Via Technologies didn't purchase National Semiconductor; they purchased the majority of Cyrix PC processor intellectual properties from National Semiconductor. National retained some properties that they've used to show off at recent shows with their information appliance gizmos. Via did NOT (and COULD not) obtain the licensing agreements with Intel through this deal.
(Disclaimer: I work for National Semi, and all of this information is available in public press releases.)
Anyway, I'm a soon-to-be graduate of a Computer Engineering program, and I think your question is an important one. However, there doesn't seem to be one solid definition of Computer Engineering vs Computer Science, so I'm going to have to go on what I've seen.
At the school I go to, engineering students do more math than computer science students, hands-down. Calculus, Statistics, Differential Equations, and Linear Algebra are requirements; CS majors only need Calc and DiffEq's OR Linear Algebra OR Stats... While CompEng majors have to take the hardest classes of CS (Discrete Structures, Operating Systems, Software Engineering, etc) plus programming courses, plus all of our Electrical Engineering core classes except for field and wave theory...
It all depends on what you want to DO after school. I work mostly in math and numerical analysis. I know CE majors who are webmonkeys and CE majors who are ASIC designers... At least with our program, there's a lot of leeway. And I've worked jobs from microelectronics to contract software development, so (my bias here) I think engineering prepares you more broadly for a large array of jobs. Still, if you love to program, a CS major is (in many ways) easier, and gives much more programming practice...
For what it's worth, I think my experiences may even have been valuable enough to share..
I worked for [insert relatively well-known semiconductor manufacturing firm here]. Before going, I did as much homework as I could in that field to fill in what gaps I thought I could. As soon as I arrived and got a handle on what my position would entail, I set up for myself a list of things to be learned before I left. While my position wasn't particularly menial, I did end up with an awful lot of down time... (Lesson one: Most good work happens in teams, so plan work to fill time gaps while waiting for teammates to complete their segments)
During this down time, I sought out different people, from my section and from others, and tried to learn as much as I could about how the business worked, how this company in particular worked, how the industry worked, how machine X worked, etc. A co-op serves two real purposes: 1, the company gets a lower-paid employee to do many things a full-time employee could do at a somewhat more closely supervised level AND can evaluate this employee for potential full-time employment in the future; 2, the co-op can evaluate the company, industry, and specific position as a future career, and can (and is expected to) use this opportunity as not only a job, but a learning experience.
So, even though my primary job function was to work on a handful of control projects (which were important; DEFINITELY focus on your work first), I had opportunity to learn Perl and SQL, work with developing some data access software, learn to use a half-dozen major software packages, poke my nose into at least half of the unit processes in semiconductor manufacture, and learn bits and pieces of the hows and whys of computer integrated manufacturing. Not bad considering I was drawing a paycheck the whole time...
The upshot is.. er.. well, how many times in the future are you going to get paid to not know what you're doing? Take this as a learning experience, and get everything you can from the experience.
I didn't see any mention of this affecting Linux. Why should it have to? I agree with the notion that wild speculation may not be worthy of the front page, but "News for Nerds" can go out of the direct "affects Linux users" world...
I attend the University of Maine, and MANY courses here rely on FirstClass... We're one of the largest active user bases of the software (in fact, that's often a difficulty because we're punching it to its limits), and it's a reasonably nice conferencing/email/chat program. The most interesting part I've seen is that UM has a web gateway into FirstClass; while I'm not sure how universal that is, it provides clientless access to nearly every feature of the software..
It seems to me that the point of SETI@home is that even this (relatively) basic search is consuming enormous amounts of CPU cycles. Searching for signals more complicated than this is just not technologically feasible yet.
One practical use I can think of is for something other than a whole chip running at 500C.
Localized high currents (only example I can think of is electrostatic discharge, but that's because silicon can't handle intentional high currents) would become more attainable if the substrate could handle short-term or even moderately-long-lived high temperatures without breaking down. Maybe not heating the entire chip up, but specific junctions/switches/gates/drivers being able to survive bursts of high current could be VERY useful..
I'd recommend using some sort of hybrid system of neural networks techniques and fuzzy logic. In fact, there's no reason not to regard a fuzzy logic system as a neural network...
And even those may be too exotic for what you're trying to accomplish. How many sensors and the like are you thinking of working with? Unless you plan to monitor enormous amounts of environmental data (temperature, humidity, light, and motion in every room at a bare minimum), your system is doomed to being relatively "stupid" anyway. But back to the point...
