You wrote: In general, though, scientists don't care about integer results-- floating point is more important.
Ugh. This makes some assumptions that are incorrect. Several sciences make extremely heavy use of integer computations versus floating point. Whats more, these are the sciences that are driving large scale computational projects today, consuming cycles at an exponentially growing rate.
Specifically I am writing about bioinformatics and related information theoretic sciences, where data mining operations, large scale (GB -> TB) database comparisons are the norm and the coming norms.
A supercomputer has always been a nebulous term. Ask 10 people for a definition, and you will get 11 answers. Fundamentally there is no single quantifyable differentiating factor between a computer and a supercomputer. It is more of a subjective view than an objective firm quantitation.
My AMD Athlon based system can theoretically hit 3.6 GFLOPs. But can I really pull that much work through it? The G4's can hit about the same amount, theoretically. Does this make them supercomputers?
IMO, hell no. I have a simple definition (non-quantitative) of a supercomputer. A supercomputer allows you to tackle the large problems you need to handle rapidly, in order for you to effectively do your science. If your calculation runs fast on your pocket calculator, over the whole range of problem sizes you are willing to consider, then for your particular problem, your calculator is a supercomputer. If you need a massive supercluster of > 1000 processors to run your BLAST jobs in a reasonable period of time (this is my domain), then that defines your supercomputer for you. If the Sony PS2 vector units tied together on a network do wonders for your problem, well...
The definition is subjective. You cannot quantify it in any reasonable way. The work on clusters is giving the folks setting up export restrictions fits on what to do for these.
And finally, a science is NOT only floating point intensive in simulation codes. In fact the majority of codes using modern methods are more limited by memory latency and bandwidth than they are on the core FP system. The quality of the compilers matter far more than the FP ability of the chips.
Just my observation as a reformed computational physicist (learning to be a bioinformatics type). Aside from this, gross generalizations tend to be incorrect....
In the end, if you cant repeat it, it ain't so...
on
Excess Heat
·
· Score: 1
Bob Park of the American Physical Society has a book out called Voodoo Science that should be on the shelf of anyone seriously considering purchasing this other book. In short, the scientific community was, as it is supposed to be, skeptical.
Many many people attempted to reproduce Pon's and Fleishmann's work. Most everyone who did obtained the null result, that is, no differences observed from normal chemical phenomenon that were expected. Those who obtained a signal, went on to refine their measurements, only to discover later on that there was no signal.
So we have a problem. A massive PR campaign, and a money/intellectual property grab. And no observable effects. This is what Irving Langmuir once called Pathological Science in that it cannot be reproduced, and there are a few believers who will not give up their belief structure (for whatever reason).
So this brings us to today, and Bob Park's book. Bob goes through the long sad history of Cold Fusion. While there are those whom would like to see a redemption of the ideas, the fact remains that the experiments are not repeatable, and when looked at with the appropriate level of skepticism and review, the signals that are observed appear to fade into the background.
There is no bias against CF, there is a bias against bad science.
Many of the first experimenters with X-rays died due to the burns they received from the systems they built. X-rays are ionizing radiation (the worst sort of radiation hazard) and should be treated with tremendous care. Building your own X-ray machine is not a wise idea unless you are a medical physicist working on X-ray machine design.
The shielding issues, the collimation issues, and every other issue you can imagine would be problematic enough. Making it safe would be quite difficult as well.
XFS is optimised for dealing with streaming media, and so deals well with high IO
and large files.
This is not entirely true. XFS will stream media if you want using GRIO (guaranteed rate IO) features if you set your file system up that way, though I am not sure the Linux version will have that.
I would refer users to the website for a more comprehensive view of XFS. Basically XFS has been running for about 6 years on IRIX machines. It is a 64 bit file system, end to end. It is journaled. It is designed for speed, both at the OS level and at the hardware level (you can hit and sustain in excess of 97% of theoretical max drive performance in various cases with SCSI systems, and large block IO). It is designed so you can have millions of files in directories, with files in the petabyte size if needed.
Basically XFS is really one of the best file systems out there.
Saying it does streaming media is like saying Linux can be used via a vt100 emulator to edit files. It can do so much more, and it does it very well.
Combine XFS with a well designed volume manager, and you can have your file system saturate your IO bus. This is nice if you need lots of IO capability. XFS based filesystems (atop XLV) sustained 7 GB/s (thats gigabytes, not gigabits, per second) several years ago in a test, reading and writing to a single file. The limiting factor was the number of spindles one could attach to the machine. We used 864 if I remember the number correctly.
Many SSC opponents were naive in thinking that if the SSC were cancelled, the
funds would go to other disciplines. But that money only existed for one purpose
-- kill the project and you kill the money, too
This is incorrect. No one assumed that the money would flow to the other fields. What the SSC did effectively do was to suck the air (and money) from other projects though, as congress wanted to limit the money given to the scientists. What many SSC opponents knew was that the zero sum game of science funding was about to get tremendously worse due to the funding problems in the SSC. Why should the condensed matter physicists have to lose funding for their research just to let the SSC continue? I speak from experience, after watching the funding in the grant that I was working under get sucked dry as monies shifted around to bolster the SSC at the expense of other science.
That this ever happened in the first place (despite the protestations that it never would) was unfortunate. Unfortunately, my group was the norm and not the exception. The grant program managers explained to us what had happened.
You also wrote
Maybe, but are they similarly far less intrigued by black holes, wormholes, the
origins of the universe, superstrings, etc.?
I would expect that they are highly focussed on CJS (aka Mad Cow disease) and the issues surrounding the processing of food, food borne illness, how to prevent contamination, are they infected, what CJS actually is (a prion, or disease state of a protein, folded in a different conformation) than they are about black holes, superstring theory, and other things that aren't likely to kill them because they ate a bad burger.
You also write:
By the way, while I contest many of your points -- because I'm in high-energy
physics -- I will freely admit that the most interesting science book I've read in
years is Kauffman's At Home in the Universe on biology, biochemistry, evolution,
etc.
I personally encourage healthy debate. I had many such debates in grad school over this with friends in HEP and friends in my own field of computational condensed matter physics. One of the most fundamental points of such debates is can we afford to do the science, as well as can we afford not to. I personally argue that not all mountains should be climbed because we can (with a nearly infinite expenditure of energy), but we need to pick our mountains more carefully. My apparant focus on the biological aspects comes with a recent shift in career to a bioinformatics focus (which is a beautiful application of statistical mechanics).
Then again I also think that we (in the US) spend far too little money on basic research, and we treat our graduate students very much like slave labor (or indentured servants, or even middle class kids on the government dole). The investment in basic research almost always pays off (economic analysis seems to put the internal rates of return at ~30% or so... I wish my stocks did as well).
Several points (based upon a long exposure to high energy physics types):
HEP types generally think that their flavor of physics is the only interesting one. It isn't, and arguably it isn't very interesting beyond their rather small group of practitioners, and some folks playing with cosmology. Yet they are good at marketing themselves.
Other, arguably more relevant to larger number of researchers, humans, and public policy areas of physics (biophysics, geophysics, condensed matter physics) get a short shrift from the press (in large part due to the quality of the marketing done by the high energy types). I personally would much rather hear about how to make a cell membrane impermeable to various viral protein capsules than hear about CPT violation in some obscure never-to-be-found-in-nature resonance. No offense intended to the HEP folks, but relevance is not something that can be sold the way some in the HEP are selling it.
Some will take issue with these points. That is fine and arguably quite good. A healthy debate on the visibility of science is good for the country and the world.
Admittedly I am biased, as I am not an HEP person. My dissertation was on molecular dynamics studies of semiconductors. What I saw while in graduate school was projects like the SSC draining all the money out of science. When the NSF went before congress to ask for more money, congress balked, as it was after all funding this massive white albatross. Couldn't all scientists use it?
