They get that fast by being really simple. They don't do speculative execution, instruction reordering, etc. They don't have TLBs, caches, or a number of other things that make circuits more complex.
All the indications I've seen will show that the SPEs will be programmed via a "job" model, not a thread model. So you have jobs ("cells") that have some code and some data, you ask the OS to ship it off to an SPE, and then go do something else while you wait for the results.
The GeForce FX's 200 gigaflops aren't all general-purpose though. A lot of them come from fixed-purpose circuits that you can't use for your own calculations. For a general purpose program, you've got about 80 gigaflops, of which you can extract 50-60 gigaflops in real-world programs.
The SPE doesn't context switch. The OS does regular scheduling on the PPE, which can execute 2 threads at the same time. The SPEs, however, are scheduled in batches. Each runs autonomously (in its own thread), and runs the same task until its completed. So when the PPE context switches, the SPEs can still run whatever they are running.
Just a couple of points. First, KDE cares way more about RAM than CPU. 256MB is a must, after that, it's diminishing returns. Second, every point release in the 3.x series has been faster than the previous one. 3.3 is a lot faster than 3.0 was.
The first footnote is not inconsistent with the source, as 300km^2 is considerably less than the 400km radius area that I cited for 90 centigray exposure. Also note that the second parts of your estimate are accrued over a very long timespan, while the PBS numbers are only over seven days. The PBS scenario is more realistic, because we are talking about an attack here, not long term environmental impact, and any survivors would be evacuated long before the seven day period was over.
With regards to your very last point, it should be noted that these reactors use a graphite core covered in SiC. They won't burn, unless the mob in question had specialized tools.
Every change requires compromise. The question is whether it is a good compromise to inconvenience some people for the greater good (within the bounds of Constitutional rights, of course). Since there is no Constitutional question here, the only question is whether it is justifiable to inconvenience some people, in return for cleaner air, water, and political stability for everyone else. In the nuclear case, it's definitely justifiable. It's not like they're being led on a forced death march. It's just that if they're afraid of living near nuclear waste (and there isn't all that much reason to be), then they have to move. There is no Constitutional guarantee that you have to feel safe where you live, or else we'd have to abolish Detroit!
I found where you got that statistic, and should point out that the 62% isn't the number of people who consider it the "most important" thing, but the number of people who consider it to be important. Thus, the 62% rating puts it just below "immigration" and "poverty", both of which proved to be non-issues in the last election.
Most people will say they care about the environment, but that doesn't mean they are the type of people deluded enough to oppose nuclear power. When asked to prioritize, few people will make it the highest priority, as someone who could be called an "environmentalist" would do.
The point is that it is not inconsistent for democrats to (purportedly) support France (which is nuclear-powered), because the majority of democrats aren't environmentalists.
A 1 megaton bomb will release 90 rad (1 rad = 1 centigray) at a radius about 400km. So it seems to me that the bomb + a nuclear reactor doesn't make the situation all that much worse!
There is a nuclear plant outside Williamsburg, VA, and while I don't live there, there is a big university (William and Mary) in the town. I know people who go there and I've never so much as heard about it. It's kind of hard to move within seeing distance of a plant, since most are on 700-2000+ acre sites, but living in a city that has a nuclear plant isn't a big deal at all.
If you're worried about nuclear waste disposal centers, don't move near one. If you live near one, move! Progress can't be held up just because some twits want to live in the middle of nowhere.
The environmentalists make up a very small percentage of democrats. Greenpeace and Sierra Club put together have just over a million members, and these are by far the biggest environmental groups in the US. "The environment" regularly makes a single-digit impact on surveys of what's important to voters, even among democrats. So your little rant is entirely unwarrented --- it is entirely consistent for a democrat to be comfortable with nuclear power.
The OS has an APUlet batch scheduler. You submit an APUlet to the scheduler, along with its timing requirements (for time-sensitive jobs), and it will set up the APU, and send the results back when its done.
It's not just a matter of being selfish vs being generous. It's about spending your money wisely. This is not an example of spending your money wisely. This is especially true if you remember that for this thing to work at all, the donations need to be in the range of $200-$500, not $20-$50! $20-$50 donations would mean getting anywhere from 2 to 4 million people donating. The show's total viewership was in that range!
It's not just for picking this "cause" instead of a charity. It's for a number of other reasons, the main one being it's ridiculous to "donate" to a for-profit operation! These people are setting themselves up to get screwed by Paramount. They'll probably just be satisfied with getting the show back on the air, and not demand anything like a cut of advertising dollars, partial ownership, broadcast rights, etc. From my point of view, it's not unlike people who spend a lot of money on lottery tickits. It's a stupid thing to do, and I'm perfectly justified in pointing that out.