A fuzzy expert system would be an easy way to build the base of an "intelligent" house controller, by establishing simple variables and rules (a la "if KITCHEN-LIGHT-LEVEL is LOW then KITCHEN-LIGHT is ON-FULL" etc). This is easy to establish initially, and gives you an "out of box" working system. Some sort of feedback technique (a la back-propagation, but modified to fit what you're trying to accomplish) can be used then with the "good/bad" remote mentioned above. Basically, you have two things you can tweak iteratively when "training" the system in this kind of a case. You could lower the weight that certain rules have, or you can modify fuzzy variable definitions (what exactly is a "LOW" LIGHT-LEVEL? maybe it's even specifically, what is a LOW KITCHEN-LIGHT-LEVEL?).
I think the idea is very workable, although I'm not sure how terrific it'd work IRL. Make sure to include an automatic override! (No, dammit, I know I'm not moving. That's because I'm READING...)
From the article
'' Noting the intensity of the Deca2000 contest and the determination of the players, Mruczek stated, "as you can see, each player's performance caused the other players to rise to the occasion. Without Robert Macauley getting the ball rolling with his score of 111,000, none of us would have reached our full potential." ''
I have no words.
I can agree with the majority of posts here, in that I don't see what Apple can possibly gain from this action. That aside, though, this employee DID sign a non-disclosure, and that IS a legally binding contract. It's not up to Joe Cubicle to decide what information is strategically important to the company, and even such a bizarre entity as Apple has to bring out the stick instead of the carrot if people are going to follow their rules.
But.. I know this is repetitive, but what can they possibly gain? Bad PR, and they can't sue for any significant amount of money (or they can sue, but if this is just a Joe Cubicle, they won't get it)...
Um, you're infringing on some of my patents here...
US 8,299,169 : Method for incorporation of all-capital letters in an online message (I already have an enormous infringment lawsuit going out against AOL on this one...)
US 8,551,212 : Method and apparatus for applying cynicism to an online post by simulating behavior to be parodied, then explaining the existence of said parody.
US 8,999,999 : Method of indicating exasperation by using hyperboly in reference to the number of occasions on which an exasperating event or class of events has happened.
They're all primarily concerned with methods of accelerating data bus transfer rates.
I don't know how I feel about this... Some of these patents are relatively new (this year, at least), so I see nothing wrong with taking this long to defend them. Some of them are from as far back as 1998, though, and it seems as though they waited a bit too long on that (although, perhaps they didn't notice until Voodoo5...)
Disclaimer: I actually believe in IP, IP law, patents, etc and am seeking a patent of my own currently
No, no, no. They expose the monkeys to radiation while they're flying, because somehow that's supposed to help them understand how human pilots might react to radiation. The monkeys die of radiation sickness.
But I see the issue here being the coming-of-age of the Internet. The facts of life are that many of these Internet companies are running out of VC, and tapping out their own reserves, and I certainly wouldn't loan a 'net venture money unless I believed they had some reasonable dream of ever repaying it!
Freedom of information is irrelevant if said information can't be provided. And a couple of *n*x/*BSD machines running on a cablemodem or DSL won't cut it for a serious Internet search service. We're now seeing the culling of the dotcoms, and the growing-up of the Internet business. Immature? Of course it's immature; the Internet business is only a handful of years old. The question to be faced now is, are these "herds of geeks" going to create functioning, lasting business models? The magic of Free information can only work if the individuals donating their tech wizardry can pay for computers, bandwidth, and enough take-out Szechuan!
From scanning the articles briefly, I'd have to guess that Microsoft won't be manufacturing these chips, but that a Microsoft subsidiary designed a chip and they'll probably have to contract out to manufacture them. It's not an enormous distinction, but it is an important one...
I agree with the point that SMP has to be programmed for, but I also think that most processor-intensive tasks aren't terrifically difficult to design for SMP. I don't have any experience with threaded *nix, but I've been a Windows app developer for several years (hey, it's a rocking student job) and I know that to code for SMP under NT, all you need to do is code a multi-threaded application and let the OS handle passing your threads around. In fact, the first project I worked on was designed to run in a threaded mode on a single-CPU system; when I worked on my dual-CPU at home, it simply flew! If *nix supports threading nearly as easily, I think it shouldn't be terrifically difficult to find well-coded multi-CPU apps.
Slashdot Mirror
Via Technologies didn't purchase National Semiconductor; they purchased the majority of Cyrix PC processor intellectual properties from National Semiconductor. National retained some properties that they've used to show off at recent shows with their information appliance gizmos. Via did NOT (and COULD not) obtain the licensing agreements with Intel through this deal.
