Speaking from experience, cows are pretty dumb. And by dumb, I mean really dumb. What you have expressed are human feelings.
Cows spook really easily. A hawk squarking above will send them crazy... now I'm pretty sure the natural predator of the cow is not a hawk. The aversion to cliffs is once again the spooking. Heights scare them. And water, cows love water.
I'm not advocating non-humane slaughter, and I think the painless breed is pretty stupid (most fences around where I grew up were electric) and pointless, just want to point out a few things here
The article specifically talk about document searching. Some of the document searching people we have require >110 Terabytes of memory (per job). On a cluster that is pretty difficult to store in ram.
A lot of supercomputers are clusters and those clusters typically don't have huge amounts of memory... they are high on compute.
My parents, public teachers for 40 years now (starting the transition to retirement) have always thought the way you do now.
They have always advocated that you teach to the bottom. The mid range and top range will either listen to the different ways of presenting the material, push ahead or help their neighbours. Of three classes teaching the same material per year, all the students were rounded up into a room big enough while the old man would introduce the material and work examples. The other two teachers would walk around and help where necessary.
Of course, there was no such subject as basic maths at their school. They made every kid do the mid level (only three levels in Australia). Over the last 15 years or so, they have had a 1% failure rate ( less than 50% final result over two years ). Just shows that it can work and people need to stop preaching how much brighter they are. Mom and Dad used it as a measure of stupidity.
The truly bright (as you would expect, a lot of slashdot believe they are) will have the rest of their lives for opportunity to harness.
Not even close. Floating point performance on a single core has still been increasing according to Moores law. 256 bit wide registers, vectorised/packed instructions, xsse4.1, etc
I see the point you're making, but what happens when I write a parallel program? I make a request to the array services daemon that I want 4 processors and that gives me a single socket (in my current, massively outdated system:) ), or in your nomenclature a single processor. I see it as having 4 processors (or cores if you will) and hence the need to get a naming scheme that is consistant across hardware and software.
Works fine until we start getting specialised cores with specialised instructions. I've adopted sockets and processors. A pizza box can 8 processors (two sockets). A single processor, in this case, is still a generalised processing element.
I think the amazing thing is that the phone is more powerful *and* only requires a battery with capacity measured in milliamps. That detail would knock an engineer from those days clean out of the ballpark.
No. After the brilliance of Ender's Game, I made it as far as Xenocide and made the decision to pull the plug. That way the good memories will last forever.
Woah! No so fast there Tex. Nobody'll get fired. Not even a reprimand. Incident reports will have to be submitted and if it is considered really, really bad, a 'problem' ticket will be requested. But that's it.
Australian government... Vogons aren't even in the same class.
The ABC (Australian Broadcast Commission) did this for a live state -vs- state football match we have here. It was quite good. Measuring sentiment of who would win based on the positiveness or negativeness of 'tweets' (or whatever the fuck they're called).
Hello Jason Levine. I'm sorry to hear about your misfortune.
I guess you could start guarding yourself by not using your name on any forums or such on the Internet. It'll take a while for your name to disappear into the archives, but will bring less attention to yourself. I mean, if I wanted to damage somebody, I'd start by collecting every post you ever made on/. and trawl for information regarding localities and events. I bet it wouldn't take too long. It is why I just made up a name to use for the Internet and there is no link to my real one.
I'm not trying to flame. It's just that there are some people on the planet that I don't think I can trust.
Wow. Modded as a troll and trollish/flamebait replies modded as insightful.
I guess you hit the nail on the head by pointed out an ugly truth about how people are reckless with their financial responsibility. I think you're right.
I'm very good with money. Not a total tight-arse, but I manage my money (even over the GFC I've increased wealth by 70%). If somebody is employed, and as part of that employment, they have to balance a budget or place orders with vendors that are the best value based on what is needed, the best employee is somebody who is good with money. Somebody with proven history of bad personal financial decisions is not going to know any better with somebody elses money.
