Another problem with making LED displayes in this way is the price. You'd need one big wafer for a monitor, and if you think the P3 is expensive, just compare the surface area of a P3 chip (~2cm^2) to that of a reasonable screen display. a 20" at 2500cm^2 screen is 1250 times as big!
First off, Intel is not an ISP. They obviously don't want to use it themselves. They most likely have it because they don't want anyone else to use it.
Intel obviously wants to sell more of their big server chips to ISPs, be it Xeons, Itaniums, or PPros back when they took out this patent. This basically forces the ISPs to buy their own computers.
Further, Intel does not want the burden of this distributed computing placed onto the user's computers. The ISPs want a reasonable turnaround, and probably aren't going to settle for idle time only. Most likely, this processing WILL cut into the user's experience. Intel has lots of other uses for that power in mind.
We can all guess what happens to the high end markets. They suffer a reduction in CPU performance that they will likely miss and possibly some jerky performance a few seconds at a time. The high end market is likely not going to settle for an ISP like this, but might get stuck paying much more for their internet connection as the typical web surfer.
The typical web surfer's processor power is also something that shouldn't be wasted. Intel is embracing cheap PCs. Software modems are perfect for intel here. When built into the motherboard, they only cost $10, and get all their processing power from the CPU. Next comes sound. The number of chips in a computer can be greatly reduced if it's capability is just handled by the CPU. Of course, this comes at a major CPU performance price. By doing this, computer prices are falling, and even low end users have a use for all that processing power. Again, the ISP's processor usage would cut into the experience of the user.
If users percieve a decrease in performance, computers all of a sudden become a worse deal, and people are less likely to buy them. Remember, there are still lots of people without computers. If people catch wind that ISPs are only bothering stealing processor power from the fastest processors, all of a sudden a low end machine and a high end machine look a lot more similar in terms of performance.
The way I see it, Intel just wants everyone to buy their own processor an to get what they paid for.
Finding a faker isn't really that hard. Generally they'll fake blocks because they think it makes them look cool to be right at the top of the stats.
It's possible to design the client software to submit a false positive roughly every 4 billion keys tested (maybe just test to see if the first 32 bits are correct). It shouldn't be too hard for their servers to keep up with testing one out of 4 billion keys. When submitting any positive, the client should point to the exact key, so that a trusted machine can examine each key more carefully. When testing the key, this machine should test 2 things, is it the winning key, and is it one that should have been returned by the clients as positive. If it is a legitimate false positive, simply log it as such. If the client shouldn't have kicked it out as a false positive, then it's likely a cheater. This is to prevent people from submitting random blocks as positive to defeat the next mechanism.
Periodically, the proportion of false positives to keys tested should be be checked on. If the proportions are out of wack use a little stats to see how likely they would be out by that much had they been legit (This isn't a difficult calculation). If it looks highly unlikely that this could happen legitimately, again, the person is probably a cheater.
That's fine if the person is honest about who they are. You have to be honest about your identity for a high position in stats, but if you were just trying to sabotage the effort, you could use someone else's E-mail address. To defeat those people, some sort of authentication needs to take place. For example, when you sign up as a new participant the server E-mails you a password which must be entered into the client. They could even let the participant select their own password by supplying it with the first returned work unit.
Faster Athalons are MUCH easier to make
on
700 MHz Athlon
·
· Score: 1
It is rumored that AMD is cranking out 800MHz parts on the plant they're already using to make all their other Athlon chips. When these higher clocked chips come off the line, they have 3 choices, they can release a higher clocked product, stockpile them to release the faster product in the future, or to mark them as a slower part. I doubt AMD is putting much or any effort into making the higher clocked Athlons at this point, the chips running this fast are probably function of their K7 design efforts over the past few years.
Switching to the.18um process will further increase clock rate, but will take some effort. However, they already have the plant, and it's almost ready. Even if they only want to produce 600-700MHz parts, it still makes sense to use their newer plant as it will reduce the manufacturing costs.
In short, with all the work AMD has done over the past few years, increasing clock rates now is cheap and easy, and there's no reason not to. If they can make a 1GHz chip while Intel is pedalling their 600MHz chips, by all means! Most people are open to buying a single processor machine in place of a dual processor machine if they can get the same performance out of it. Naturally, this solution doesn't scale nearly as well, to compete with 4 or 8 processor computers.
Switching to dual processors, however requires a new chipset and new motherboards. Currently AMD isn't making either. I believe, that a chipset manufacturer, could make a SMP chipset based on the specs that AMD has released thus far, if not, AMD could simply release more specs to permit it.
Since the chipset manufacturers aren't likely to make SMP chipsets until the Athlon is already sucessful, AMD will have to produce on themselves or wait a while. Designing the chipset would be expensive, and still won't sell until there are SMP boards.
