"...time travel is a ridiculous concept that belongs only in bad science fiction, not serious discussion."
This is slashdot, last time I checked there wasn't all that much serious discussion.
They do actually implement the version number, but don't generally publish it as numbers confuse people. http://en.wikipedia.org/wiki/NTFS#Versions so it about the same (ish) time line as ext2->3->4
I do, to my shame, have an MCSE in Windows 2000, which was purely for the purpose of getting that first job, which it did help me get. Now I have experience it is less than worthless (as well as very out of date).
You don't need education if you are smart, only people who need the benefit of a piece of paper to show they have a brain need one. Interview skills and experience will get you most jobs, even if you don't have the required qualifications. I speak as someone who does not have a degree and now work for the university I dropped out of for their IT department.
If you are smart you don't need the piece of paper to indicate such. It may take you a little longer to get moving upwards but experience is really what they want and you only get that by doing. If however you are not able to convince others you have a brain then get a piece of paper as this will help you.
I don't have a degree (in fact dropped out during my second year), but now work for the IT department of the university I went to, and I like to think my prospects for the future are good. But it did take a little while in a shitty job to get some experience to get this far.
err, you seem to have missed something fairly major in your understanding. Specifically about what constitutes a 'core'. These cards are based on the same chip in the GT280, so they have 240 stream processors, which are very good at specific types of calculation (If I was wiser I could tell you what types but I'm sure you can use google yourself). I believe that each of the chips has a 512 bit wide bus to 4GiB of memory. I'm not sure what the memory allocation per stream processor is but I think the other parts of the chip control what goes where. There probably are some bottlenecks but I don't know enough about it to be able to give useful information on the subject. There is something like 102GB/s memory bandwidth per 240 core chip.
In much the same way that the current Quadro FX cards are based on the same chip as the gaming gforce cards. But still the most expensive gaming card is ~£400, but you'll pay ~£1500 for the top of the line FX5700.
It's because workstation graphics cards are configured for accuracy above all else, where as gaming cards are configured for speed. Having a few pixels being wrong does not affect gaming at all, getting the numbers wrong in simulations is going to cause problems.
Mostly the people who use these cards care about OpenGL support, but some people do use them under Windows and DirectX.
This type of computing came in with the gforce 8 range when CUDA (Computer Unified Device Architecture) brought C programming to the massively parallel graphics chips. Which has allowed nVidia to port the Ageia PhysX technology to the gforce cards so a separate addin card is not necessary.
I believe that ATi are doing something similar with their FireGL cards, which again are based on the same chip as their Radeon cards. This is why they have both moved from Shader/Vertex to Unified Stream processors. This is a really interesting development if you happen to work in a research establishment, otherwise please move along nothing to see here.
Right so eliptic curves being intimately related to modular forms, which in turn can be used to prove Fermat's last theorem would be easy to do with a computer. Plus any program for finding proofs would be so complicated that there would be bugs in the code, leading to problems with the proofs it generated. That and mathematicians would not generally accept the results until they have been checked by peer review.
An awful lot of proofs actually go in non logical directions and incorporate very different areas of mathematics. In order to get this level of information in a program it would have to be extremely complex and open to inconsistencies.
You can read http://en.wikipedia.org/wiki/Four_color_theorem which shows that mathematicians do not all accept unverifiable proofs, such as those generated from a computer. Mathematicians like fact, absolute truth, not some method which cannot be checked. http://xkcd.com/435/
There may be some programs which can produce interesting information on which areas in maths may be linked (pattern matching) or for giving ideas as to the validity of statements but they will never become a large method of solving proofs*.
now incorperate the chipset, well northbridge, into the TDP, AMD is not that far off so it really does come down to AMDs performance at this power level. And if they can ship machines sub 2lb with 12-14" screens for $800 that's a big saving on the $2000 you'd pay for this sort of vaio. What with ATi becoming a force again in graphics and proving to be more power efficient than recent nVidia offerings this might become more interesting than you suggest.
you really should have read the article, it suggests that AMD is in fact going to miss the segment entirely. Going for full function laptops in slimmer lighter cases for sub $1000 (they cite the macbookair as a good example of the form factor, but not price obviously), essentially going you can pay twice what a netbook costs but actually get a proper computer, which seems like as good a ploy as any for AMD. They do have some competition from the core2 which does scale down to low power quite well, but intel chipsets are not always the most power friendly and AMD save on the northbridge with their design.