I actually have a major problem with being ripped off (that is, paying the same as everybody else or even above 80% of that price) and that part of me argues the finer points with anything involving cash. Always looking for the edge.
Not at all absurd. I realise that the gpu is a compute workhorse. That's not the issue. It is the data transfer rate to and from the card. Transferring 3GiB takes quite a while. Pulling the results back takes a while also. That's what kills it. The cpu can get the work done in that time.
I'm using the cuda blas routines, examples from the sdk and those published as 'glorious almighty' codes. Everything that the card does is timed as it is all time to solution.
I've found that an O(n^3) algorithm or less should be run on cpu. The overhead of moving to gpu memory is just too high. The gen2 pci is faster, but that just means I do #pragma omp parallel for and set the number of processors to 2.
The comparisons of gpu and cpu code are not fair. They talk about highly optimised code for the gpu but totally neglect the cpu code (only use a O2 with the gcc compiler and that's it). On a E5430 Xeon, intel compiler and well written code, an O(n^3) or less is faster.
I'd say yes, copyright of academic works should be abolished.
I work at a University and would be happy for anybody to have any work I produce. The tax payers pay my wage so the tax payers should benefit. I see it pretty simply.
The groups I work with should share work also. We already have an edge over other research groups worldwide cause we did something first and are 'in the zone'. If another research group overtakes us, well, we dropped the ball.
Slashdot Mirror
5- you'd be amazed how picky (or impossible...so far) it is to coax a chunk of plant tissue into creating a whole new plant out of it's cells
Buy it a drink?
You haven't been around many cattle have you?
Speaking from experience, cows are pretty dumb. And by dumb, I mean really dumb. What you have expressed are human feelings.
Cows spook really easily. A hawk squarking above will send them crazy ... now I'm pretty sure the natural predator of the cow is not a hawk. The aversion to cliffs is once again the spooking. Heights scare them. And water, cows love water.
I'm not advocating non-humane slaughter, and I think the painless breed is pretty stupid (most fences around where I grew up were electric) and pointless, just want to point out a few things here
The article specifically talk about document searching. Some of the document searching people we have require >110 Terabytes of memory (per job). On a cluster that is pretty difficult to store in ram.
A lot of supercomputers are clusters and those clusters typically don't have huge amounts of memory ... they are high on compute.
My parents, public teachers for 40 years now (starting the transition to retirement) have always thought the way you do now.
They have always advocated that you teach to the bottom. The mid range and top range will either listen to the different ways of presenting the material, push ahead or help their neighbours. Of three classes teaching the same material per year, all the students were rounded up into a room big enough while the old man would introduce the material and work examples. The other two teachers would walk around and help where necessary.
Of course, there was no such subject as basic maths at their school. They made every kid do the mid level (only three levels in Australia). Over the last 15 years or so, they have had a 1% failure rate ( less than 50% final result over two years ). Just shows that it can work and people need to stop preaching how much brighter they are. Mom and Dad used it as a measure of stupidity.
The truly bright (as you would expect, a lot of slashdot believe they are) will have the rest of their lives for opportunity to harness.
CPU speed has stagnated,
Not even close. Floating point performance on a single core has still been increasing according to Moores law. 256 bit wide registers, vectorised/packed instructions, xsse4.1, etc
I see the point you're making, but what happens when I write a parallel program? I make a request to the array services daemon that I want 4 processors and that gives me a single socket (in my current, massively outdated system :) ), or in your nomenclature a single processor. I see it as having 4 processors (or cores if you will) and hence the need to get a naming scheme that is consistant across hardware and software.
And please don't say threads :s
Works fine until we start getting specialised cores with specialised instructions. I've adopted sockets and processors. A pizza box can 8 processors (two sockets). A single processor, in this case, is still a generalised processing element.
I think the amazing thing is that the phone is more powerful *and* only requires a battery with capacity measured in milliamps. That detail would knock an engineer from those days clean out of the ballpark.
That was my immediate thought as well.