Star Trek isn't a cause. It's entertainment, and bad entertainment at that. It's fundementally different from something like the Hubble Space Telescope.
Yes, they are free to do with their money as they please, and we're free to call them morons for it. Beyond that, ruling an island nation would be cool. This is just utterly moronic. I mean, I could pay people to repeatedly kick me in the nuts, but I'm not that stupid...
The cell processor is only really fast when the spus are in use, which means 32-bit non-branching floating-point arithmatic.
Actually, the APUs handle both floating-point and integer vector arithmetic (like Altivec or SSE2). And it has a branch unit, and since its so simple, it's pipeline is probably a reasonable length. So it would likely be good on branchy integer code, as long as it stayed within a small dataset. I can think of uses for that --- certain fluid dynamics, structural analysis, or thermal analysis problems could fit nicely. Also, it could be good at solving constraint systems, numeric integration, and a number of other algorithms where runtime can be quite large for small datasets.
The SPEs don't context switch. They process a single APUlet until it signals conclusion, and then stops until another APUlet is loaded. I presume you could interrupt it and load a different APUlet, but the APUs aren't cores that you can schedule threads on.
they bought into that they could actually use the 6.2gflops meaningfully.
The EE wasn't some magic architecture that could only do 6.2 gigaflops on a single benchmark. It was a simple vector architecture. Developers knew that. If they thought they could get 6.2 gigaflops out of it, but they didn't know how to program vector architectures, then whose fault is that? I don't see anybody complaining about the benchmarks published for things like the Earth Simulator, because it's pretty hard to get close to peak on that thing too!
With regards to what Blachford said --- I refuse to attribute journalistic hype to IBM and Sony. It's the journalists making far out claims, not IBM. A lot of people are legitimately excited about the thing, not because they believe they can actually get 256gigaflops running Quake III, but because they see the 256gigaflops peak, and salivate at the real numbers they could post in their apps.
Slashdot Mirror
Yes, KDE uses a lot of RAM. But being a memory hog and a processor hog are two different things :)
They get that fast by being really simple. They don't do speculative execution, instruction reordering, etc. They don't have TLBs, caches, or a number of other things that make circuits more complex.
All the indications I've seen will show that the SPEs will be programmed via a "job" model, not a thread model. So you have jobs ("cells") that have some code and some data, you ask the OS to ship it off to an SPE, and then go do something else while you wait for the results.
The GeForce FX's 200 gigaflops aren't all general-purpose though. A lot of them come from fixed-purpose circuits that you can't use for your own calculations. For a general purpose program, you've got about 80 gigaflops, of which you can extract 50-60 gigaflops in real-world programs.
The SPE doesn't context switch. The OS does regular scheduling on the PPE, which can execute 2 threads at the same time. The SPEs, however, are scheduled in batches. Each runs autonomously (in its own thread), and runs the same task until its completed. So when the PPE context switches, the SPEs can still run whatever they are running.
Just a couple of points. First, KDE cares way more about RAM than CPU. 256MB is a must, after that, it's diminishing returns. Second, every point release in the 3.x series has been faster than the previous one. 3.3 is a lot faster than 3.0 was.
this source
The first footnote is not inconsistent with the source, as 300km^2 is considerably less than the 400km radius area that I cited for 90 centigray exposure. Also note that the second parts of your estimate are accrued over a very long timespan, while the PBS numbers are only over seven days. The PBS scenario is more realistic, because we are talking about an attack here, not long term environmental impact, and any survivors would be evacuated long before the seven day period was over.
With regards to your very last point, it should be noted that these reactors use a graphite core covered in SiC. They won't burn, unless the mob in question had specialized tools.
Every change requires compromise. The question is whether it is a good compromise to inconvenience some people for the greater good (within the bounds of Constitutional rights, of course). Since there is no Constitutional question here, the only question is whether it is justifiable to inconvenience some people, in return for cleaner air, water, and political stability for everyone else. In the nuclear case, it's definitely justifiable. It's not like they're being led on a forced death march. It's just that if they're afraid of living near nuclear waste (and there isn't all that much reason to be), then they have to move. There is no Constitutional guarantee that you have to feel safe where you live, or else we'd have to abolish Detroit!
I found where you got that statistic, and should point out that the 62% isn't the number of people who consider it the "most important" thing, but the number of people who consider it to be important. Thus, the 62% rating puts it just below "immigration" and "poverty", both of which proved to be non-issues in the last election.
