I'm curious--what exactly happens when Deep Blue plays another instance of Deep Blue? Would having a chessmaster observe a series of games like this create an advantage? Also, how does Deep Blue determine what it's most optimal first move would be? Is it random (from a selection of openings) or deterministic?
"...and they still provide a quality product at a fair price."
Weeeeeeeell, I wouldn't go quite that far. While I think the product is certainly not poor quality, I recall that they have a ways to go.
IANACCIE or an experienced network engineer, but I used to have a lab with quite an array of Cisco equipment. I frequently heard (from lab engineers) that IOS was not terribly consistent (command wise) from one device to another, and that they were not impressed with the quality of some of the 'production' IOS releases we received.
The production environment I worked in used Cisco 12000's, and they had OC-12/45 line cards that seemed to fail more often than one would expect.
Keep in mind that Declan McCullagh is the author of this article. He isn't exactly a spring chicken in tech news reporting circles--he's been writing for Wired for quite a while.
Declan--savvy writer that he is--very likely did this for a particular reason. Keep in mind that people who make a living out of (hopefully) careful observation of a situation, as well as grammar, do not make "mistakes" like this.
When I read this, I imagined an Italian IT support engineer wildly gesticulating on a phone to a Dell Computer rep about the problems with his "viedo card" QUE COSA CON IL CARTA VIEDO, EH?!?
I am presently a consultant to the Federal government, and while there are some exceptions, I have to say that I am completely unimpressed. I thought that folks at the State of Oregon were lackadaisical--that is, until I started working with the Feds.
Granted, I've encountered SOME IT groups in government that work hard. For the most part, many don't (it takes folks in government roughly six times the amount of time to do ANYTHING involving systems administration.) However, I have never seen a stranger combination of feeling "entitled to job" combined with an overall professional malaise. Last time I checked, if you really, truly didn't add value to an organization, you weren't helping it survive.
The "job security" provided by government work is a misnomer. In the event that your division gets axed (which does happen--granted, every five to ten years), you end up holding the bag with skills that are way outdated, along with the remnants of bad cultural habits...
Sorry for the rant, but really, it's frustrating when you're trying to do something quite good for the organization you're working for...
I've noticed a number of disparate posts about criticisms about the practical design of this storage system. While these criticisms certainly have validity, I'd offer this counterpoint:
Sure, IBM's view of the architecture's abilities may be idealistic. That being said, hasn't creative engineering been the force behind making innovative, yet seemingly impractical, designs feasible? What about Shannon's law, dial-up modem bandwidth and the development of better encoding schemes (I admit that I have only a rudimentary knowledge of Shannon's law).
For instance, if a module goes out, why would the architecture still be "blocking?" Wouldn't it be possible to scale up the bandwidth of adjacent busses to accommodate for this? (In the electric utility business, critical transmission paths are designed with 'double contingency'--i.e., if one critical transmission line goes down, another line will be able to pick up the remaining capacity. If THAT line goes down, a third line can handle the capacity of the other two. Wasteful, yes, but better than the unpleasant consequences of urban areas losing a large percentage of their electric power.)
Another valid point raised surrounded the issue of water cooling. Again, wouldn't good industrial design and advances in materials science (pipe sealing components, etc.) possibly help us here?
By the way, is this concept really new? After all, haven't the folks at n-Cube been using "3-D" interconnection schemes for a long time (i.e., MediaCube servers used to serve broadcast-quality movies at hotels, cruise ships, etc.)?
You see, DC cannot travel long distances at high voltages or with a lot of amps. AC, on the other hand, can, and Tesla recognized this.
I hear people say this all the time. I work as a consultant to Bonneville Power Administration (US DOE), which is the holder of many long-distance power intertie lines. The high voltage DC "tie-line" that connects BPA (Oregon) to Los Angeles (California ISO) regularly sends hundreds of megawatts an hour to California, with far lower loss than the AC tie-lines. I know, because I work on the software that helps schedule transmission:-)
I'm not entirely sure why this is, but perhaps someone in the energy business can fill in why...
Interesting point made about my mailbox being USPS property (which, from what I understand, permits Federal penalties to be assessed for tampering with my mail and/or mailbox).
Does this mean that the USPS could, in a broad sense, have the ability to set up usps.com accounts for all of us, and apply similar protections?
Is the PRC trying to block out ALL offensive traffic, or MOST offensive traffic? What about the use of techniques like steganography to encrypt data into seemingly "harmless" pictures, mp3's, etc.?
If I were a government, I'd never want to try to get into this sort of "information freedom" cold war. It's companies like Cisco, Yahoo and the Chinese government vs. 1,273,111,290 Chinese people, some of whom have had the benefit of an American graduate education in computer science and mathematics. I'd say the odds are slanted toward the people...