Arguably that was part of the problem, the lack of congressional education. The other part of the problem was the selling of HEP as Physics. It isn't, and it has demonstratably damaged the entire research community when that view was pushed. Yet there are still some that push it (see the article pointed to at the root of this thread).
No, the interesting problems in physics come from all the physics disciplines. The High Energy Physics (HEP) types still haven't learned that interesting physics to the public and interesting High Energy Physics are not identical. I would suspect that people in the U.K. are far less intrigued by the parity violation experiments as they are with protein folding diseases, specifically prions that are suspected to cause CJS. That is biophysics, biochemistry, biology, condensed matter physics, etc. The binding of small molecules to receptor sites to promote or inhibit life processes is arguably more interesting, and related to molecular dynamics, molecular recognition, statistical mechanics, etc. It is also the basis for drug discovery, without which we would have no pharmaceutical products.
As one of the presenters at the Indy SGI Linux university, I am disappointed to hear of this situation. That was one of the best LU's that I presented at, with one of the most energetic audiences.
If you email me offline at landman@sgi.com, I might be able to make things happen.
Large memory/IO capable systems running a standard OS, with standard tools, and a well known ABI/API
Highly scalable and reconfigurable modular computing. If you need more power, add more C-Bricks. If you need more IO, add more P or X bricks.
Large application base: Linux has captured mindshare of devolopers. Applications are being ported at a furious rate. It is becoming the dominant platform for software development (over Solaris and other similar Unices).
There are many other reasons as well, but frankly this type of machine is what many people have been waiting for. The total cost of ownership of all those Sun machines is far larger than of this machine. The performance of this machine is significantly ahead of your typical Sun machine.
One of the nicest features of this machine is that you can reconfigure it with a reboot (no recabling) to come up as a single large machine or as multiple medium machines, or many single machines. You can configure the computer to your needs, not shoehorn the problem to fit within a solaris boxes limitations. And unlike on other OSes, the partitioning actually works here.
High performance programming is likely to be dominated by Fortran/C/C++ for quite some time. There are some well known reasons why one needs to take any Java based numerical application with a truckload of salt.
Moreover, the write once run anywhere myth has been debunked in many places. JIT's in general cannot achieve what a good optimizing compiler can do (PC's generally do not have good optimizing compilers, though they are starting to appear on Linux platforms). A very good way to tell if your C++ compiler is poorly optimizing, is if your Java and your C++ application (identical ones, doing identical work) run at the same speed.
All that said, Java, should it ever become standardized, could be an interesting platform to work on such applications. It is not a computational powerhouse like Fortran, or a string processing powerhouse like Perl, but it has its uses. Once it is standardized, many of the design flaws (numerical, etc) can be fixed properly. After that it might start to get interesting as a distributed computational tool (more of a controlling tool than the tool for calculation).
Yes, photons can be emitted or absorbed. Photons carry momentum. Momentum of a photon is well known E=p*c with p the momentum of the photon. If you really require a newtonian description, an absorbed or emitted photon can impart a change of momentum to the source of the photon. Having that absorption or emission occur over a time interval could show a nice delta_P over a delta_time which has dimensions of force (also called an impulse/impact).
As for the rest of it, yes, you are wrong. Completely. Please review your copy of Sears, Zemansky, and Young, "University Physics" for more details.
You wrote: For example, the healthy blue glow *grin* seen in the water in fission reactors is due to electrons moving faster than light in the water.
Oy vey. Lets try this one again... correctly this time...
In water, with an index of refraction Nw (Nw=c/(velocity of light WITHIN water)), it is entirely possible that an electron exceeds the velocity of light WITHIN water (which is not c, but c/Nw). The (unhealthy) blue glow is Cerenkov radiation. The electrons do NOT go faster than c.
... I await some nutjob somewhere to start a petition mandating that Creationism (you know, that conservative brand of pseudoscience that *ALL* of the republican presidential nominees have indicated support for) be taught in our Michigan schools. That would naturally follow the censorship garbage being foisted only 150 miles from my house. Shortly after that I expect that the decalogue will be nailed to the front entrance of all schools, and that enforced religious classes (with a slight preference for some odd brand of christianity) will ensue.
No thank you.
This wave of pseudo-conservatism is as sickening as the halleluhya's proclaimed by the aforementioned candidates. Going back to the past "ideals" when we basically ignored problems/reality and replaced it with an Ozzie+Harriet+2.4 kids just doesn't work. It leads down a slippery slope that far too many of the conservative folks seem to fall down.
In government, you get what you pay for. You pay for internet access, and you get it. You pay for censorship and you will get it. The law of unintended consequences also indicates that when you go down that slippery slope, other very important literary and scientific works get caught up in the fray. Allowing this to proceed to its unnatural and ridiculous conclusion, and you find a Krystalnacht here in the US.
Censorship begets ignorance. Ignorance begets stupidity. Stupidity takes us backwards. That is where these pseudo-conservatives want to take us.
Look with great suspicion at anyone who proclaims themselves as here to help protect you from the evils that are out there. Their evils are likely not your evils. Their religion is likely not your religion, and their politics are likely not your politics. Despite this, they want to control your thought processes: by adjusting your education (ala creationism), by adjusting your reading (ala censorship in libraries/internet), etc.
Just remember precisely who is doing this to you, and do not forget for a second that it is their agenda and needs they are trying to put forth, not yours and your families.
Also remember that you may send these folks a message of GET OUT OF MY BACKYARD by voting down their proposals, their candidates, etc.
One of the more interesting aspects of any "new" theory is that it needs to supply some testable predictions in order for anyone to have a chance at falsifying it. As it turns out, what the experimentalists are looking for are one of two things, either agreement with the predicitions of the theory, or disagreements. Seems obvious, but it is important to note that agreement between measurement and theory does not prove the theory, yet disagreement between measurement and theory does disprove the theory.
Moreover, many people seem to have problems in general with the concept of a theory. I will give you an example of a theory: Sum over the all of the Forces = Sum over all the masses * accelleration of each mass This theory gives me predictive power, if I assign velocity to be the time integral of acceleration, and the position to be the time integral of velocity. Now using this theory (actually all of these theories) together, I get a simple predictive equation that relates position to velocity and acceleration (or in this case applied force per unit mass).
X=X0 + V0*t + (1/2)*(Total Force/Mass)*t*t
With this theoretical predictive equation, I can model position changes as a function of time. All the measurements I do may seem to indicate or in the parlance of the experimentalist, lend support to the theory, but no number of measurements can prove the theory correct. But I can always disprove it, simply by showing where it does not predict accurately.
People mistakenly believe that when something is a theory, it is not accepted/acceptable as a representation of reality. This is not correct. A theory is a predictive system which can and is tested, and is intrinsically falseafiable. That is the very nature and heart of science, that one can falsify a theory by finding out where it doesn't work. A theory that is not falseafiable is not scientific, it is a religious issue. This is precisely why on other threads the poor folks who posit creationism as an alternative to an evolutionary like process are beating their heads against a wall that they shouldn't be wasting their time on. Creationism is religion not science, evolution and related theories are testable, falseafiable, and modifiable.
When a theory is found to be false, some significant salvage work goes on to see what parts are useful and accurately predict observations, and where it fails. This post-mortem analysis is usually hidden from view of most of us, but practitioners of the science want to know what works and what doesn't. The parts the seem to work are usually kept for empirical reasons, as it is easier to explain regions of validity without subscribing to the theories deeper innards.
There is a great example of this. I just gave above a brief descrption of Newton's laws. (note: A law is in this case a widely accepted theory). That equation I gave above is used by millions of school kids in order to figure out where the car stops, or the cannon ball lands, etc.