He HAS to do things in his first year, being as 2010 is a right off because of the midterms, then he's running again. So really the first year is the time when the president can actually get something done...
My point was that Americans spend more on healthcare per capita than most other developed countries and yet they still don't have universal coverage. That is expensive and inefficient, now if the Americans spent less than everyone else then mayhap you would have a point, efficiency being work done per unit money, so if you cost more you do less work per unit cost.
As I understand it Medicare spends about 2% of their costs on administration. Your average insurance company spends in the region of 20-25% on administration. You have to have health care, I would personally go for the one which is going to cost me less.
Please note my numbers may be wrong but for the same level of care if you don't have to worry about profits then you _can_ do things cheaper.
Er the rest of the world knew this 8 years ago, why do you think the rest of the world makes fun of the american system (well in Europe we certainly do).
What most amuses me is that the republicans like to paint the democrats as "tax and spend", where as I see the republicans as "and spend". At least the democrats generate income before spending it. Which is how in 8 years, mostly prosperous years, they've gone from the largest budget surplus in american history see here to the largest budget deficit in history see here
No you miss the point entirely. Parity data can give you all the data back, but ONLY if you know what the other bits you have are. Parity is a checksum* so lets assume you have 3 disks for any given disk slice you have 2 parts data to one part parity. If you add the corresponding bit values of each of the data part, and using only 1 bit answer, eg 1100 + 1010 = 0110, this gives you the third part the parity. So to replace any missing disk, which will contain both data and parity you need to know what the parity is to work out any missing data, and you also need to know both bits of data (when they happen to exist on the surviving drive) to fill in the missing parity
His point is also only valid if you have 12TB of data, because they HDD manufacturers say the failure rate is 1 12TB of data (using some other measure).
*May be the wrong word but I can't think of a better one right now.
I may be wrong about how parity is actually worked out (I can't be bothered to check wikipedia) but it is essentially along those lines, the data is manipulated in such a way that the system can take 1 missing part and reconstruct from what they have, and this means that they MUST read all the data from ALL the remaining drives.
RAID controllers probably don't read the entire surface during a rebuild but rather just the parity portions of the disk. This means in a RAID5 of 1TB disks. You are reading 1TB of data. Which would likely mean that you have a 1 in 12 chance of getting an URE. This may be an acceptable risk for some.
How exactly do you expect the parity partition, which being RAID 5 is split across every disk, to work out what the 'lost' data is without reading all the corresponding stripes on the other disks?
Firstly a MegaByte is 1000000 Bytes, a Mebibyte is 1048676 Bytes.
And really the formatting is not a huge issue in any way. For example a 250GB disk (manufacturers number) shows up in all OSes I've seen as 232GiB. Which tallies nicely with 250GB = 232.830644GiB, so you may be losing up to 0.4% (approximately) to formatting, which really isn't anything to the perception which you are losing ~14% because of the inability of users to work with numbers in different bases.
Slashdot Mirror
Are there any american universities that are competent?
"...time travel is a ridiculous concept that belongs only in bad science fiction, not serious discussion." This is slashdot, last time I checked there wasn't all that much serious discussion.
Step 3 (which really should be step 1) is 'Be Apple', there really are no unknowns in this equasion.
