Actually, I would be more interested in finding out how Leia got away/wasn't known to Vader.
As to it not making sense for Luke going to Tatooine, I thought it actually did - wasn't Luke given to "Uncle Owen" (whom I presume is Anakin's half brother) to be raised?
Give Lucas some credit. He "Spoils" something that shouldn't be news, but holds back the piece that is actually more intriguing (at least I hope it is).
I bet you can find legions of them dumb enough to get the tat. I'm willing to bet that the winner will be the person willing to devote the most square footage for it.
If Dell took on AMD it would probably end up killing AMD.
Intel's strength has always been its ability to drop prices for large customers to levels below that of its competitors. For AMD to become a Dell supplier, it would have to lower its prices to the point where they were not only unprofitable but probably bleeding money from an opened artery.
Intel might not want anything more than AMD to make an offer to Dell that they can't match. Even if it didn't kill AMD, it would put AMD in the place where Intel believes it belongs.
I probably should have said MOT683xx, PowerPC or mobile Pentiums, which seem to be the primary choice for most high-end embedded applications that I am seeing, instead of the big three.
The CNet article was pretty vague and seemed to imply that the custom instruction capability provided better number crunching capabilities than these traditional server/workstation processors.
Pretty skimpy blurb - I suspect that the product is either a) vapourware or b) a lot more limited than is discussed in the article.
From the article, I presume that the processor's microinstruction memory can be updated with special information embedded in the executable file. This is not as unique as you might think: virtually all Intel and AMD processors have the ability to have their microinstruction memory updated during the boot process - this is used up upload microinstruction updates/corrections without requiring a new chip. What's implied in the article's title (note, not in the text), is that this memory can be updated on the fly which, I presume, means that the instruction set changes for each process that is loaded.
The immediate questions that I have are:
1. How is this accomplished in a multi-process (ie Linux) operating environment where the op-codes given to the custom instructions in one process are the same as another? Does this mean that the microinstruction memory must be reset during task switch to avoid this problem? What is the task switch latency penalty for this process?
I suspect that only one process with custom instructions can be run at any one time and this is why the device is designed for embedded applications.
2. How do you debug these new instructions? Again, I don't read anything about the debugger in the article (part of the "Company's own C/C++ compiler"). I would expect the debugger to be critical to understanding whether or not the custom instructions work the way that is expected.
I would never consider a part which did not have simulator/emulator support and I don't know of too many people who would either, so this is a pretty important question.
3. Where is the data showing "performing encryption or digital video processing on blocks of data, can be executed in single clock cycles"?
In my experience, unless very small blocks of data are being processed, the biggest bottleneck is memory speed/latency, not the number of instruction cycles.
4. How much space is there in the microinstruction memory for the custom instruction definitions?
This has implications on the number of custom instructions that can be implemented along with the total number that can be implemented.
I can see some advantages to having custom instructions optimize specific operations in a single process embedded controller (such as the ones used in a router, in camera video compression and basic cyptography).
I don't see it being a significant threat to the current Xeon, Opteron or Itanium processored systems.
Before going ahead and spending an hour creating your masterpiece, wouldn't it be a good idea that you first check to see what the conditions are like?
Right now in Toronto (Eastern Daylight Savings Time) it is 11:30 AM and in Dublin it is 4:27:56 PM IST (Irish Standard Time?).
It might be a good idea to wait a few hours before submitting your project, designed to be intuitively obvious to any extraterrestrial beings that there is intelligent life on Earth...
Remlik wrote I always wondered why bombers leave these things behind in rather conspicious places for long periods of time. Its almost as if they want the bombs to be found before they detonate...seems like a short fuse and a decent hiding place would serve them better.
From what I understand, bombers are a criminal class unto themselves - Theodore Kaczynski (the Unabomber) actually fit the "profile" quite closely. They are generally very timid and afraid of confrontation (which is why they leave bombs) and if anything happens that makes them nervous, they tend to drop the bombs where they are and run. One of the things that make them nervous is the unstable nature of their bombs - many of them are built from match heads or other pyrotechnics that can be found in hardware stores and can't be traced back to them.
The meticulous planner with access to military grade hardware in fiction is just that; most bombers are paranoid loners that put bombs together out of a variety of random pieces.
A bit of a navel gazing question for you; what form do you think A.I. will take when somebody finally comes up with a program that is accepted as intelligent?