That equation (actually the theory underneath it) is false. Or, stated another way, it has limited regions of applicability. One cannot and should not use that underlying theory to calculate Hohmann transfer orbits (minimum energy orbits for visiting our neighboring worlds). You can get most of the details right, but due to relativistic effects associated with curved space-time, a small error in calculation can spell disaster for the mission. NASA just had an issue just like this crop up (though theirs was more along the lines of showing the stupidity we have here in the US in still using the english measurement system... the English are not using it themselves!!!). You need to deal with relativistic effects, solar wind effects, magnetic effects, etc. That is to say that the theory is at best incomplete, and at worst wrong. On the other hand, no one is going to haul out the Einsteinian equations in order to calculate this orbit (or last time I checked they wouldn't do it), even though that theory (GR) is more accurate than Newton's theory of gravitation.
Ok, why have I said all this? Simple. As it turns out, many people have posited fractional quantum numbers for many years. All sorts of properties and measurements have been postulated, all sorts of behaviors have been theorized. To this date, in the system that is under study, none has been observed as far as I know. I do not know of any confirmation of this fellows work, but I do know of much disparagement of his methods (e.g. pointing out the flaws in his efforts).
If I were the average non-scientist on the street, I would be seriously skeptical about these things. There is always a romantic notion about some forgotten scientist somewhere finding something where no one else could, but largely this doesn't happen. I can tell you from personal experience that most scientists have enough built in skepticism to avoid the eureka factor if at all possible. It is better to be cautious than to be labelled a charlatan.
No body knows what this fellow found. However, he has been making claims of fractional quantum number stuff in hydrogen for quite some time. As it turns out, Hydrogen is the one element for which we have a very workable theory. This theory specifically precludes fractional quantum numbers. The reason for that is that they would be observable in the spectra of the atom. They are not observable. Either there is a complicated reason for non-observability, or they don't exist. My bet is on the latter.
As a side note, there is such a thing as a fractional quantum effect (the fractional quantum hall effect) though it is quite difficult to explain. There are also filling of fractional quantum Landau levels. There is nothing sacred about the integer versus the fraction in physics. We are not diophantinians.
My best guess is that this stuff is fake or another effect is poisoning his results. We see that all the time. That is precisely the purpose of the journal article, to call attention to a specific thing, and get smart people thinking about this. Why hasn't he published? I would be asking this.
I have been having this discussion with folks at the SC conference for several years. In short there are some bits of fundamental physics in the way of continuous shrinkage of components (and higher clock speeds as it turns out).
The punchline is this, and it is quite simple: You cannot continue to shrink feature size and maintain classical (e.g. transistor-like) behaviour out out your devices. You will eventually wind up in a quantum realm (length scales of a few angstroms per feature), and you have many difficulties to contend with in this realm which will prevent the behaviour you need in your devices from occuring.
Right now, the transistors are on the mesoscopic scale, so quantum effects are quite small. As you diminish the size of the feature (the wire, the transistor, etc), you are effectively replacing a classical object with a quantum mechanical object. This will occur when your length scale is a small multiple of the planck wavelength of your current carrying particle (electron or hole). You may introduce odd electronic levels in your bandstructure, which deplete/inject carriers causing extreme weirdness.
Of course, there is another more insidious problem. That of defects in the structures. You cannot (thanks to thermodynamics) grow a defect free material. Worse than that, if you ever figure out how to grow a defect free material, it will not remain defect free for long. This is important as structures become smaller, and defects make up larger and larger relative fractions of the materials.
Defects can add or remove charge carriers from the material, which may alter the electronic properties significantly. In the metallic portions of the chip, a defect in a small "wire" ("1D" quantum well) may significantly perterb the "flow" of current... Moreover, current in a quantum system is defined a little bit differently than in a classical system, so circuts that depend upon a classical flowing current are going to need to be redesigned.
In short, transfering from the mesoscopic scale to the quantum scale for circuits does not make a great deal of sense. It is slightly less sensible than running conventional silicon at GHz speeds. At these speeds, you need waveguides to carry current, not wires. This is a major problem that I have not seen any solution to yet.
Where are we going to get the great speedups we need in the future? Really you have few options.
1) a new material. The problem is that the economics of Silicon are hard to beat today.
2) a totally new computing mechanism. Today, computation depends upon the flow of fermions (electrons) which are spin 1/2 type particles. If we could figure out how to tap bosons (spin 1 particles, such as photons, etc) for the computational aspects, we could concievably exploit the boson wavefunction overlaps (bosons unlike fermions can have the same sets of quantum numbers simultaneously, and actually "be" in the same place at the same time...) This way, each "wire" could be carrying many signals simultaneously, and the computations that the CPU performs could concievably be massively parallel.
This is a great dream, but it will not likely happen in our lifetimes.
By denying the basis for current biological thought, and replacing it with religious mysticism, these folks are simply damaging their children. These children will believe as they are taught, regardless of the soundness or lack of soundness of the principles. Moreover this "educational" body has ignored the advice of others wiser than they, and blindly taken a step 100 years backwards.
To think that those who pay for their sins will be the very children they purport to help. These children will grow up, go to college, and have the rather stark reality of the world impressed upon them. I can tell you from direct experience that it is quite sad when a student stands up at the back of the introductory physics class you are teaching to correct you about things like the age of the universe (after all that light couldn't have been traveling for more than 6000 years), the big bang, etc. I felt rather bad for these folks, but as I have always said, you can drag a student towards knowledge, but you cannot make them think.
Colleges will look upon Kansas educated applicants with disfavor. After all, their preliminary biological education is incorrect, regardless of what the board tries to mandate. Businesses will be correctly suspicious of graduates from Kansas secondary institutions, as if the board mandated this, what other damage could they have done? Pi = 3? Pharmaceuticals, and related organizations will likely not wish to remain in a state that denys the very basis for the science that these organizations use to produce their products.
All of this because the board saw fit to put its own petty desires over the true educational needs of the students, ignoring the experts. By going purely political (and religious) on this, they have caused inestimable damage to the state. I wonder how many years and millions of tax dollars it will take to fix the problems that are caused by this?
Folks, I personally try to chose political candidates for office only after reading about their views and asking them questions. Democracy is hard folks, you have to fight to learn the relevant information. Voting a party line is a disaster waiting to happen, as has been demonstrated here. If you don't know or like any of the candidates, then dont vote for them. Straight party line tickets are an insult to the democratic ideal of the country, where each person in public office is there because they have the best interests of the public at heart, and will act accordingly. Straight party tickets mean you get ugly personal agendas from any crackpot group that is sneaky enough to get on the ticket.
Here you have a crackpot minority who has just leveraged a change that they should not have been able to do. Next will come prayer in schools. I do not want my daughter praying to Allah or Jesus in her school, as school is for education, and not religion. I control her access to that, and I do not want someone elses religion forced down hers or my throat. Nor should the people of Kansas allow this.
If some people wish to remain ignoramuses for life, you cannot necessarily coerce them into learning the error of their ways. There is no cure for stupidity and stubbornness. However, there is no reason that said idiots should ever have the opportunity to impose their own particular idiosyncratic belief systems upon others. That is the very heart of the democratic ideal, that crackpotism will largely be contained by exposing the ideas of the candidate to valid scrutiny. Voting a straight party ticket deprives you of that scrutiny, and allows the crackpotism to go unchecked. Folks, get out there and grill your candidates before voting. The damage they do may be to your own family.
There are some basic and rather fundamental things that are being missed in this discussion. Anecdotal evidence for abuse is easy to find, but the most important issues are being missed by focussing on individual cases.