They do actually implement the version number, but don't generally publish it as numbers confuse people. http://en.wikipedia.org/wiki/NTFS#Versions so it about the same (ish) time line as ext2->3->4
I do, to my shame, have an MCSE in Windows 2000, which was purely for the purpose of getting that first job, which it did help me get. Now I have experience it is less than worthless (as well as very out of date).
You don't need education if you are smart, only people who need the benefit of a piece of paper to show they have a brain need one. Interview skills and experience will get you most jobs, even if you don't have the required qualifications. I speak as someone who does not have a degree and now work for the university I dropped out of for their IT department.
If you are smart you don't need the piece of paper to indicate such. It may take you a little longer to get moving upwards but experience is really what they want and you only get that by doing. If however you are not able to convince others you have a brain then get a piece of paper as this will help you.
I don't have a degree (in fact dropped out during my second year), but now work for the IT department of the university I went to, and I like to think my prospects for the future are good. But it did take a little while in a shitty job to get some experience to get this far.
There are 240 shaders on the GT280, which is why there is something like 1.4bn transistors on this chip. http://en.wikipedia.org/wiki/GeForce_200_Series#GeForce_GTX_200
err, you seem to have missed something fairly major in your understanding. Specifically about what constitutes a 'core'. These cards are based on the same chip in the GT280, so they have 240 stream processors, which are very good at specific types of calculation (If I was wiser I could tell you what types but I'm sure you can use google yourself). I believe that each of the chips has a 512 bit wide bus to 4GiB of memory. I'm not sure what the memory allocation per stream processor is but I think the other parts of the chip control what goes where. There probably are some bottlenecks but I don't know enough about it to be able to give useful information on the subject. There is something like 102GB/s memory bandwidth per 240 core chip.
In much the same way that the current Quadro FX cards are based on the same chip as the gaming gforce cards. But still the most expensive gaming card is ~£400, but you'll pay ~£1500 for the top of the line FX5700.
It's because workstation graphics cards are configured for accuracy above all else, where as gaming cards are configured for speed. Having a few pixels being wrong does not affect gaming at all, getting the numbers wrong in simulations is going to cause problems.
Mostly the people who use these cards care about OpenGL support, but some people do use them under Windows and DirectX.
This type of computing came in with the gforce 8 range when CUDA (Computer Unified Device Architecture) brought C programming to the massively parallel graphics chips. Which has allowed nVidia to port the Ageia PhysX technology to the gforce cards so a separate addin card is not necessary.
I believe that ATi are doing something similar with their FireGL cards, which again are based on the same chip as their Radeon cards. This is why they have both moved from Shader/Vertex to Unified Stream processors. This is a really interesting development if you happen to work in a research establishment, otherwise please move along nothing to see here.
Yes, yes there is, I'm shocked you didn't already have that fact in your head...
Because one is shiny and the other is vista.
Right so eliptic curves being intimately related to modular forms, which in turn can be used to prove Fermat's last theorem would be easy to do with a computer. Plus any program for finding proofs would be so complicated that there would be bugs in the code, leading to problems with the proofs it generated. That and mathematicians would not generally accept the results until they have been checked by peer review.
An awful lot of proofs actually go in non logical directions and incorporate very different areas of mathematics. In order to get this level of information in a program it would have to be extremely complex and open to inconsistencies.
You can read http://en.wikipedia.org/wiki/Four_color_theorem which shows that mathematicians do not all accept unverifiable proofs, such as those generated from a computer. Mathematicians like fact, absolute truth, not some method which cannot be checked. http://xkcd.com/435/
There may be some programs which can produce interesting information on which areas in maths may be linked (pattern matching) or for giving ideas as to the validity of statements but they will never become a large method of solving proofs*.