My own feeling is that the first A.I. program will simulate a simple life form (like a worm) instead of a highly complex and communicative form like humans. This goes against what Dr. Minsky believes A.I. should be, but I can't honestly believe that our first interaction with an intelligent mechanism would with something with similar capabilities to ourselves, but with something with the same mental capabilities and capacities as a bug.
The important aspects of Aritficial Intelligence will be making sense of its environments and learning from experience. To demonstrate that the Intelligence is learning is observing and testing the Intelligence's application of this knowledge.
I've been involved on the periphery of a not-so-local high school's (Rick Hansen Secondary School - Team 1241 "Force 6") development project and I'm disappointed in the extremely high cost of entry (ie to be registered and to get a kit), the sophistication of the projects as well as the other costs associated with it. It is essentially impossible to field a team for less than $35k CAN ($25k+ USD) to be successful. This includes money for the kits as well as travel expenses and, amazingly enough, promotional materials that are needed to ask for sponsorship funds.
The high cost of entry really bars schools from low-income inner city neighborhoods, which are the ones that would probably benefit the most from the experience. These schools also do not have contacts/parents in industry that could help as mentors and sponsors. This is probably the biggest issue I have with USFIRST right now.
The robot task is such that high school kids cannot work through them without substantial help from experienced engineers and what the kinds get out of the program (as well as put into it) depends primarily on how the sponsor engineers allow the kids to do. The best sponsors are high level advisors and make sure the kids plan out the designs themselves and help them think through the problems that they encounter rather than do the design themselves. I'm sure there are a lot of cases where the kids are barely able to play around with the robots before the competition because of the amount of time the sponsors put into the robots.
There is too much emphasis on the necessary fund raising. The Rick Hansen team had created a promotional DVD along with glossy brochures; there is an irony that these materials can be produced quite cheaply because they give the impression that the team has more money than they know what to do with.
Rather than limiting the kids to the materials supplied in the (incredibly expensive) kits, I would prefer seeing something where the bare minimum was provided by FIRST and the majority of parts were to be found at Home Depot/Digikey by the kids themselves. I think this would limit the price somewhat, would allow the kids to spend more time on design, building and experimenting (which is what FIRST should be all about anyway).
There should also be a restriction on how much the sponsors can do - clearly there are a lot of teams that benefit from corporate tool rooms with trained tool makers and do not rely on industrial arts rooms with the students learning how to machine parts on their own. To help enforce this, I believe that each team, to qualify must provide documentation on the robot to prove that the students were primarily responsible for the design and this documentation could be made available by USFIRST as guides for later teams.
Regardless of the warts, USFIRST is the best opportunity kids have to learn, design and compete with others. The events are amazing, fun and energetic experiences that are barely controlled chaos. The kids have a lot of fun, FIRST is a great way to build school spirit and it gives a few kids an opportunity to see if engineering/computer science is the way they want to go in life.
The worldwide production of optical discs is approximately 20 billion per year and optical discs are being adopted widely.
What is it minus AOL?
Extending this thread, it's too bad Sony didn't work on this with P The "Bounty" version of the AOL disk could pre-emptively clean up those annoying coffee drips and the "Charmin" version, well the AOL disks would finally actually be useful.
I have some experience with the type of lawyers IBM hires as litigators in social situations - I was involved as a witness in a product conterfiting suit about ten years ago.
To make me feel at ease, here in Toronto they took me out to dinner (Shopsey's at Yonge and Front) followed by a show at Second City.
Over dinner, the conversation ranged over such fascinating topics as the low percentage of both Canadians and Americans that could name all the judges on the countries supreme court, the need for all engineers to have mandatory training in evidence collection, documenting and testifying, explanation of the correct terms to use in the deposition process, which of the firm's offices were best for aspiring new hires and how much BMWs cost in Canada.
During the show, one of the lawyers actually started taking notes, recording what he thought were inappropriately used registered statements in the show - during the intermission they debated on the differences between Canadian and US law and whether or not the useage would be legal in the US.
At the end of the evening, they were happy that they were able to "sneak in a few minutes" talking about business, as this would allow them to claim the evening expenses as well as the time spent over dinner and the show as billable hours.
All in all, I found it to be a pretty traumatic evening.
At no time did I get the feeling that these people were normal human beings. On a personal level, they can't function with normal human beings. On a professional level, they are more than competent and although they will bleed you dry.
But, if they are working for you, I'm sure they'll devote more than 100% of their waking hours to your case. So, I guess looking at it from this dimension, they are excellent friends to have.
