useful (and present no special difficulties) for servers that handle many simultaneous connections - in other words, reducing the space and electricity requirements of server farms. That's a significant market.
Second, parallelism is a far broader problem than this tiny company's single product; it's now a problem for Intel and AMD, too - in other words, for everybody. Any effective solution is highly unlikely to be specific to this particular chip.
Sure, it would be great if these guys (or anybody else) made some breakthrough in parallel programming, but that doesn't appear to be the problem they've tackled. You say shoving together a bunch of cores is boring, but to me replacing a cluster with a single $500 chip would be fantastic.
I am interested how this will stack up to Sun's Niagara chip. 600 MHz is pretty slow nowadays.
I agree with your point, though not because China has the nuclear bomb in particular. Their strength is their economy. In the long run this is all that matters. With money you can develop or buy arms and other technology, buy limited natural resources (oil), and a louder voice at the UN, etc. Much has been made of specific transfers of sensitive technology to China. But the bigger picture is that all this is inevitable so long as their economic growth is outstripping ours; they will come closer to parity with us in all other respects. And we are so helpful in assisting them with our massive trade deficit, each year bringing our nations closer to economic parity. Maybe it will turn out well and the hypothesis that capitalism breeds democracy will be proven correct; on the other hand, maybe not. It seems like an awfully big gamble to me, so I'm surprised there isn't more concern and debate about it.
OK, who modded this +5 insightful? Come on people, he didn't even site any figures. $1.2 trillion for 113 million households is a little over $10K / household, and yes that will buy you something. Even if you want to go with individuals instead of households (which is totally unreasonable since homes, not individuals, are wired), it's still $4000 (not $1000) per capita. And $1.2 trillion is a conservative prediction of the costs of the war; I don't think it includes e.g. the high gas prices resulting from all the chaos in the middle east.
The best analogy for wired Internet (including fiber) is sewage, water, and roads. We are not going to have a dozen different fiber networks to each home so we can let the market decide which is best. That is why the "free market" for bandwidth is a failure in the US. Compared to global norms for price and service, we are losing the race and falling further behind all the time. This is not an ideology-based prediction, but actual fact about the current state of affairs.
Now I guess you can say it doesn't matter because so much Internet usage is recreational. Perhaps you could sit along any Interstate and complain about all the people towing jet-skis and RVs since after all the Interstate highway system was touted during the red-scare '50s for the purpose of national defense, not wasting time. Yet roads (including "unnecessary" travel) are vital to the economy and our way of life. Nothing is better correlated with economic growth than the ease and speed of communication, which for the forseeable future means the Internet.
I see one possible trump card that might allow an efficient market for bandwidth by private industry, and that is wireless. Maybe eventually it will be good enough, and the spectrum regulated properly so as to create a competitive marketplace. Maybe one day, but not yet.
The requirement for ID to enter parks seems so asinine that I have a very hard time imagining they'll actually go forward with it. Since the last Congressional elections did bring a change in which party holds the majority, isn't there any chance Real ID will be shot down?
It's nothing more than America's version of "show me your papers."
There's nothing in the second law of thermodynamics that says some reactor in Alabama couldn't make better use of the heat it generates. Just like different cars getting different mpg.
I've been using linux on the desktop for 10 years, but I'm not sure it's the right choice for somebody who wants things to work all the time, or even be easily fixable. I still loathe jobs like making OpenGL acceleration work, and printing, and power management. Sometimes they work out of the box, sometimes you spend hours and never fix it (except power management of course - it never, ever works out of the box). And package management always seems to deteriorate over time and eventually crumble to pieces, leaving unresolvable dependencies or a corrupted database (and I'm talking about portage, rpm, and apt!) That said, I'm addicted to the control. I have my config files how I want them, I have weird custom setups, and to me Linux is still worth it, so far.
I was really surprised when the guy didn't end the sentence "I might just do the unthinkable: I might move to..." with "Macintosh." I think their slogan of "it just works" has a lot of truth to it, so long as you stay in the walled garden.