That is, the companies primary motivation to bring on an H1B person is to lower their employment costs. Period. Moreover the apocryphal nonesense promolguated by many tech companies as to why they need more H1B's is absolutely insulting. The one and only reason they want H1Bs here is to lower their costs. That is to raise their profits.
While you may argue one way or another about how good this is, remember that a business is in business to make a profit for those that own the business. Period. The H1B mechanism allows them to lower costs (which count against profits).
What this creates unfortunately is a technological underclass of worker. That is, a person who is bound to a particular employer for a period of time via the H1B mechanism. The argument is that the H1B provides a method for the employer to pave the way for the H1B to enter the country as a contributing member supplying a badly needed skill.
The problem is that the skills are not all that badly needed. This is rather grossly misrepresented by our high tech lobbyists in DC. The H1B effect is not just to bind a set of workers more closely to an employer, but to also depress the market for wages for those not on the H1B, as they are competing for the same jobs as the H1B people.
So if you are starting to get the point I am trying to make, that employers are using this to help contain their costs, well, then you are on the right track of understanding what is going on here. I do not apologise for this policy, I abhor it. It is important to understand it regardless of how frustrating it is to deal with it.
I would prefer a free and open job market, but it doesn't exist. The high tech salaries are exploding far faster than profits and growth. This should be something that H1B people can take advantage of. Unfortunately with the program in place as it is now, they cannot. They are effectively excluded from this market. The market for H1B people is in fact an attempt to regulate and control the wages for the non-H1B types.
So we are left with a legal binding contract, placing a person into the bowels of an organization for a set period of time. Last I heard, we outlawed that practice in this country (US) 135 years ago.
Clearly the AC who posted this is more than a little biased. Certainly they are more than a little wrong.
SGI is a systems and solutions company. The article definitely says that SGI will not, quoting the poor misguided AC, "ditch Unix for NT". Quite emphatically, SGI is fully in support of what its customers request in terms of OSes and support.
What I have seen in recent months has been pure FUD spewing from little Sun driods and lackeys about SGI. It is amazing how many times I run into people who have heard this nonesense from a Sun lackey, and never questioned it. Well, I guess the old salt is true, you can fool most of the people most of the time.
Regardless, SGI is alive, and on the 21st of July we will hear how healthy it is, and possibly more about its focus and direction w.r.t OSes. This developer document looks like the first salve.
GaAs has some problems. But Low Temperature grown GaAs is promising. Basically an MBE grown material, you can engineer in whatever defects you like... or sort of... You can drive the material As rich easily (it prefers this), or Ga rich (harder). The difficult part is understanding the correct dopants for GaAs, as the defect behaviors are different. Interstitials are highly mobile in GaAs. Also, defect complexes are very important electrically to the material. The high temperature grown material requires an overpressure of As gas to grow in the requisite stoichiometry. The low temperature grown material requires an MBE setup which is difficult to use on a mass production line. GaAs is the material of the future, and it always will be.
Before any of you pull out your AFMs and start building these things, remember that thermodynamics is going to make life really hard on you, unless you can erect massive diffusion barriers (never mind electromigration, the "wires" will interdiffuse unless you keep the stuff real cold).
All you need are a few atoms to migrate in your 5 atom width device and voila, no more device. Migration barriers for self diffusion in Si tend to be only a few eV high at most (some barriers are around 1 eV if my memory serves me). The atoms will sample these barriers around 10**12 1/s, so it is quite likely that at room temperature you will see effects in a short period of time.
Does anyone remember the threading defects in blue solid state lasers when they first came out? They would work for only a few seconds, and then die from thermodynamic driven diffusion, threading defects (basically releaving strain in the lattice by displacing a line of atoms).
I suspect the 5 atom problems will be harder to overcome.
Folks, I think that there might be some substantial points that were glossed over by the article. The article makes charges of a cover up, supression, and all sorts of other nastiness. What it really doesn't cover is that the general scientific view at the moment that the experiment had fatal flaws due to very poor design/implementation, had shoddy data analysis and error identification, and a myriad of other things that need to be fixed long before the data will be acceptable by the scientific community.
Add to this an interpretation based upon a non-existant theory, and unsupportability due to missing information (e.g. where are those darned neutrons... should be streaming outta there like mad). Also add to this a non-conservation of a conserved quantity, and you get an audience with a high degree of scepticism which is properly placed.
Now go overboard, have some computer science professor from MIT patent a theory (almost unheard of) on how it works, have many others try the thing and start patenting techniques like mad, and what you have looks suspiciously like a money grab/gamble. If the thing is real, these people are rich. If it is not, it is just a reputation, which may be repaired.
This is not suppression. It was bad science. Was there some effect being witnessed? Who knows? The measurement process was sufficiently bad as to effectively nullify any data. This is what people complained about.
I remember in grad school (and after in my pseudo postdoc), I did months and months of calculations to test a theory we were putting forth. I had a nagging doubt about my results, but just wrote about it to my coworkers. Well, the referees picked it up and blasted us for it. Sure enough, my nagging doubt turned out to be problematic. Should I call up some (generalized science-illiterate) journalist type to complain bitterly of suppression or should I read their comments carefully and see why they rejected the paper?
The right answer is the latter. You know you have a crackpot when you see the former. I think this article is an example of the former. I could go off on a tangent about journalists, what you read/see on TV, and all that, but I will not. Someone claimed suppression. They missed the point, they didn't understand the peer review process. If you are going to claim something fantastic, you had better be prepared to defend your views with data, theories that explain something and match known existing observations, etc. Crying suppression is the best way to marginalize yourself.
I noticed a well defined lack of focus... (hows that for an oxymoron) to the subject matters. I kept bouncing between the computational materials talks, the computational biology talks, and the computational chemistry talks.
I should have submitted something in retrospect. I had 4-5 projects that I could have written about. At least this year I will be ready with 2 or 3.
If SGI is truly interested in pushing interoperability, what they ought to be pushing, IMO, is open-standards, cross-platform interoperability mechanisms such as CORBA, Java and LDAP
Several points. Corba is supported by several vendors commercial ORBs, as well as non-commercial OpenSource ORBs (Zope, et al). LDAP and crew are supported on IRIX as well, using commercial software or the LDAP software from UofM.
Java is a moving target. This is one of the reasons that many companies have banded together to define standards for things like real time and other related technologies. It appears that the hype-meisters of the technology seem to forget that every subtle tweak, every existing standard they ignore (OpenGL, Optimizer, VRML, etc) means longer lead times to real product. Further, the API/spec d'jour is generally frustrating, and it has given Java a reputation for write many times to run everywhere.
All that said, SGI is supporting 1.2 Java (relabeled as 2.0 by Sun).
I see a great deal of discussion on what Linux does, what IRIX does, and competition between them. I would ask that this be looked on somewhat differently. Linux on the SGI product line not only enhances the breadth of the product line, but it vastly increases the size of the market in which they will play.
There are some places where Linux cannot play today, and IRIX does a great job. There are places where IRIX is not as cost effective as Linux, and thus Linux is the appropriate platform. Where they overlap is not competition, but user choice. It is a function of application requirements, and end user preferences.
I am obviously biased as far as this goes, but I see this offering as a positive step for the market, it opens up many new choices that others previously had wished to remain closed.
hehe... I remember my a Nobel laureate of some note saying something about the sad state of affairs in physics, and catching hell for it.
I remember going to a talk by Lederman on some of this stuff in 92, and he gave the establishment non-answer.
I had to chose between the postdoc path and the feed the family path. I fed the family.
Part of the reason that I posted my tome was that I felt a responsibility to have the next round of cannon fodder look upon any such reports with open eyes. Why are there fewer students here? Is it a population cycle, or an out of phase supply and demand thing? Whether it was NAS or NSF (I think you are right actually...), that this was quoted verbatim to impressionable undergrads annoys me.