I've known quite a few "communications specialists" (I think the term "PR" has been out of vogue for 20+ years now) that have been embarrassed by what they have said; when it comes back to them. I would rephrase the basic statement as:
The most effective PR people know how to shade the truth/lie in such a way that the message, regardless of how misleading, cannot be challenged as being inaccurate.
An experienced communications specialist would come out with the statement:
"The Senator is taking a hiatus from active service to better understand how to reduce his own reliance on foreign products while minimizing any potential financial repercussions."
which is much better than saying:
"The Senator is drying out at the Betty Ford clinic after realizing that he can no longer afford to drink Chivas by the boatload."
Is it a co-incidence that you would mention Dr. Strangelove with Britain being the home of Jack The Ripper?
Where are all the mineshaft's to protect us from the Cobalt-Thorium G? And, where can I find the catalog so that I can choose the 10 woman I'm suposed to service?
Yeah, this is somewhat off topic, but I'm having a crappy day and any kind of diversion is a good thing.
People may listen to Orson Well's version once I year, but I always pull out my CDs of Jeff Wayne's musical version of "War of the Worlds" (with Richard Burton narrating).
Now, if Spielberg were to make a film of this version I would definitely go see it, even if T. Cruise was starring in it.
Slashdot Mirror
Actually, I would be more interested in finding out how Leia got away/wasn't known to Vader.
As to it not making sense for Luke going to Tatooine, I thought it actually did - wasn't Luke given to "Uncle Owen" (whom I presume is Anakin's half brother) to be raised?
Give Lucas some credit. He "Spoils" something that shouldn't be news, but holds back the piece that is actually more intriguing (at least I hope it is).
myke
I thought the correct quote was:
Those who can, do. Those who can't, teach.
Those who can't teach, teach gym.
(At least according to Woody Allen).
myke
the old "Wizard of Id" cartoon in which a group of protesters are imploring the king to give up animal testing.
His response was to ask for volunteers to take the animals' places.
myke
For years, I thought it stood for "Fucked Up Design" because it was always applied to Microsoft...
myke
the Second "Treehouse of Horror".
Bones Crushed... Organs leaking vital fluids... Loss of Appetite...
Or, my favourite:
It's so simple, I don't know why nobody has seen this before. The solution to Fermat's Last Theorem is....Gak
myke
"a college student"?
I bet you can find legions of them dumb enough to get the tat. I'm willing to bet that the winner will be the person willing to devote the most square footage for it.
myke
If Dell took on AMD it would probably end up killing AMD.
Intel's strength has always been its ability to drop prices for large customers to levels below that of its competitors. For AMD to become a Dell supplier, it would have to lower its prices to the point where they were not only unprofitable but probably bleeding money from an opened artery.
Intel might not want anything more than AMD to make an offer to Dell that they can't match. Even if it didn't kill AMD, it would put AMD in the place where Intel believes it belongs.
myke
To go, you follow the glow...
myke
I probably should have said MOT683xx, PowerPC or mobile Pentiums, which seem to be the primary choice for most high-end embedded applications that I am seeing, instead of the big three.
The CNet article was pretty vague and seemed to imply that the custom instruction capability provided better number crunching capabilities than these traditional server/workstation processors.
Thanx for pointing this out,
myke
Pretty skimpy blurb - I suspect that the product is either a) vapourware or b) a lot more limited than is discussed in the article.
From the article, I presume that the processor's microinstruction memory can be updated with special information embedded in the executable file. This is not as unique as you might think: virtually all Intel and AMD processors have the ability to have their microinstruction memory updated during the boot process - this is used up upload microinstruction updates/corrections without requiring a new chip. What's implied in the article's title (note, not in the text), is that this memory can be updated on the fly which, I presume, means that the instruction set changes for each process that is loaded.
The immediate questions that I have are:
1. How is this accomplished in a multi-process (ie Linux) operating environment where the op-codes given to the custom instructions in one process are the same as another? Does this mean that the microinstruction memory must be reset during task switch to avoid this problem? What is the task switch latency penalty for this process?
I suspect that only one process with custom instructions can be run at any one time and this is why the device is designed for embedded applications.
2. How do you debug these new instructions? Again, I don't read anything about the debugger in the article (part of the "Company's own C/C++ compiler"). I would expect the debugger to be critical to understanding whether or not the custom instructions work the way that is expected.