(Disclaimer: I work for National Semi, and all of this information is available in public press releases.)
The only problem
With haiku is you just get
Anyway, I'm a soon-to-be graduate of a Computer Engineering program, and I think your question is an important one. However, there doesn't seem to be one solid definition of Computer Engineering vs Computer Science, so I'm going to have to go on what I've seen.
At the school I go to, engineering students do more math than computer science students, hands-down. Calculus, Statistics, Differential Equations, and Linear Algebra are requirements; CS majors only need Calc and DiffEq's OR Linear Algebra OR Stats... While CompEng majors have to take the hardest classes of CS (Discrete Structures, Operating Systems, Software Engineering, etc) plus programming courses, plus all of our Electrical Engineering core classes except for field and wave theory...
It all depends on what you want to DO after school. I work mostly in math and numerical analysis. I know CE majors who are webmonkeys and CE majors who are ASIC designers... At least with our program, there's a lot of leeway. And I've worked jobs from microelectronics to contract software development, so (my bias here) I think engineering prepares you more broadly for a large array of jobs. Still, if you love to program, a CS major is (in many ways) easier, and gives much more programming practice...
For what it's worth, I think my experiences may even have been valuable enough to share..
I worked for [insert relatively well-known semiconductor manufacturing firm here]. Before going, I did as much homework as I could in that field to fill in what gaps I thought I could. As soon as I arrived and got a handle on what my position would entail, I set up for myself a list of things to be learned before I left. While my position wasn't particularly menial, I did end up with an awful lot of down time... (Lesson one: Most good work happens in teams, so plan work to fill time gaps while waiting for teammates to complete their segments)
During this down time, I sought out different people, from my section and from others, and tried to learn as much as I could about how the business worked, how this company in particular worked, how the industry worked, how machine X worked, etc. A co-op serves two real purposes: 1, the company gets a lower-paid employee to do many things a full-time employee could do at a somewhat more closely supervised level AND can evaluate this employee for potential full-time employment in the future; 2, the co-op can evaluate the company, industry, and specific position as a future career, and can (and is expected to) use this opportunity as not only a job, but a learning experience.
So, even though my primary job function was to work on a handful of control projects (which were important; DEFINITELY focus on your work first), I had opportunity to learn Perl and SQL, work with developing some data access software, learn to use a half-dozen major software packages, poke my nose into at least half of the unit processes in semiconductor manufacture, and learn bits and pieces of the hows and whys of computer integrated manufacturing. Not bad considering I was drawing a paycheck the whole time...
The upshot is.. er.. well, how many times in the future are you going to get paid to not know what you're doing? Take this as a learning experience, and get everything you can from the experience.
Well, no kidding. There you go... Blame Canada!
whee.
I didn't see any mention of this affecting Linux. Why should it have to? I agree with the notion that wild speculation may not be worthy of the front page, but "News for Nerds" can go out of the direct "affects Linux users" world...
I attend the University of Maine, and MANY courses here rely on FirstClass... We're one of the largest active user bases of the software (in fact, that's often a difficulty because we're punching it to its limits), and it's a reasonably nice conferencing/email/chat program. The most interesting part I've seen is that UM has a web gateway into FirstClass; while I'm not sure how universal that is, it provides clientless access to nearly every feature of the software..
It was called ROB, and while it was neat, it was horrible. I mean, if you had two hours to blow per Gyromite level, it was amusing.
Now, far neater than ROB was the PowerGlove. At least I can hook that up to my PC and program the puppy (:
It seems to me that the point of SETI@home is that even this (relatively) basic search is consuming enormous amounts of CPU cycles. Searching for signals more complicated than this is just not technologically feasible yet.
Check out MindRover
One practical use I can think of is for something other than a whole chip running at 500C.
Localized high currents (only example I can think of is electrostatic discharge, but that's because silicon can't handle intentional high currents) would become more attainable if the substrate could handle short-term or even moderately-long-lived high temperatures without breaking down. Maybe not heating the entire chip up, but specific junctions/switches/gates/drivers being able to survive bursts of high current could be VERY useful..
I'd recommend using some sort of hybrid system of neural networks techniques and fuzzy logic. In fact, there's no reason not to regard a fuzzy logic system as a neural network...
And even those may be too exotic for what you're trying to accomplish. How many sensors and the like are you thinking of working with? Unless you plan to monitor enormous amounts of environmental data (temperature, humidity, light, and motion in every room at a bare minimum), your system is doomed to being relatively "stupid" anyway. But back to the point...