Most people will say they care about the environment, but that doesn't mean they are the type of people deluded enough to oppose nuclear power. When asked to prioritize, few people will make it the highest priority, as someone who could be called an "environmentalist" would do.
The point is that it is not inconsistent for democrats to (purportedly) support France (which is nuclear-powered), because the majority of democrats aren't environmentalists.
A 1 megaton bomb will release 90 rad (1 rad = 1 centigray) at a radius about 400km. So it seems to me that the bomb + a nuclear reactor doesn't make the situation all that much worse!
There is a nuclear plant outside Williamsburg, VA, and while I don't live there, there is a big university (William and Mary) in the town. I know people who go there and I've never so much as heard about it. It's kind of hard to move within seeing distance of a plant, since most are on 700-2000+ acre sites, but living in a city that has a nuclear plant isn't a big deal at all.
If you're worried about nuclear waste disposal centers, don't move near one. If you live near one, move! Progress can't be held up just because some twits want to live in the middle of nowhere.
The environmentalists make up a very small percentage of democrats. Greenpeace and Sierra Club put together have just over a million members, and these are by far the biggest environmental groups in the US. "The environment" regularly makes a single-digit impact on surveys of what's important to voters, even among democrats. So your little rant is entirely unwarrented --- it is entirely consistent for a democrat to be comfortable with nuclear power.
Here in Virginia, the cops won't touch you for 20 mile an hour over the limit, while back home the cops will ding you for 1 mph over the limit.
Heh. Go 5mph over the limit in Vienna, VA, sometime. The cops, who have nothing better to do, will nail your ass.
Sounds quite comforting, really. "Pebble bed?" Reminds me of a babbling brook...
Physics?
You're right. In that case, see the dictionary definition.
The OS has an APUlet batch scheduler. You submit an APUlet to the scheduler, along with its timing requirements (for time-sensitive jobs), and it will set up the APU, and send the results back when its done.
It's not just a matter of being selfish vs being generous. It's about spending your money wisely. This is not an example of spending your money wisely. This is especially true if you remember that for this thing to work at all, the donations need to be in the range of $200-$500, not $20-$50! $20-$50 donations would mean getting anywhere from 2 to 4 million people donating. The show's total viewership was in that range!
It's not just for picking this "cause" instead of a charity. It's for a number of other reasons, the main one being it's ridiculous to "donate" to a for-profit operation! These people are setting themselves up to get screwed by Paramount. They'll probably just be satisfied with getting the show back on the air, and not demand anything like a cut of advertising dollars, partial ownership, broadcast rights, etc. From my point of view, it's not unlike people who spend a lot of money on lottery tickits. It's a stupid thing to do, and I'm perfectly justified in pointing that out.
Star Trek isn't a cause. It's entertainment, and bad entertainment at that. It's fundementally different from something like the Hubble Space Telescope.
Yes, they are free to do with their money as they please, and we're free to call them morons for it. Beyond that, ruling an island nation would be cool. This is just utterly moronic. I mean, I could pay people to repeatedly kick me in the nuts, but I'm not that stupid...
The cell processor is only really fast when the spus are in use, which means 32-bit non-branching floating-point arithmatic.
Actually, the APUs handle both floating-point and integer vector arithmetic (like Altivec or SSE2). And it has a branch unit, and since its so simple, it's pipeline is probably a reasonable length. So it would likely be good on branchy integer code, as long as it stayed within a small dataset. I can think of uses for that --- certain fluid dynamics, structural analysis, or thermal analysis problems could fit nicely. Also, it could be good at solving constraint systems, numeric integration, and a number of other algorithms where runtime can be quite large for small datasets.
The SPEs don't context switch. They process a single APUlet until it signals conclusion, and then stops until another APUlet is loaded. I presume you could interrupt it and load a different APUlet, but the APUs aren't cores that you can schedule threads on.
they bought into that they could actually use the 6.2gflops meaningfully.
The EE wasn't some magic architecture that could only do 6.2 gigaflops on a single benchmark. It was a simple vector architecture. Developers knew that. If they thought they could get 6.2 gigaflops out of it, but they didn't know how to program vector architectures, then whose fault is that? I don't see anybody complaining about the benchmarks published for things like the Earth Simulator, because it's pretty hard to get close to peak on that thing too!
With regards to what Blachford said --- I refuse to attribute journalistic hype to IBM and Sony. It's the journalists making far out claims, not IBM. A lot of people are legitimately excited about the thing, not because they believe they can actually get 256gigaflops running Quake III, but because they see the 256gigaflops peak, and salivate at the real numbers they could post in their apps.