I would never consider a part which did not have simulator/emulator support and I don't know of too many people who would either, so this is a pretty important question.
3. Where is the data showing "performing encryption or digital video processing on blocks of data, can be executed in single clock cycles"?
In my experience, unless very small blocks of data are being processed, the biggest bottleneck is memory speed/latency, not the number of instruction cycles.
4. How much space is there in the microinstruction memory for the custom instruction definitions?
This has implications on the number of custom instructions that can be implemented along with the total number that can be implemented.
I can see some advantages to having custom instructions optimize specific operations in a single process embedded controller (such as the ones used in a router, in camera video compression and basic cyptography).
I don't see it being a significant threat to the current Xeon, Opteron or Itanium processored systems.
myke
Regardless who wrote it.
myke
Before going ahead and spending an hour creating your masterpiece, wouldn't it be a good idea that you first check to see what the conditions are like?
Right now in Toronto (Eastern Daylight Savings Time) it is 11:30 AM and in Dublin it is 4:27:56 PM IST (Irish Standard Time?).
It might be a good idea to wait a few hours before submitting your project, designed to be intuitively obvious to any extraterrestrial beings that there is intelligent life on Earth...
myke
Remlik wrote I always wondered why bombers leave these things behind in rather conspicious places for long periods of time. Its almost as if they want the bombs to be found before they detonate...seems like a short fuse and a decent hiding place would serve them better.
From what I understand, bombers are a criminal class unto themselves - Theodore Kaczynski (the Unabomber) actually fit the "profile" quite closely. They are generally very timid and afraid of confrontation (which is why they leave bombs) and if anything happens that makes them nervous, they tend to drop the bombs where they are and run. One of the things that make them nervous is the unstable nature of their bombs - many of them are built from match heads or other pyrotechnics that can be found in hardware stores and can't be traced back to them.
The meticulous planner with access to military grade hardware in fiction is just that; most bombers are paranoid loners that put bombs together out of a variety of random pieces.
myke
Hey Joanne,
A bit of a navel gazing question for you; what form do you think A.I. will take when somebody finally comes up with a program that is accepted as intelligent?
My own feeling is that the first A.I. program will simulate a simple life form (like a worm) instead of a highly complex and communicative form like humans. This goes against what Dr. Minsky believes A.I. should be, but I can't honestly believe that our first interaction with an intelligent mechanism would with something with similar capabilities to ourselves, but with something with the same mental capabilities and capacities as a bug.
The important aspects of Aritficial Intelligence will be making sense of its environments and learning from experience. To demonstrate that the Intelligence is learning is observing and testing the Intelligence's application of this knowledge.
What are your thoughts?
Thanx,
myke
Sadly, there's a lot of truth to this.
I've been involved on the periphery of a not-so-local high school's (Rick Hansen Secondary School - Team 1241 "Force 6") development project and I'm disappointed in the extremely high cost of entry (ie to be registered and to get a kit), the sophistication of the projects as well as the other costs associated with it. It is essentially impossible to field a team for less than $35k CAN ($25k+ USD) to be successful. This includes money for the kits as well as travel expenses and, amazingly enough, promotional materials that are needed to ask for sponsorship funds.
The high cost of entry really bars schools from low-income inner city neighborhoods, which are the ones that would probably benefit the most from the experience. These schools also do not have contacts/parents in industry that could help as mentors and sponsors. This is probably the biggest issue I have with USFIRST right now.
The robot task is such that high school kids cannot work through them without substantial help from experienced engineers and what the kinds get out of the program (as well as put into it) depends primarily on how the sponsor engineers allow the kids to do. The best sponsors are high level advisors and make sure the kids plan out the designs themselves and help them think through the problems that they encounter rather than do the design themselves. I'm sure there are a lot of cases where the kids are barely able to play around with the robots before the competition because of the amount of time the sponsors put into the robots.
There is too much emphasis on the necessary fund raising. The Rick Hansen team had created a promotional DVD along with glossy brochures; there is an irony that these materials can be produced quite cheaply because they give the impression that the team has more money than they know what to do with.
Rather than limiting the kids to the materials supplied in the (incredibly expensive) kits, I would prefer seeing something where the bare minimum was provided by FIRST and the majority of parts were to be found at Home Depot/Digikey by the kids themselves. I think this would limit the price somewhat, would allow the kids to spend more time on design, building and experimenting (which is what FIRST should be all about anyway).