Be skeptical... hopefully the pragmatism of the next generation.
You write people who get ph.d. don't care about earning less money
I disagree. I care a great deal about earning less money, I want to earn more. I have a Ph.D. I did want to work as a scientist, but I did not want to be poor, and I have a family to feed.
If you look at companies like Vertex Pharmaceuticals, Inktomi, etc you will see groups of scientists who thought that they could make some money. Bunches of Ph.D types who figured that they could make a buck or two. Of course, I didn't mention my favorite example company and some of its founders.
Dr. Andy Grove is a physicist by training. As were Dr. Gordon Moore, and most of the original alumni of Intel. I don't think Andy is gonna trade his billions in for a little more knowledge about semiconductors and quantum wells.
Grad students want many things. Not all of them the same. Ph.D types want many things, not all of them the same. Recognition from ones peers is always nice, but it is difficult to translate that into feeding a hungry family on nearly slave wages.
If you ever have to make that choice, you will understand. I had to, and many others did as well. So we tell a cautionary tale to the next generation of hopefuls.
There is nothing so magical, so powerful, so incredible as knowing that you are the first person to see something... to know something... to pull back the shroud of uncertainty a little bit. To know that you are the first person there...
...but as I said, the pay sucks, and when you are combing your pockets and your car for change in order to buy next weeks groceries...
Slashdot Mirror
Ugh. This makes some assumptions that are incorrect. Several sciences make extremely heavy use of integer computations versus floating point. Whats more, these are the sciences that are driving large scale computational projects today, consuming cycles at an exponentially growing rate.
Specifically I am writing about bioinformatics and related information theoretic sciences, where data mining operations, large scale (GB -> TB) database comparisons are the norm and the coming norms.
A supercomputer has always been a nebulous term. Ask 10 people for a definition, and you will get 11 answers. Fundamentally there is no single quantifyable differentiating factor between a computer and a supercomputer. It is more of a subjective view than an objective firm quantitation.
My AMD Athlon based system can theoretically hit 3.6 GFLOPs. But can I really pull that much work through it? The G4's can hit about the same amount, theoretically. Does this make them supercomputers?
IMO, hell no. I have a simple definition (non-quantitative) of a supercomputer. A supercomputer allows you to tackle the large problems you need to handle rapidly, in order for you to effectively do your science. If your calculation runs fast on your pocket calculator, over the whole range of problem sizes you are willing to consider, then for your particular problem, your calculator is a supercomputer. If you need a massive supercluster of > 1000 processors to run your BLAST jobs in a reasonable period of time (this is my domain), then that defines your supercomputer for you. If the Sony PS2 vector units tied together on a network do wonders for your problem, well...
The definition is subjective. You cannot quantify it in any reasonable way. The work on clusters is giving the folks setting up export restrictions fits on what to do for these.
And finally, a science is NOT only floating point intensive in simulation codes. In fact the majority of codes using modern methods are more limited by memory latency and bandwidth than they are on the core FP system. The quality of the compilers matter far more than the FP ability of the chips.
Just my observation as a reformed computational physicist (learning to be a bioinformatics type). Aside from this, gross generalizations tend to be incorrect....
Many many people attempted to reproduce Pon's and Fleishmann's work. Most everyone who did obtained the null result, that is, no differences observed from normal chemical phenomenon that were expected. Those who obtained a signal, went on to refine their measurements, only to discover later on that there was no signal.
So we have a problem. A massive PR campaign, and a money/intellectual property grab. And no observable effects. This is what Irving Langmuir once called Pathological Science in that it cannot be reproduced, and there are a few believers who will not give up their belief structure (for whatever reason).
So this brings us to today, and Bob Park's book. Bob goes through the long sad history of Cold Fusion. While there are those whom would like to see a redemption of the ideas, the fact remains that the experiments are not repeatable, and when looked at with the appropriate level of skepticism and review, the signals that are observed appear to fade into the background.
There is no bias against CF, there is a bias against bad science.
Many of the first experimenters with X-rays died due to the burns they received from the systems they built. X-rays are ionizing radiation (the worst sort of radiation hazard) and should be treated with tremendous care. Building your own X-ray machine is not a wise idea unless you are a medical physicist working on X-ray machine design.
The shielding issues, the collimation issues, and every other issue you can imagine would be problematic enough. Making it safe would be quite difficult as well.
XFS is optimised for dealing with streaming media, and so deals well with high IO and large files.
This is not entirely true. XFS will stream media if you want using GRIO (guaranteed rate IO) features if you set your file system up that way, though I am not sure the Linux version will have that.
I would refer users to the website for a more comprehensive view of XFS. Basically XFS has been running for about 6 years on IRIX machines. It is a 64 bit file system, end to end. It is journaled. It is designed for speed, both at the OS level and at the hardware level (you can hit and sustain in excess of 97% of theoretical max drive performance in various cases with SCSI systems, and large block IO). It is designed so you can have millions of files in directories, with files in the petabyte size if needed.
Basically XFS is really one of the best file systems out there.
Saying it does streaming media is like saying Linux can be used via a vt100 emulator to edit files. It can do so much more, and it does it very well.
Combine XFS with a well designed volume manager, and you can have your file system saturate your IO bus. This is nice if you need lots of IO capability. XFS based filesystems (atop XLV) sustained 7 GB/s (thats gigabytes, not gigabits, per second) several years ago in a test, reading and writing to a single file. The limiting factor was the number of spindles one could attach to the machine. We used 864 if I remember the number correctly.
XFS scales provided the LVM scales.
Many SSC opponents were naive in thinking that if the SSC were cancelled, the funds would go to other disciplines. But that money only existed for one purpose -- kill the project and you kill the money, too
This is incorrect. No one assumed that the money would flow to the other fields. What the SSC did effectively do was to suck the air (and money) from other projects though, as congress wanted to limit the money given to the scientists. What many SSC opponents knew was that the zero sum game of science funding was about to get tremendously worse due to the funding problems in the SSC. Why should the condensed matter physicists have to lose funding for their research just to let the SSC continue? I speak from experience, after watching the funding in the grant that I was working under get sucked dry as monies shifted around to bolster the SSC at the expense of other science.
That this ever happened in the first place (despite the protestations that it never would) was unfortunate. Unfortunately, my group was the norm and not the exception. The grant program managers explained to us what had happened.
You also wrote
Maybe, but are they similarly far less intrigued by black holes, wormholes, the origins of the universe, superstrings, etc.?
I would expect that they are highly focussed on CJS (aka Mad Cow disease) and the issues surrounding the processing of food, food borne illness, how to prevent contamination, are they infected, what CJS actually is (a prion, or disease state of a protein, folded in a different conformation) than they are about black holes, superstring theory, and other things that aren't likely to kill them because they ate a bad burger. You also write:
By the way, while I contest many of your points -- because I'm in high-energy physics -- I will freely admit that the most interesting science book I've read in years is Kauffman's At Home in the Universe on biology, biochemistry, evolution, etc.
I personally encourage healthy debate. I had many such debates in grad school over this with friends in HEP and friends in my own field of computational condensed matter physics. One of the most fundamental points of such debates is can we afford to do the science, as well as can we afford not to. I personally argue that not all mountains should be climbed because we can (with a nearly infinite expenditure of energy), but we need to pick our mountains more carefully. My apparant focus on the biological aspects comes with a recent shift in career to a bioinformatics focus (which is a beautiful application of statistical mechanics).
Then again I also think that we (in the US) spend far too little money on basic research, and we treat our graduate students very much like slave labor (or indentured servants, or even middle class kids on the government dole). The investment in basic research almost always pays off (economic analysis seems to put the internal rates of return at ~30% or so... I wish my stocks did as well).