Given that AMD is cash strapped as it is, and the vast majority of the market is for single processor systems, designing their own chipset would probably be a grave mistake. I personally think AMD is on the right track now, they just need motherboards for their current Athlons to proceed.
On the surface, most people would want to keep their existing wires, and not have to replace them. This is, of course, a reasonable concern, and of significant importance if it meant re-wiring a building.
However, the time when every PC needs a Gigabit connection is still a while off. 10Mbit is still sufficient for the vast majority of networked PCs. The vast majority of network performance problems stem from frequent collisions, which in turn results from too many users hooked upto one port of a switch. Only a few will truely benefit from 100Mbit.
Most small to medium size companies can easily run all their servers off single 100Mbit ethernet connections, and have no problems at all. 1Gbit ethernet cards are really only needed in enterprise servers, and it will probably stay that way for some time (at least 5 years). Most PCs can't even saturate 100Mbits while providing or storing/displaying meaningful data.
I really don't see the problem with laying new lines between a handful of servers and switches which are normally placed in a central server room.
I would suggest going the way of the origional poster, set a limit for total cost of hardware only. Hardware prices will have to come from some indepented 3rd party, to rule out one party getting sweet deals. Ingram Micro's price lists sounds like a fair source. Setting a price limit on hardware eliminates any arguement about one side getting a better computer. It also rules out arguements about certain devices having better drivers for one OS than the other.
If we start to get into price of OS, and software, it would be perfectly fair for MS to start blowing the total cost of ownership horn.
Linux would still have to compete in the big computer arena, something it's not good at. However, if Linux were to loose here - which is entirely possible, it would be a clear indication that Linux still needs lots of work on large servers.
Any test where Microsoft picks the hardware, and the Linux community is not allowed to improve support for that hardware prior to the deadline cannot be concluded as fair. However, if the two are asked to build the fastest server they can for $50,000, we have to accept the resluts, good or bad.
While the K7 has multiple processor support, Kryotech's releases indicate that they are making cooling devices for uniprocessor systems only.
Kryotech clearly has the ability to cool a multi-processor system, in an older document of theirs they claimed to be able to cool processors up to 1KW. This looks like a newer cooling system - one which might not have quite as much capacity, but I would certainly expect them to be able to do 4 processors without much difficulty.
Only time will tell. I certainly hope they do come out with a multi-processor solution, even if I can't justify spending that much money on a computer for myself.
Slashdot Mirror
Another problem with making LED displayes in this way is the price. You'd need one big wafer for a monitor, and if you think the P3 is expensive, just compare the surface area of a P3 chip (~2cm^2) to that of a reasonable screen display. a 20" at 2500cm^2 screen is 1250 times as big!
First off, Intel is not an ISP. They obviously don't want to use it themselves. They most likely have it because they don't want anyone else to use it.
Intel obviously wants to sell more of their big server chips to ISPs, be it Xeons, Itaniums, or PPros back when they took out this patent. This basically forces the ISPs to buy their own computers.
Further, Intel does not want the burden of this distributed computing placed onto the user's computers. The ISPs want a reasonable turnaround, and probably aren't going to settle for idle time only. Most likely, this processing WILL cut into the user's experience. Intel has lots of other uses for that power in mind.
We can all guess what happens to the high end markets. They suffer a reduction in CPU performance that they will likely miss and possibly some jerky performance a few seconds at a time. The high end market is likely not going to settle for an ISP like this, but might get stuck paying much more for their internet connection as the typical web surfer.
The typical web surfer's processor power is also something that shouldn't be wasted. Intel is embracing cheap PCs. Software modems are perfect for intel here. When built into the motherboard, they only cost $10, and get all their processing power from the CPU. Next comes sound. The number of chips in a computer can be greatly reduced if it's capability is just handled by the CPU. Of course, this comes at a major CPU performance price. By doing this, computer prices are falling, and even low end users have a use for all that processing power. Again, the ISP's processor usage would cut into the experience of the user.
If users percieve a decrease in performance, computers all of a sudden become a worse deal, and people are less likely to buy them. Remember, there are still lots of people without computers. If people catch wind that ISPs are only bothering stealing processor power from the fastest processors, all of a sudden a low end machine and a high end machine look a lot more similar in terms of performance.
The way I see it, Intel just wants everyone to buy their own processor an to get what they paid for.
Finding a faker isn't really that hard. Generally they'll fake blocks because they think it makes them look cool to be right at the top of the stats.