*until the busy beaver function has been solved
now incorperate the chipset, well northbridge, into the TDP, AMD is not that far off so it really does come down to AMDs performance at this power level. And if they can ship machines sub 2lb with 12-14" screens for $800 that's a big saving on the $2000 you'd pay for this sort of vaio. What with ATi becoming a force again in graphics and proving to be more power efficient than recent nVidia offerings this might become more interesting than you suggest.
you really should have read the article, it suggests that AMD is in fact going to miss the segment entirely. Going for full function laptops in slimmer lighter cases for sub $1000 (they cite the macbookair as a good example of the form factor, but not price obviously), essentially going you can pay twice what a netbook costs but actually get a proper computer, which seems like as good a ploy as any for AMD. They do have some competition from the core2 which does scale down to low power quite well, but intel chipsets are not always the most power friendly and AMD save on the northbridge with their design.
I like that they were allowed to call 98% of their results junk without having much understanding of the other 2%.
That does not end well, see...
He HAS to do things in his first year, being as 2010 is a right off because of the midterms, then he's running again. So really the first year is the time when the president can actually get something done...
My point was that Americans spend more on healthcare per capita than most other developed countries and yet they still don't have universal coverage. That is expensive and inefficient, now if the Americans spent less than everyone else then mayhap you would have a point, efficiency being work done per unit money, so if you cost more you do less work per unit cost.
As I understand it Medicare spends about 2% of their costs on administration. Your average insurance company spends in the region of 20-25% on administration. You have to have health care, I would personally go for the one which is going to cost me less.
Please note my numbers may be wrong but for the same level of care if you don't have to worry about profits then you _can_ do things cheaper.
Ah but the market will fix it because the market knows best, government involvement would cause it to be more inefficient and cost more...*
*this is a joke, obviously.
Er the rest of the world knew this 8 years ago, why do you think the rest of the world makes fun of the american system (well in Europe we certainly do).
What most amuses me is that the republicans like to paint the democrats as "tax and spend", where as I see the republicans as "and spend". At least the democrats generate income before spending it. Which is how in 8 years, mostly prosperous years, they've gone from the largest budget surplus in american history see here to the largest budget deficit in history see here
No you miss the point entirely. Parity data can give you all the data back, but ONLY if you know what the other bits you have are. Parity is a checksum* so lets assume you have 3 disks for any given disk slice you have 2 parts data to one part parity. If you add the corresponding bit values of each of the data part, and using only 1 bit answer, eg 1100 + 1010 = 0110, this gives you the third part the parity. So to replace any missing disk, which will contain both data and parity you need to know what the parity is to work out any missing data, and you also need to know both bits of data (when they happen to exist on the surviving drive) to fill in the missing parity
| 1 | 1 | 0 | 0 | data 1
| + | + | + | + |
| 1 | 0 | 1 | 0 | data 2
=
| 0 | 1 | 1 | 0 | parity
His point is also only valid if you have 12TB of data, because they HDD manufacturers say the failure rate is 1 12TB of data (using some other measure).
*May be the wrong word but I can't think of a better one right now.
I may be wrong about how parity is actually worked out (I can't be bothered to check wikipedia) but it is essentially along those lines, the data is manipulated in such a way that the system can take 1 missing part and reconstruct from what they have, and this means that they MUST read all the data from ALL the remaining drives.
RAID controllers probably don't read the entire surface during a rebuild but rather just the parity portions of the disk. This means in a RAID5 of 1TB disks. You are reading 1TB of data. Which would likely mean that you have a 1 in 12 chance of getting an URE. This may be an acceptable risk for some.
How exactly do you expect the parity partition, which being RAID 5 is split across every disk, to work out what the 'lost' data is without reading all the corresponding stripes on the other disks?
Firstly a MegaByte is 1000000 Bytes, a Mebibyte is 1048676 Bytes. And really the formatting is not a huge issue in any way. For example a 250GB disk (manufacturers number) shows up in all OSes I've seen as 232GiB. Which tallies nicely with 250GB = 232.830644GiB, so you may be losing up to 0.4% (approximately) to formatting, which really isn't anything to the perception which you are losing ~14% because of the inability of users to work with numbers in different bases.