Some will take issue with these points. That is fine and arguably quite good. A healthy debate on the visibility of science is good for the country and the world.
Admittedly I am biased, as I am not an HEP person. My dissertation was on molecular dynamics studies of semiconductors. What I saw while in graduate school was projects like the SSC draining all the money out of science. When the NSF went before congress to ask for more money, congress balked, as it was after all funding this massive white albatross. Couldn't all scientists use it?
Arguably that was part of the problem, the lack of congressional education. The other part of the problem was the selling of HEP as Physics. It isn't, and it has demonstratably damaged the entire research community when that view was pushed. Yet there are still some that push it (see the article pointed to at the root of this thread).
No, the interesting problems in physics come from all the physics disciplines. The High Energy Physics (HEP) types still haven't learned that interesting physics to the public and interesting High Energy Physics are not identical. I would suspect that people in the U.K. are far less intrigued by the parity violation experiments as they are with protein folding diseases, specifically prions that are suspected to cause CJS. That is biophysics, biochemistry, biology, condensed matter physics, etc. The binding of small molecules to receptor sites to promote or inhibit life processes is arguably more interesting, and related to molecular dynamics, molecular recognition, statistical mechanics, etc. It is also the basis for drug discovery, without which we would have no pharmaceutical products.
If you email me offline at landman@sgi.com, I might be able to make things happen.
There are many other reasons as well, but frankly this type of machine is what many people have been waiting for. The total cost of ownership of all those Sun machines is far larger than of this machine. The performance of this machine is significantly ahead of your typical Sun machine.
One of the nicest features of this machine is that you can reconfigure it with a reboot (no recabling) to come up as a single large machine or as multiple medium machines, or many single machines. You can configure the computer to your needs, not shoehorn the problem to fit within a solaris boxes limitations. And unlike on other OSes, the partitioning actually works here.
Moreover, the write once run anywhere myth has been debunked in many places. JIT's in general cannot achieve what a good optimizing compiler can do (PC's generally do not have good optimizing compilers, though they are starting to appear on Linux platforms). A very good way to tell if your C++ compiler is poorly optimizing, is if your Java and your C++ application (identical ones, doing identical work) run at the same speed.
All that said, Java, should it ever become standardized, could be an interesting platform to work on such applications. It is not a computational powerhouse like Fortran, or a string processing powerhouse like Perl, but it has its uses. Once it is standardized, many of the design flaws (numerical, etc) can be fixed properly. After that it might start to get interesting as a distributed computational tool (more of a controlling tool than the tool for calculation).
As for the rest of it, yes, you are wrong. Completely. Please review your copy of Sears, Zemansky, and Young, "University Physics" for more details.
Oy vey.
Lets try this one again... correctly this time...
In water, with an index of refraction Nw (Nw=c/(velocity of light WITHIN water)), it is entirely possible that an electron exceeds the velocity of light WITHIN water (which is not c, but c/Nw). The (unhealthy) blue glow is Cerenkov radiation. The electrons do NOT go faster than c.
No thank you.
This wave of pseudo-conservatism is as sickening as the halleluhya's proclaimed by the aforementioned candidates. Going back to the past "ideals" when we basically ignored problems/reality and replaced it with an Ozzie+Harriet+2.4 kids just doesn't work. It leads down a slippery slope that far too many of the conservative folks seem to fall down.
In government, you get what you pay for. You pay for internet access, and you get it. You pay for censorship and you will get it. The law of unintended consequences also indicates that when you go down that slippery slope, other very important literary and scientific works get caught up in the fray. Allowing this to proceed to its unnatural and ridiculous conclusion, and you find a Krystalnacht here in the US.
Censorship begets ignorance. Ignorance begets stupidity. Stupidity takes us backwards. That is where these pseudo-conservatives want to take us.
Look with great suspicion at anyone who proclaims themselves as here to help protect you from the evils that are out there. Their evils are likely not your evils. Their religion is likely not your religion, and their politics are likely not your politics. Despite this, they want to control your thought processes: by adjusting your education (ala creationism), by adjusting your reading (ala censorship in libraries/internet), etc.
Just remember precisely who is doing this to you, and do not forget for a second that it is their agenda and needs they are trying to put forth, not yours and your families.
Also remember that you may send these folks a message of GET OUT OF MY BACKYARD by voting down their proposals, their candidates, etc.
Moreover, many people seem to have problems in general with the concept of a theory. I will give you an example of a theory: Sum over the all of the Forces = Sum over all the masses * accelleration of each mass This theory gives me predictive power, if I assign velocity to be the time integral of acceleration, and the position to be the time integral of velocity. Now using this theory (actually all of these theories) together, I get a simple predictive equation that relates position to velocity and acceleration (or in this case applied force per unit mass).
X=X0 + V0*t + (1/2)*(Total Force/Mass)*t*t
With this theoretical predictive equation, I can model position changes as a function of time. All the measurements I do may seem to indicate or in the parlance of the experimentalist, lend support to the theory, but no number of measurements can prove the theory correct. But I can always disprove it, simply by showing where it does not predict accurately.
People mistakenly believe that when something is a theory, it is not accepted/acceptable as a representation of reality. This is not correct. A theory is a predictive system which can and is tested, and is intrinsically falseafiable. That is the very nature and heart of science, that one can falsify a theory by finding out where it doesn't work. A theory that is not falseafiable is not scientific, it is a religious issue. This is precisely why on other threads the poor folks who posit creationism as an alternative to an evolutionary like process are beating their heads against a wall that they shouldn't be wasting their time on. Creationism is religion not science, evolution and related theories are testable, falseafiable, and modifiable.
When a theory is found to be false, some significant salvage work goes on to see what parts are useful and accurately predict observations, and where it fails. This post-mortem analysis is usually hidden from view of most of us, but practitioners of the science want to know what works and what doesn't. The parts the seem to work are usually kept for empirical reasons, as it is easier to explain regions of validity without subscribing to the theories deeper innards.
There is a great example of this. I just gave above a brief descrption of Newton's laws. (note: A law is in this case a widely accepted theory). That equation I gave above is used by millions of school kids in order to figure out where the car stops, or the cannon ball lands, etc.
That equation (actually the theory underneath it) is false. Or, stated another way, it has limited regions of applicability. One cannot and should not use that underlying theory to calculate Hohmann transfer orbits (minimum energy orbits for visiting our neighboring worlds). You can get most of the details right, but due to relativistic effects associated with curved space-time, a small error in calculation can spell disaster for the mission. NASA just had an issue just like this crop up (though theirs was more along the lines of showing the stupidity we have here in the US in still using the english measurement system... the English are not using it themselves!!!). You need to deal with relativistic effects, solar wind effects, magnetic effects, etc. That is to say that the theory is at best incomplete, and at worst wrong. On the other hand, no one is going to haul out the Einsteinian equations in order to calculate this orbit (or last time I checked they wouldn't do it), even though that theory (GR) is more accurate than Newton's theory of gravitation.
Ok, why have I said all this? Simple. As it turns out, many people have posited fractional quantum numbers for many years. All sorts of properties and measurements have been postulated, all sorts of behaviors have been theorized. To this date, in the system that is under study, none has been observed as far as I know. I do not know of any confirmation of this fellows work, but I do know of much disparagement of his methods (e.g. pointing out the flaws in his efforts).
If I were the average non-scientist on the street, I would be seriously skeptical about these things. There is always a romantic notion about some forgotten scientist somewhere finding something where no one else could, but largely this doesn't happen. I can tell you from personal experience that most scientists have enough built in skepticism to avoid the eureka factor if at all possible. It is better to be cautious than to be labelled a charlatan.