A fuzzy expert system would be an easy way to build the base of an "intelligent" house controller, by establishing simple variables and rules (a la "if KITCHEN-LIGHT-LEVEL is LOW then KITCHEN-LIGHT is ON-FULL" etc). This is easy to establish initially, and gives you an "out of box" working system. Some sort of feedback technique (a la back-propagation, but modified to fit what you're trying to accomplish) can be used then with the "good/bad" remote mentioned above. Basically, you have two things you can tweak iteratively when "training" the system in this kind of a case. You could lower the weight that certain rules have, or you can modify fuzzy variable definitions (what exactly is a "LOW" LIGHT-LEVEL? maybe it's even specifically, what is a LOW KITCHEN-LIGHT-LEVEL?).
I think the idea is very workable, although I'm not sure how terrific it'd work IRL. Make sure to include an automatic override! (No, dammit, I know I'm not moving. That's because I'm READING...)
"I'm sorry, I can't do that..."
From the article '' Noting the intensity of the Deca2000 contest and the determination of the players, Mruczek stated, "as you can see, each player's performance caused the other players to rise to the occasion. Without Robert Macauley getting the ball rolling with his score of 111,000, none of us would have reached our full potential." '' I have no words.
I can agree with the majority of posts here, in that I don't see what Apple can possibly gain from this action. That aside, though, this employee DID sign a non-disclosure, and that IS a legally binding contract. It's not up to Joe Cubicle to decide what information is strategically important to the company, and even such a bizarre entity as Apple has to bring out the stick instead of the carrot if people are going to follow their rules.
But.. I know this is repetitive, but what can they possibly gain? Bad PR, and they can't sue for any significant amount of money (or they can sue, but if this is just a Joe Cubicle, they won't get it)...
Um, you're infringing on some of my patents here...
US 8,299,169 : Method for incorporation of all-capital letters in an online message (I already have an enormous infringment lawsuit going out against AOL on this one...)
US 8,551,212 : Method and apparatus for applying cynicism to an online post by simulating behavior to be parodied, then explaining the existence of said parody.
US 8,999,999 : Method of indicating exasperation by using hyperboly in reference to the number of occasions on which an exasperating event or class of events has happened.
They're all primarily concerned with methods of accelerating data bus transfer rates.
I don't know how I feel about this... Some of these patents are relatively new (this year, at least), so I see nothing wrong with taking this long to defend them. Some of them are from as far back as 1998, though, and it seems as though they waited a bit too long on that (although, perhaps they didn't notice until Voodoo5...)
Disclaimer: I actually believe in IP, IP law, patents, etc and am seeking a patent of my own currently
Yes, it is bad for your 8 month old baby to be watching Sesame Street. He/She should be watching Teletubbies, complete with gay purple one!
No, no, no. They expose the monkeys to radiation while they're flying, because somehow that's supposed to help them understand how human pilots might react to radiation. The monkeys die of radiation sickness.
Sheesh. STUDY your 80's movies! (;
Freedom of information is irrelevant if said information can't be provided. And a couple of *n*x/*BSD machines running on a cablemodem or DSL won't cut it for a serious Internet search service. We're now seeing the culling of the dotcoms, and the growing-up of the Internet business. Immature? Of course it's immature; the Internet business is only a handful of years old. The question to be faced now is, are these "herds of geeks" going to create functioning, lasting business models? The magic of Free information can only work if the individuals donating their tech wizardry can pay for computers, bandwidth, and enough take-out Szechuan!
The "Tasmanian Tiger" is actually not a tiger at all. It's more similar to a wolf.
Confused me, too, but as long as it's a completely useless fact, it's worth memorizing since it could win me a game of Trivial Pursuit someday..
From scanning the articles briefly, I'd have to guess that Microsoft won't be manufacturing these chips, but that a Microsoft subsidiary designed a chip and they'll probably have to contract out to manufacture them. It's not an enormous distinction, but it is an important one...
I agree with the point that SMP has to be programmed for, but I also think that most processor-intensive tasks aren't terrifically difficult to design for SMP. I don't have any experience with threaded *nix, but I've been a Windows app developer for several years (hey, it's a rocking student job) and I know that to code for SMP under NT, all you need to do is code a multi-threaded application and let the OS handle passing your threads around. In fact, the first project I worked on was designed to run in a threaded mode on a single-CPU system; when I worked on my dual-CPU at home, it simply flew! If *nix supports threading nearly as easily, I think it shouldn't be terrifically difficult to find well-coded multi-CPU apps.