I was concerned that Slashdot's servers were going to have a problem with all the traffic for this article... It's about damn time Slashdot got Slashdotted:-)
Best wishes for even more fun, love, and trippy 3 am discoveries in life!
Wouldn't it be possible to build filters to absorb the "intermod" that's leaking on to adjacent frequencies?
I suspect that no self-respecting hacker/engineer would a) want to actually transmit this sort of interference or b) want to raise the ire of the FCC, as hams do have an understandable penchant for locating and attempting to end the source of the interference.
"Costs matter. Price the service to reach critical mass first; adjust the pricing for profitability later; obsessively drive down internal costs at every opportunity."
ADJUST THE PRICING FOR PROFITABILITY LATER? Perhaps I'm reactionary, but reading this nearly made be lose all respect for the consultant who wrote the article.
Business should be about using analytical methods to develop a plan that will EVENTUALLY be profitable (helped by economies of scale, decreasing production costs, etc.). Adjusting for profitability later merely led to "burn rate," and the subsequent waste of millions of dollars from clueless investors.
When I was in school in 1992, I didn't have CIS available to me--only CS. I really despaired, since I appreciated, but didn't like the rigors of math (the irony being that I did poorly in discrete structures but passed the equivalent content in philosophy), but ran a profitable business assembling, repairing and selling clone PCs. My dean, knowing that I wasn't stupid, but unmotivated, forced me to leave CS (major number three, after EMT certification and philosophy). I ended up relying on my self-taught hacking experience from when I was in third grade, and relied on education from my mentors at companies I worked at. I became a competent systems engineer, and have enjoyed my work. My only regret is that I had held business majors in disdain, something that delayed a much-encouraged move (from friends who are coders and management types) into technical project management.
What I'm getting at is this: it's futile to argue the merits of CS vs. CIS vs. law vs. art history. I think it's more critical to pursue one's passions, tempered by one's abilities (I am, for instance, a medium built Asian guy--probably not smart to go for that defensive line position in football). Who cares what people think: do what makes you happy and which you have a passion for. There is a time and place for everyone to be...
Slashdot Mirror
I'm curious--what exactly happens when Deep Blue plays another instance of Deep Blue? Would having a chessmaster observe a series of games like this create an advantage? Also, how does Deep Blue determine what it's most optimal first move would be? Is it random (from a selection of openings) or deterministic?
"...and they still provide a quality product at a fair price."
Weeeeeeeell, I wouldn't go quite that far. While I think the product is certainly not poor quality, I recall that they have a ways to go.
IANACCIE or an experienced network engineer, but I used to have a lab with quite an array of Cisco equipment. I frequently heard (from lab engineers) that IOS was not terribly consistent (command wise) from one device to another, and that they were not impressed with the quality of some of the 'production' IOS releases we received.
The production environment I worked in used Cisco 12000's, and they had OC-12/45 line cards that seemed to fail more often than one would expect.
Has anyone else had these sorts of experiences?
Keep in mind that Declan McCullagh is the author of this article. He isn't exactly a spring chicken in tech news reporting circles--he's been writing for Wired for quite a while.
Declan--savvy writer that he is--very likely did this for a particular reason. Keep in mind that people who make a living out of (hopefully) careful observation of a situation, as well as grammar, do not make "mistakes" like this.
viedo card
When I read this, I imagined an Italian IT support engineer wildly gesticulating on a phone to a Dell Computer rep about the problems with his "viedo card" QUE COSA CON IL CARTA VIEDO, EH?!?
Very true, very true...
-Pete
P.S. Don't even get me started on how long it takes for me to get "unapproved software" installed.
I am presently a consultant to the Federal government, and while there are some exceptions, I have to say that I am completely unimpressed. I thought that folks at the State of Oregon were lackadaisical--that is, until I started working with the Feds.
Granted, I've encountered SOME IT groups in government that work hard. For the most part, many don't (it takes folks in government roughly six times the amount of time to do ANYTHING involving systems administration.) However, I have never seen a stranger combination of feeling "entitled to job" combined with an overall professional malaise. Last time I checked, if you really, truly didn't add value to an organization, you weren't helping it survive.
The "job security" provided by government work is a misnomer. In the event that your division gets axed (which does happen--granted, every five to ten years), you end up holding the bag with skills that are way outdated, along with the remnants of bad cultural habits...
Sorry for the rant, but really, it's frustrating when you're trying to do something quite good for the organization you're working for...
I've noticed a number of disparate posts about criticisms about the practical design of this storage system. While these criticisms certainly have validity, I'd offer this counterpoint:
Sure, IBM's view of the architecture's abilities may be idealistic. That being said, hasn't creative engineering been the force behind making innovative, yet seemingly impractical, designs feasible? What about Shannon's law, dial-up modem bandwidth and the development of better encoding schemes (I admit that I have only a rudimentary knowledge of Shannon's law).