No body knows what this fellow found. However, he has been making claims of fractional quantum number stuff in hydrogen for quite some time. As it turns out, Hydrogen is the one element for which we have a very workable theory. This theory specifically precludes fractional quantum numbers. The reason for that is that they would be observable in the spectra of the atom. They are not observable. Either there is a complicated reason for non-observability, or they don't exist. My bet is on the latter.
As a side note, there is such a thing as a fractional quantum effect (the fractional quantum hall effect) though it is quite difficult to explain. There are also filling of fractional quantum Landau levels. There is nothing sacred about the integer versus the fraction in physics. We are not diophantinians.
My best guess is that this stuff is fake or another effect is poisoning his results. We see that all the time. That is precisely the purpose of the journal article, to call attention to a specific thing, and get smart people thinking about this. Why hasn't he published? I would be asking this.
Be smart. Be skeptical.
I have been having this discussion with folks at the SC conference for several years. In short there are some bits of fundamental physics in the way of continuous shrinkage of components (and higher clock speeds as it turns out).
The punchline is this, and it is quite simple: You cannot continue to shrink feature size and maintain classical (e.g. transistor-like) behaviour out out your devices. You will eventually wind up in a quantum realm (length scales of a few angstroms per feature), and you have many difficulties to contend with in this realm which will prevent the behaviour you need in your devices from occuring.
Right now, the transistors are on the mesoscopic scale, so quantum effects are quite small. As you diminish the size of the feature (the wire, the transistor, etc), you are effectively replacing a classical object with a quantum mechanical object. This will occur when your length scale is a small multiple of the planck wavelength of your current carrying particle (electron or hole). You may introduce odd electronic levels in your bandstructure, which deplete/inject carriers causing extreme weirdness.
Of course, there is another more insidious problem. That of defects in the structures. You cannot (thanks to thermodynamics) grow a defect free material. Worse than that, if you ever figure out how to grow a defect free material, it will not remain defect free for long. This is important as structures become smaller, and defects make up larger and larger relative fractions of the materials.
Defects can add or remove charge carriers from the material, which may alter the electronic properties significantly. In the metallic portions of the chip, a defect in a small "wire" ("1D" quantum well) may significantly perterb the "flow" of current... Moreover, current in a quantum system is defined a little bit differently than in a classical system, so circuts that depend upon a classical flowing current are going to need to be redesigned.
In short, transfering from the mesoscopic scale to the quantum scale for circuits does not make a great deal of sense. It is slightly less sensible than running conventional silicon at GHz speeds. At these speeds, you need waveguides to carry current, not wires. This is a major problem that I have not seen any solution to yet.
Where are we going to get the great speedups we need in the future? Really you have few options.
1) a new material. The problem is that the economics of Silicon are hard to beat today.
2) a totally new computing mechanism. Today, computation depends upon the flow of fermions (electrons) which are spin 1/2 type particles. If we could figure out how to tap bosons (spin 1 particles, such as photons, etc) for the computational aspects, we could concievably exploit the boson wavefunction overlaps (bosons unlike fermions can have the same sets of quantum numbers simultaneously, and actually "be" in the same place at the same time...) This way, each "wire" could be carrying many signals simultaneously, and the computations that the CPU performs could concievably be massively parallel.
This is a great dream, but it will not likely happen in our lifetimes.
...it isn't any more true.
By denying the basis for current biological thought, and replacing it with religious mysticism, these folks are simply damaging their children. These children will believe as they are taught, regardless of the soundness or lack of soundness of the principles. Moreover this "educational" body has ignored the advice of others wiser than they, and blindly taken a step 100 years backwards.
To think that those who pay for their sins will be the very children they purport to help. These children will grow up, go to college, and have the rather stark reality of the world impressed upon them. I can tell you from direct experience that it is quite sad when a student stands up at the back of the introductory physics class you are teaching to correct you about things like the age of the universe (after all that light couldn't have been traveling for more than 6000 years), the big bang, etc. I felt rather bad for these folks, but as I have always said, you can drag a student towards knowledge, but you cannot make them think.
Colleges will look upon Kansas educated applicants with disfavor. After all, their preliminary biological education is incorrect, regardless of what the board tries to mandate. Businesses will be correctly suspicious of graduates from Kansas secondary institutions, as if the board mandated this, what other damage could they have done? Pi = 3? Pharmaceuticals, and related organizations will likely not wish to remain in a state that denys the very basis for the science that these organizations use to produce their products.
All of this because the board saw fit to put its own petty desires over the true educational needs of the students, ignoring the experts. By going purely political (and religious) on this, they have caused inestimable damage to the state. I wonder how many years and millions of tax dollars it will take to fix the problems that are caused by this?
Folks, I personally try to chose political candidates for office only after reading about their views and asking them questions. Democracy is hard folks, you have to fight to learn the relevant information. Voting a party line is a disaster waiting to happen, as has been demonstrated here. If you don't know or like any of the candidates, then dont vote for them. Straight party line tickets are an insult to the democratic ideal of the country, where each person in public office is there because they have the best interests of the public at heart, and will act accordingly. Straight party tickets mean you get ugly personal agendas from any crackpot group that is sneaky enough to get on the ticket.
Here you have a crackpot minority who has just leveraged a change that they should not have been able to do. Next will come prayer in schools. I do not want my daughter praying to Allah or Jesus in her school, as school is for education, and not religion. I control her access to that, and I do not want someone elses religion forced down hers or my throat. Nor should the people of Kansas allow this.
If some people wish to remain ignoramuses for life, you cannot necessarily coerce them into learning the error of their ways. There is no cure for stupidity and stubbornness. However, there is no reason that said idiots should ever have the opportunity to impose their own particular idiosyncratic belief systems upon others. That is the very heart of the democratic ideal, that crackpotism will largely be contained by exposing the ideas of the candidate to valid scrutiny. Voting a straight party ticket deprives you of that scrutiny, and allows the crackpotism to go unchecked. Folks, get out there and grill your candidates before voting. The damage they do may be to your own family.
Folks:
There are some basic and rather fundamental things that are being missed in this discussion. Anecdotal evidence for abuse is easy to find, but the most important issues are being missed by focussing on individual cases.
That is, the companies primary motivation to bring on an H1B person is to lower their employment costs. Period. Moreover the apocryphal nonesense promolguated by many tech companies as to why they need more H1B's is absolutely insulting. The one and only reason they want H1Bs here is to lower their costs. That is to raise their profits.
While you may argue one way or another about how good this is, remember that a business is in business to make a profit for those that own the business. Period. The H1B mechanism allows them to lower costs (which count against profits).
What this creates unfortunately is a technological underclass of worker. That is, a person who is bound to a particular employer for a period of time via the H1B mechanism. The argument is that the H1B provides a method for the employer to pave the way for the H1B to enter the country as a contributing member supplying a badly needed skill.
The problem is that the skills are not all that badly needed. This is rather grossly misrepresented by our high tech lobbyists in DC. The H1B effect is not just to bind a set of workers more closely to an employer, but to also depress the market for wages for those not on the H1B, as they are competing for the same jobs as the H1B people.
So if you are starting to get the point I am trying to make, that employers are using this to help contain their costs, well, then you are on the right track of understanding what is going on here. I do not apologise for this policy, I abhor it. It is important to understand it regardless of how frustrating it is to deal with it.
I would prefer a free and open job market, but it doesn't exist. The high tech salaries are exploding far faster than profits and growth. This should be something that H1B people can take advantage of. Unfortunately with the program in place as it is now, they cannot. They are effectively excluded from this market. The market for H1B people is in fact an attempt to regulate and control the wages for the non-H1B types.