There should also be a restriction on how much the sponsors can do - clearly there are a lot of teams that benefit from corporate tool rooms with trained tool makers and do not rely on industrial arts rooms with the students learning how to machine parts on their own. To help enforce this, I believe that each team, to qualify must provide documentation on the robot to prove that the students were primarily responsible for the design and this documentation could be made available by USFIRST as guides for later teams.
Regardless of the warts, USFIRST is the best opportunity kids have to learn, design and compete with others. The events are amazing, fun and energetic experiences that are barely controlled chaos. The kids have a lot of fun, FIRST is a great way to build school spirit and it gives a few kids an opportunity to see if engineering/computer science is the way they want to go in life.
myke
From the article:
The worldwide production of optical discs is approximately 20 billion per year and optical discs are being adopted widely.
What is it minus AOL?
Extending this thread, it's too bad Sony didn't work on this with P The "Bounty" version of the AOL disk could pre-emptively clean up those annoying coffee drips and the "Charmin" version, well the AOL disks would finally actually be useful.
myke
I would have though properly grounded tin-foil gloves would be more appropriate in light of this article.
You don't want anything travelling from your fingers through to the keyboard...
myke
I have one too. When I found out that it was rare, I immediately placed it in a plastic bag and bought another copy.
I priced it a few years ago and its not worth anything close to what you would think...
It is a nice keepsake tho,
myke
I have some experience with the type of lawyers IBM hires as litigators in social situations - I was involved as a witness in a product conterfiting suit about ten years ago.
To make me feel at ease, here in Toronto they took me out to dinner (Shopsey's at Yonge and Front) followed by a show at Second City.
Over dinner, the conversation ranged over such fascinating topics as the low percentage of both Canadians and Americans that could name all the judges on the countries supreme court, the need for all engineers to have mandatory training in evidence collection, documenting and testifying, explanation of the correct terms to use in the deposition process, which of the firm's offices were best for aspiring new hires and how much BMWs cost in Canada.
During the show, one of the lawyers actually started taking notes, recording what he thought were inappropriately used registered statements in the show - during the intermission they debated on the differences between Canadian and US law and whether or not the useage would be legal in the US.
At the end of the evening, they were happy that they were able to "sneak in a few minutes" talking about business, as this would allow them to claim the evening expenses as well as the time spent over dinner and the show as billable hours.
All in all, I found it to be a pretty traumatic evening.
At no time did I get the feeling that these people were normal human beings. On a personal level, they can't function with normal human beings. On a professional level, they are more than competent and although they will bleed you dry.
But, if they are working for you, I'm sure they'll devote more than 100% of their waking hours to your case. So, I guess looking at it from this dimension, they are excellent friends to have.
myke
I've known quite a few "communications specialists" (I think the term "PR" has been out of vogue for 20+ years now) that have been embarrassed by what they have said; when it comes back to them. I would rephrase the basic statement as:
The most effective PR people know how to shade the truth/lie in such a way that the message, regardless of how misleading, cannot be challenged as being inaccurate.
An experienced communications specialist would come out with the statement:
"The Senator is taking a hiatus from active service to better understand how to reduce his own reliance on foreign products while minimizing any potential financial repercussions."
which is much better than saying:
"The Senator is drying out at the Betty Ford clinic after realizing that he can no longer afford to drink Chivas by the boatload."
myke
Is it a co-incidence that you would mention Dr. Strangelove with Britain being the home of Jack The Ripper?
Where are all the mineshaft's to protect us from the Cobalt-Thorium G? And, where can I find the catalog so that I can choose the 10 woman I'm suposed to service?
Yeah, this is somewhat off topic, but I'm having a crappy day and any kind of diversion is a good thing.
myke
Hmmm... Clearly some testing is required.
Maybe if somebody could forward it, I could test it out by buying something that will prove that it is actually Gates' card.
I'm thinking that South Dakota should be adequate for this task.
myke
shoulders of giants."
And, with recent events, with all this weight on them they have been pushed into the Muck.
Obligatory Bundy quote:
"Everybody say 'Thanks Darl.'"
myke
I thought you were going to say that Michael Jackson was sued because the equipment used to build him used Linux.
myke
People may listen to Orson Well's version once I year, but I always pull out my CDs of Jeff Wayne's musical version of "War of the Worlds" (with Richard Burton narrating).
Now, if Spielberg were to make a film of this version I would definitely go see it, even if T. Cruise was starring in it.
myke