It's possible to design the client software to submit a false positive roughly every 4 billion keys tested (maybe just test to see if the first 32 bits are correct). It shouldn't be too hard for their servers to keep up with testing one out of 4 billion keys. When submitting any positive, the client should point to the exact key, so that a trusted machine can examine each key more carefully. When testing the key, this machine should test 2 things, is it the winning key, and is it one that should have been returned by the clients as positive. If it is a legitimate false positive, simply log it as such. If the client shouldn't have kicked it out as a false positive, then it's likely a cheater. This is to prevent people from submitting random blocks as positive to defeat the next mechanism.
Periodically, the proportion of false positives to keys tested should be be checked on. If the proportions are out of wack use a little stats to see how likely they would be out by that much had they been legit (This isn't a difficult calculation). If it looks highly unlikely that this could happen legitimately, again, the person is probably a cheater.
That's fine if the person is honest about who they are. You have to be honest about your identity for a high position in stats, but if you were just trying to sabotage the effort, you could use someone else's E-mail address. To defeat those people, some sort of authentication needs to take place. For example, when you sign up as a new participant the server E-mails you a password which must be entered into the client. They could even let the participant select their own password by supplying it with the first returned work unit.
It is rumored that AMD is cranking out 800MHz parts on the plant they're already using to make all their other Athlon chips. When these higher clocked chips come off the line, they have 3 choices, they can release a higher clocked product, stockpile them to release the faster product in the future, or to mark them as a slower part. I doubt AMD is putting much or any effort into making the higher clocked Athlons at this point, the chips running this fast are probably function of their K7 design efforts over the past few years.
.18um process will further increase clock rate, but will take some effort. However, they already have the plant, and it's almost ready. Even if they only want to produce 600-700MHz parts, it still makes sense to use their newer plant as it will reduce the manufacturing costs.
Switching to the
In short, with all the work AMD has done over the past few years, increasing clock rates now is cheap and easy, and there's no reason not to. If they can make a 1GHz chip while Intel is pedalling their 600MHz chips, by all means! Most people are open to buying a single processor machine in place of a dual processor machine if they can get the same performance out of it. Naturally, this solution doesn't scale nearly as well, to compete with 4 or 8 processor computers.
Switching to dual processors, however requires a new chipset and new motherboards. Currently AMD isn't making either. I believe, that a chipset manufacturer, could make a SMP chipset based on the specs that AMD has released thus far, if not, AMD could simply release more specs to permit it.
Since the chipset manufacturers aren't likely to make SMP chipsets until the Athlon is already sucessful, AMD will have to produce on themselves or wait a while. Designing the chipset would be expensive, and still won't sell until there are SMP boards.
Given that AMD is cash strapped as it is, and the vast majority of the market is for single processor systems, designing their own chipset would probably be a grave mistake. I personally think AMD is on the right track now, they just need motherboards for their current Athlons to proceed.
On the surface, most people would want to keep their existing wires, and not have to replace them. This is, of course, a reasonable concern, and of significant importance if it meant re-wiring a building.
However, the time when every PC needs a Gigabit connection is still a while off. 10Mbit is still sufficient for the vast majority of networked PCs. The vast majority of network performance problems stem from frequent collisions, which in turn results from too many users hooked upto one port of a switch. Only a few will truely benefit from 100Mbit.
Most small to medium size companies can easily run all their servers off single 100Mbit ethernet connections, and have no problems at all. 1Gbit ethernet cards are really only needed in enterprise servers, and it will probably stay that way for some time (at least 5 years). Most PCs can't even saturate 100Mbits while providing or storing/displaying meaningful data.
I really don't see the problem with laying new lines between a handful of servers and switches which are normally placed in a central server room.
I would suggest going the way of the origional poster, set a limit for total cost of hardware only. Hardware prices will have to come from some indepented 3rd party, to rule out one party getting sweet deals. Ingram Micro's price lists sounds like a fair source. Setting a price limit on hardware eliminates any arguement about one side getting a better computer. It also rules out arguements about certain devices having better drivers for one OS than the other.
If we start to get into price of OS, and software, it would be perfectly fair for MS to start blowing the total cost of ownership horn.
Linux would still have to compete in the big computer arena, something it's not good at. However, if Linux were to loose here - which is entirely possible, it would be a clear indication that Linux still needs lots of work on large servers.
Any test where Microsoft picks the hardware, and the Linux community is not allowed to improve support for that hardware prior to the deadline cannot be concluded as fair. However, if the two are asked to build the fastest server they can for $50,000, we have to accept the resluts, good or bad.
While the K7 has multiple processor support, Kryotech's releases indicate that they are making cooling devices for uniprocessor systems only.
Kryotech clearly has the ability to cool a multi-processor system, in an older document of theirs they claimed to be able to cool processors up to 1KW. This looks like a newer cooling system - one which might not have quite as much capacity, but I would certainly expect them to be able to do 4 processors without much difficulty.
Only time will tell. I certainly hope they do come out with a multi-processor solution, even if I can't justify spending that much money on a computer for myself.