For instance, if a module goes out, why would the architecture still be "blocking?" Wouldn't it be possible to scale up the bandwidth of adjacent busses to accommodate for this? (In the electric utility business, critical transmission paths are designed with 'double contingency'--i.e., if one critical transmission line goes down, another line will be able to pick up the remaining capacity. If THAT line goes down, a third line can handle the capacity of the other two. Wasteful, yes, but better than the unpleasant consequences of urban areas losing a large percentage of their electric power.)
Another valid point raised surrounded the issue of water cooling. Again, wouldn't good industrial design and advances in materials science (pipe sealing components, etc.) possibly help us here?
By the way, is this concept really new? After all, haven't the folks at n-Cube been using "3-D" interconnection schemes for a long time (i.e., MediaCube servers used to serve broadcast-quality movies at hotels, cruise ships, etc.)?
You see, DC cannot travel long distances at high voltages or with a lot of amps. AC, on the other hand, can, and Tesla recognized this.
:-)
I hear people say this all the time. I work as a consultant to Bonneville Power Administration (US DOE), which is the holder of many long-distance power intertie lines. The high voltage DC "tie-line" that connects BPA (Oregon) to Los Angeles (California ISO) regularly sends hundreds of megawatts an hour to California, with far lower loss than the AC tie-lines. I know, because I work on the software that helps schedule transmission
I'm not entirely sure why this is, but perhaps someone in the energy business can fill in why...
Working on E-Tag 1.7? E-mail pblee@bpa.gov!
Interesting point made about my mailbox being USPS property (which, from what I understand, permits Federal penalties to be assessed for tampering with my mail and/or mailbox).
Does this mean that the USPS could, in a broad sense, have the ability to set up usps.com accounts for all of us, and apply similar protections?
Is the PRC trying to block out ALL offensive traffic, or MOST offensive traffic? What about the use of techniques like steganography to encrypt data into seemingly "harmless" pictures, mp3's, etc.?
If I were a government, I'd never want to try to get into this sort of "information freedom" cold war. It's companies like Cisco, Yahoo and the Chinese government vs. 1,273,111,290 Chinese people, some of whom have had the benefit of an American graduate education in computer science and mathematics. I'd say the odds are slanted toward the people...
I was concerned that Slashdot's servers were going to have a problem with all the traffic for this article... It's about damn time Slashdot got Slashdotted :-)
Best wishes for even more fun, love, and trippy 3 am discoveries in life!
Cheers,
-Pete
PS. I've included your wedding present below:
>
Wouldn't it be possible to build filters to absorb the "intermod" that's leaking on to adjacent frequencies?
I suspect that no self-respecting hacker/engineer would a) want to actually transmit this sort of interference or b) want to raise the ire of the FCC, as hams do have an understandable penchant for locating and attempting to end the source of the interference.
"Costs matter. Price the service to reach critical mass first; adjust the pricing for profitability later; obsessively drive down internal costs at every opportunity."
ADJUST THE PRICING FOR PROFITABILITY LATER? Perhaps I'm reactionary, but reading this nearly made be lose all respect for the consultant who wrote the article.
Business should be about using analytical methods to develop a plan that will EVENTUALLY be profitable (helped by economies of scale, decreasing production costs, etc.). Adjusting for profitability later merely led to "burn rate," and the subsequent waste of millions of dollars from clueless investors.
I find it highly ironic that an ad professional who recently sent out an e-mail ad campaign has "NO SPAM" in his e-mail address :)
Nice argument--I agree with this...
Yeah, no joke--I believe the name for TeX comes from a Greek letter...
I think a very fascinating concept for a language, although it does look all greek to me :-)
When I was in school in 1992, I didn't have CIS available to me--only CS. I really despaired, since I appreciated, but didn't like the rigors of math (the irony being that I did poorly in discrete structures but passed the equivalent content in philosophy), but ran a profitable business assembling, repairing and selling clone PCs. My dean, knowing that I wasn't stupid, but unmotivated, forced me to leave CS (major number three, after EMT certification and philosophy). I ended up relying on my self-taught hacking experience from when I was in third grade, and relied on education from my mentors at companies I worked at. I became a competent systems engineer, and have enjoyed my work. My only regret is that I had held business majors in disdain, something that delayed a much-encouraged move (from friends who are coders and management types) into technical project management. What I'm getting at is this: it's futile to argue the merits of CS vs. CIS vs. law vs. art history. I think it's more critical to pursue one's passions, tempered by one's abilities (I am, for instance, a medium built Asian guy--probably not smart to go for that defensive line position in football). Who cares what people think: do what makes you happy and which you have a passion for. There is a time and place for everyone to be...