Oh brother. It's past time for a Godwin's Law-style rule about thermodynamics on slashdot. Every time somebody propses conserving wasted energy, somebody else retorts that perpetual machines are impossible or something to that effect.
This seems a lot simpler and fairer than traffic shaping by protocol.
There's no need for fixed transfer limits. And shaping by protocol is the problem, not the solution, since the content (including the protocol) is really none of the carrier's business.
Timesharing CPU schedulers have been solving this problem better for, what, 45 years now? You don't look at the filename of the executable somebody is running to see if you will schedule it. You don't suddenly kill their process if they exceed 60 seconds of CPU time. Instead, you simply de-prioritize "cpu hogs" - or in this case, bandwidth hogs. If you are a bandwidth hog, your "prime time" bandwidth should fall very low - lower than others who *only* use bandwidth at that time - but at 3am it should ramp up again, since you're only "competing" with other bandwidth hogs.
Erasing traumatic events is not helpful. Learning to accept and cope with a past traumatic event is.
I wouldn't criticize the research on the basis of potential applications just yet. Maybe it will lead to a breakthrough in AI, or drugs that vastly improve memory, or combat Alzheimer's. Unfortunately it could instead lead to a "morning-after" date rape drug or a way to enforce anti-compete clauses in employment contracts. It's just too early to say.
Now I'll have to keep thinking about old memories or I'll lose all the good ones.;)
I'm sure recalling old memories periodically does refresh them (though I'll bet it also changes them over time, like an Nth generation photocopy). This is different though, they're describing a subconscious chemical process.
It's self-regulating. The speculators go after high-risk stuff. If they go too high-risk, they lose all of their money and cease to be speculators.
That all depends on the system. What about all the sleazy loan agents who sold ballooning loans to people who could barely pay the teaser rate? Those agents got their comission and they are long gone. Now it's up to the Fed to pump billions into the economy to compensate for their actions. The rest of us will repay that handout through inflation.
One of the reasons that they are so successful here is that there is a relatively low stigma associated with failure in the US.
I guess you mean "business failure," because the consequences for personal bankruptcy in the US are pretty severe - moreso after the 2005 bankruptcy law changes.
he demonizes guys for figuring out how to make money in the current tax environment. It's not their fault that the government has weird tax rules, and they are not bad guys for working within those rules. They aren't even bending them or subverting their intent - they are just playing the game.
They are not just playing the game. They are also writing the rulebook. Big difference.
American students are non-science jobs because that's what our economy rewards. Dentists don't have to contend with global competition. Apparently the envisioned future is that the Chinese and Mexicans will do all the work while we sit back and "manage" them, e.g. continue glutting ourselves by skimming all the profits off their work. Personally I think we're headed for trouble.
There's a computer store on my way to work with a sign out front that says, "Hate Vista? We have: XP Desktops! XP Laptops! Hurry, supplies limited!" I'm not making this up.
Read his remark: "it would be an incredible experiment to shut down the whole internet for five years and see what sort of art is produced over that span." Do you think that's a serious call to shut down the Internet? I don't. I think it's an off-the-cuff call for musicians to interact more with each other and audiences. Do I personally agree that the Internet will turn all musicians into Moby? Nah. Then again, Elton John might just have insight into the musical world that you and I do not.
From the blurb, if you're not a "server" I guess you don't need an IPv6 address... then again what does that even mean? I think it would be a critical mistake to start making any actual distinction between client servers on the Internet. To me the Internet would be fundamentally different if I could no longer log in remotely, receive VOIP phonecalls, and host my family photos and a few other files - partly because these are important applications for me, and partly because there simply should not be different classes of nodes.
Slashdot Mirror
Second, parallelism is a far broader problem than this tiny company's single product; it's now a problem for Intel and AMD, too - in other words, for everybody. Any effective solution is highly unlikely to be specific to this particular chip.