We're about to drop $250K on a GPU cluster and if performance increases to that amount in 6 years, why on earth would we buy now?
Dammit, there's just no win when you fork out for clusters (of any kind).
Should spend 50K now, stick the rest into stocks and buy 50K every year. Of course, the dudes up the tree don't like that kind of thinking.
Java, it's like the girl next door.
She ain't pretty, but she'll do what you want.
Not sure if somebody above has said it yet, but you need to install powertop and run it as root.
That'll increase battery performance.
No. After the brilliance of Ender's Game, I made it as far as Xenocide and made the decision to pull the plug. That way the good memories will last forever.
tits yeah this'd boobs be great ass cause then nipple there'd be no sex need for twitter pussy. Overall quality would flange increase.
Woah! No so fast there Tex. Nobody'll get fired. Not even a reprimand. Incident reports will have to be submitted and if it is considered really, really bad, a 'problem' ticket will be requested. But that's it.
Australian government ... Vogons aren't even in the same class.
The ABC (Australian Broadcast Commission) did this for a live state -vs- state football match we have here. It was quite good. Measuring sentiment of who would win based on the positiveness or negativeness of 'tweets' (or whatever the fuck they're called).
Here
I'd say you were a dyslexic psychopath.
Do toilet flushes swirl in the opposite direction on this planet?
What? Up?
Hello Jason Levine. I'm sorry to hear about your misfortune.
I guess you could start guarding yourself by not using your name on any forums or such on the Internet. It'll take a while for your name to disappear into the archives, but will bring less attention to yourself. I mean, if I wanted to damage somebody, I'd start by collecting every post you ever made on /. and trawl for information regarding localities and events. I bet it wouldn't take too long. It is why I just made up a name to use for the Internet and there is no link to my real one.
I'm not trying to flame. It's just that there are some people on the planet that I don't think I can trust.
Wow. Modded as a troll and trollish/flamebait replies modded as insightful.
I guess you hit the nail on the head by pointed out an ugly truth about how people are reckless with their financial responsibility. I think you're right.
Actually, I like it.
I'm very good with money. Not a total tight-arse, but I manage my money (even over the GFC I've increased wealth by 70%). If somebody is employed, and as part of that employment, they have to balance a budget or place orders with vendors that are the best value based on what is needed, the best employee is somebody who is good with money. Somebody with proven history of bad personal financial decisions is not going to know any better with somebody elses money.
I actually have a major problem with being ripped off (that is, paying the same as everybody else or even above 80% of that price) and that part of me argues the finer points with anything involving cash. Always looking for the edge.
Who would be best for placing orders?
That was a lovely, optimistic post. Nope, I'm gonna sleep well tonight. No cold sweats for me.
Not at all absurd. I realise that the gpu is a compute workhorse. That's not the issue. It is the data transfer rate to and from the card. Transferring 3GiB takes quite a while. Pulling the results back takes a while also. That's what kills it. The cpu can get the work done in that time.
I'm using the cuda blas routines, examples from the sdk and those published as 'glorious almighty' codes. Everything that the card does is timed as it is all time to solution.
I've found that an O(n^3) algorithm or less should be run on cpu. The overhead of moving to gpu memory is just too high. The gen2 pci is faster, but that just means I do #pragma omp parallel for and set the number of processors to 2.
The comparisons of gpu and cpu code are not fair. They talk about highly optimised code for the gpu but totally neglect the cpu code (only use a O2 with the gcc compiler and that's it). On a E5430 Xeon, intel compiler and well written code, an O(n^3) or less is faster.
I'd say yes, copyright of academic works should be abolished.
I work at a University and would be happy for anybody to have any work I produce. The tax payers pay my wage so the tax payers should benefit. I see it pretty simply.
The groups I work with should share work also. We already have an edge over other research groups worldwide cause we did something first and are 'in the zone'. If another research group overtakes us, well, we dropped the ball.
Wow. I really cannot understand why this is a troll. Simple maybe but not a troll. Must be attracting trigger happy nerds with this headline.