So we are left with a legal binding contract, placing a person into the bowels of an organization for a set period of time. Last I heard, we outlawed that practice in this country (US) 135 years ago.
Apparantly there are loopholes to be closed.
Clearly the AC who posted this is more than a little biased. Certainly they are more than a little wrong.
SGI is a systems and solutions company. The article definitely says that SGI will not, quoting the poor misguided AC, "ditch Unix for NT". Quite emphatically, SGI is fully in support of what its customers request in terms of OSes and support.
What I have seen in recent months has been pure FUD spewing from little Sun driods and lackeys about SGI. It is amazing how many times I run into people who have heard this nonesense from a Sun lackey, and never questioned it. Well, I guess the old salt is true, you can fool most of the people most of the time.
Regardless, SGI is alive, and on the 21st of July we will hear how healthy it is, and possibly more about its focus and direction w.r.t OSes. This developer document looks like the first salve.
GaAs has some problems. But Low Temperature grown GaAs is promising. Basically an MBE grown material, you can engineer in whatever defects you like... or sort of... You can drive the material As rich easily (it prefers this), or Ga rich (harder). The difficult part is understanding the correct dopants for GaAs, as the defect behaviors are different. Interstitials are highly mobile in GaAs. Also, defect complexes are very important electrically to the material. The high temperature grown material requires an overpressure of As gas to grow in the requisite stoichiometry. The low temperature grown material requires an MBE setup which is difficult to use on a mass production line. GaAs is the material of the future, and it always will be.
Before any of you pull out your AFMs and start building these things, remember that thermodynamics is going to make life really hard on you, unless you can erect massive diffusion barriers (never mind electromigration, the "wires" will interdiffuse unless you keep the stuff real cold).
All you need are a few atoms to migrate in your 5 atom width device and voila, no more device. Migration barriers for self diffusion in Si tend to be only a few eV high at most (some barriers are around 1 eV if my memory serves me). The atoms will sample these barriers around 10**12 1/s, so it is quite likely that at room temperature you will see effects in a short period of time.
Does anyone remember the threading defects in blue solid state lasers when they first came out? They would work for only a few seconds, and then die from thermodynamic driven diffusion, threading defects (basically releaving strain in the lattice by displacing a line of atoms).
I suspect the 5 atom problems will be harder to overcome.
Folks, I think that there might be some substantial points that were glossed over by the article. The article makes charges of a cover up, supression, and all sorts of other nastiness. What it really doesn't cover is that the general scientific view at the moment that the experiment had fatal flaws due to very poor design/implementation, had shoddy data analysis and error identification, and a myriad of other things that need to be fixed long before the data will be acceptable by the scientific community.
Add to this an interpretation based upon a non-existant theory, and unsupportability due to missing information (e.g. where are those darned neutrons... should be streaming outta there like mad). Also add to this a non-conservation of a conserved quantity, and you get an audience with a high degree of scepticism which is properly placed.
Now go overboard, have some computer science professor from MIT patent a theory (almost unheard of) on how it works, have many others try the thing and start patenting techniques like mad, and what you have looks suspiciously like a money grab/gamble. If the thing is real, these people are rich. If it is not, it is just a reputation, which may be repaired.
This is not suppression. It was bad science. Was there some effect being witnessed? Who knows? The measurement process was sufficiently bad as to effectively nullify any data. This is what people complained about.
I remember in grad school (and after in my pseudo postdoc), I did months and months of calculations to test a theory we were putting forth. I had a nagging doubt about my results, but just wrote about it to my coworkers. Well, the referees picked it up and blasted us for it. Sure enough, my nagging doubt turned out to be problematic. Should I call up some (generalized science-illiterate) journalist type to complain bitterly of suppression or should I read their comments carefully and see why they rejected the paper?
The right answer is the latter. You know you have a crackpot when you see the former. I think this article is an example of the former. I could go off on a tangent about journalists, what you read/see on TV, and all that, but I will not. Someone claimed suppression. They missed the point, they didn't understand the peer review process. If you are going to claim something fantastic, you had better be prepared to defend your views with data, theories that explain something and match known existing observations, etc. Crying suppression is the best way to marginalize yourself.
I noticed a well defined lack of focus... (hows that for an oxymoron) to the subject matters. I kept bouncing between the computational materials talks, the computational biology talks, and the computational chemistry talks.
I should have submitted something in retrospect. I had 4-5 projects that I could have written about. At least this year I will be ready with 2 or 3.
If SGI is truly interested in pushing interoperability, what they ought
to be pushing, IMO, is open-standards, cross-platform interoperability
mechanisms such as CORBA, Java and LDAP
Several points. Corba is supported by several vendors commercial ORBs, as well as non-commercial OpenSource ORBs (Zope, et al). LDAP and crew are supported on IRIX as well, using commercial software or the LDAP software from UofM.
Java is a moving target. This is one of the reasons that many companies have banded together to define standards for things like real time and other related technologies. It appears that the hype-meisters of the technology seem to forget that every subtle tweak, every existing standard they ignore (OpenGL, Optimizer, VRML, etc) means longer lead times to real product. Further, the API/spec d'jour is generally frustrating, and it has given Java a reputation for write many times to run everywhere.
All that said, SGI is supporting 1.2 Java (relabeled as 2.0 by Sun).
Folks:
I see a great deal of discussion on what Linux does, what IRIX does, and competition between them. I would ask that this be looked on somewhat differently. Linux on the SGI product line not only enhances the breadth of the product line, but it vastly increases the size of the market in which they will play.
There are some places where Linux cannot play today, and IRIX does a great job. There are places where IRIX is not as cost effective as Linux, and thus Linux is the appropriate platform. Where they overlap is not competition, but user choice. It is a function of application requirements, and end user preferences.
I am obviously biased as far as this goes, but I see this offering as a positive step for the market, it opens up many new choices that others previously had wished to remain closed.
hehe... I remember my a Nobel laureate of some note saying something about the sad state of affairs in physics, and catching hell for it.
I remember going to a talk by Lederman on some of this stuff in 92, and he gave the establishment non-answer.
I had to chose between the postdoc path and the feed the family path. I fed the family.
Part of the reason that I posted my tome was that I felt a responsibility to have the next round of cannon fodder look upon any such reports with open eyes. Why are there fewer students here? Is it a population cycle, or an out of phase supply and demand thing? Whether it was NAS or NSF (I think you are right actually...), that this was quoted verbatim to impressionable undergrads annoys me.
Be skeptical... hopefully the pragmatism of the next generation.
You write people who get ph.d. don't care about earning less money
I disagree. I care a great deal about earning less money, I want to earn more. I have a Ph.D. I did want to work as a scientist, but I did not want to be poor, and I have a family to feed.
If you look at companies like Vertex Pharmaceuticals, Inktomi, etc you will see groups of scientists who thought that they could make some money. Bunches of Ph.D types who figured that they could make a buck or two. Of course, I didn't mention my favorite example company and some of its founders.
Dr. Andy Grove is a physicist by training. As were Dr. Gordon Moore, and most of the original alumni of Intel. I don't think Andy is gonna trade his billions in for a little more knowledge about semiconductors and quantum wells.
Grad students want many things. Not all of them the same. Ph.D types want many things, not all of them the same. Recognition from ones peers is always nice, but it is difficult to translate that into feeding a hungry family on nearly slave wages.
If you ever have to make that choice, you will understand. I had to, and many others did as well. So we tell a cautionary tale to the next generation of hopefuls.
There is nothing so magical, so powerful, so incredible as knowing that you are the first person to see something... to know something... to pull back the shroud of uncertainty a little bit. To know that you are the first person there...
...but as I said, the pay sucks, and when you are combing your pockets and your car for change in order to buy next weeks groceries...
Been there, done that.... never ever again.