Sure, it would be great if these guys (or anybody else) made some breakthrough in parallel programming, but that doesn't appear to be the problem they've tackled. You say shoving together a bunch of cores is boring, but to me replacing a cluster with a single $500 chip would be fantastic.
I am interested how this will stack up to Sun's Niagara chip. 600 MHz is pretty slow nowadays.
I agree with your point, though not because China has the nuclear bomb in particular. Their strength is their economy. In the long run this is all that matters. With money you can develop or buy arms and other technology, buy limited natural resources (oil), and a louder voice at the UN, etc. Much has been made of specific transfers of sensitive technology to China. But the bigger picture is that all this is inevitable so long as their economic growth is outstripping ours; they will come closer to parity with us in all other respects. And we are so helpful in assisting them with our massive trade deficit, each year bringing our nations closer to economic parity. Maybe it will turn out well and the hypothesis that capitalism breeds democracy will be proven correct; on the other hand, maybe not. It seems like an awfully big gamble to me, so I'm surprised there isn't more concern and debate about it.
Now I guess you can say it doesn't matter because so much Internet usage is recreational. Perhaps you could sit along any Interstate and complain about all the people towing jet-skis and RVs since after all the Interstate highway system was touted during the red-scare '50s for the purpose of national defense, not wasting time. Yet roads (including "unnecessary" travel) are vital to the economy and our way of life. Nothing is better correlated with economic growth than the ease and speed of communication, which for the forseeable future means the Internet.
I see one possible trump card that might allow an efficient market for bandwidth by private industry, and that is wireless. Maybe eventually it will be good enough, and the spectrum regulated properly so as to create a competitive marketplace. Maybe one day, but not yet.
It's nothing more than America's version of "show me your papers."
Maybe they could use the warm water to grow slime mold for biodiesel in cooling tanks by the reactor.
There's nothing in the second law of thermodynamics that says some reactor in Alabama couldn't make better use of the heat it generates. Just like different cars getting different mpg.
I was really surprised when the guy didn't end the sentence "I might just do the unthinkable: I might move to..." with "Macintosh." I think their slogan of "it just works" has a lot of truth to it, so long as you stay in the walled garden.
Timesharing CPU schedulers have been solving this problem better for, what, 45 years now? You don't look at the filename of the executable somebody is running to see if you will schedule it. You don't suddenly kill their process if they exceed 60 seconds of CPU time. Instead, you simply de-prioritize "cpu hogs" - or in this case, bandwidth hogs. If you are a bandwidth hog, your "prime time" bandwidth should fall very low - lower than others who *only* use bandwidth at that time - but at 3am it should ramp up again, since you're only "competing" with other bandwidth hogs.
So? What significance do you think you see in a simple grammatical error?
American students are non-science jobs because that's what our economy rewards. Dentists don't have to contend with global competition. Apparently the envisioned future is that the Chinese and Mexicans will do all the work while we sit back and "manage" them, e.g. continue glutting ourselves by skimming all the profits off their work. Personally I think we're headed for trouble.
Maybe awesome for 1976 is awful for 2007.
There's a computer store on my way to work with a sign out front that says, "Hate Vista? We have: XP Desktops! XP Laptops! Hurry, supplies limited!" I'm not making this up.
Actually this is a way for chips to give off *even more* heat. In fact it might be a good technology for hair dryers.
Don't show up at all and you can avoid the anxiety of face-to-face conversations and the resulting workload. Yay!
But open standards for state documents is not a campaign issue. I don't see how my vote has any influence over it.
You're only a counter example if you don't suck :)
Read his remark: "it would be an incredible experiment to shut down the whole internet for five years and see what sort of art is produced over that span." Do you think that's a serious call to shut down the Internet? I don't. I think it's an off-the-cuff call for musicians to interact more with each other and audiences. Do I personally agree that the Internet will turn all musicians into Moby? Nah. Then again, Elton John might just have insight into the musical world that you and I do not.