17 posts • joined 2 Feb 2009
Shades of STRETCH
This smells like the IBM 7030. Sometimes project managers just try to reach too far. The 7030 was a beautiful design which had elements of every modern supercomputer but it was just too complex to get the clock up to speed and it was too costly to manufacture. Seymour ate their lunch with the simple CDC 6600. That translates to modern chip design. If you can't get the clock up to expectations you need to go back to the drawing board. I would say that applies to Itanium. Any CPU that needs that much cache and still can't break the sound barrier on clock rate as well as being delayed for so long doesn't smell of success.
I'm putting on my raincoat before the black rain falls.
This is bollocks. Did anyone ever hear of size_t n? These people are idiots. The problem is that the coders don't bother to check the size in the code relative to the buffer. Banning memcpy is not going to do a damn thing to make software more secure. What we need to do is ban stupid developers who don't bother to size check data before they move it. It's rather hard to imagine how you can create a function that can read the mind of the developer and also determine how much virtual memory is available at a given moment. If you could do that the computer could just develop its own code.
One has to wonder...
After all the various nuclear accidents that have more or less flooded the whole stratosphere with radiation how do these self appointed gourmet boffins claim to be able to pick out these things? Next thing you know they'll spot Sputnik or Mir flying over the distillery. So, they can tell the difference between Bikini and Chernobyl? If they truly can I guess it's a whole new way to market vintage whiskey to yuppie idiots with more money than sense. You can glow in the dark with your favorite flavor of nuclear fallout.
Hmm, I think I'll brew up a batch of microbrew beer aged with spent nuclear fuel rods.
I think HP will change course
Let's face it. The Itanic is sinking. HP would be insane not to port HP-UX to x64. The Nehalem is a killer architecture and Tukwila is yesterday's news long overdue. The future of SPARC is largely up to Fujitsu. As I see things though once the eight core Nehalem EX comes out a 64 socket server running that chip will probably blow even a Cray off the planet. And if IBM are smart they will be the ones to sell it. They don't even need to port AIX to it. They already have Linux and WebSphere. IBM can still keep their increasingly POWER based i and z series markets, but we all know the killer app of today is the front end web not the DBMS, which IBM also has covered. So they can let Microsoft deal with the OS. But so does HP with their alliance with Oracle. So it really depends on who beats the other to the punch. In my view HP's at least publically stated unwillingness to port HP-UX to Nehalem Xeon servers is a mistake. Of course HP has often concealed business strategies so they could spring a new version of HP-UX or even Linux on Nehalem at the last minute just to upset the market. A company that produces a pink digital camera branded by Gwen Stefani is capable of anything.
The downside to electronic medical records
The University of Virginia (not where these records were stolen, AFAIK) were one of the pioneers in electronic online medical records but this just goes to show that the value is only as good as the dedication of the people who operate the system. Indeed even the Veterans Administration in Washington, D.C. were early pioneers in online radiology image availability. There was a lot more emphasis at UVA on automation then on security or even quality of care. They had an automated patient record, but they still seemed to be able to send home babies with the wrong mothers or have epidemics of VRE or MRSA in the medical center. To their credit they did address these issues, but this just goes to show that throwing technology at something or automating it doesn't fix anything if all you're doing is automating a broken system.
Oracle, Sun and HP
I see IBM as the loser in this. Oracle will likely continue to partner with HP. They would be insane not to. Oracle likes to get paid for commercial use of their products, but I've found them very reasonable to work with. They provided me with an evaluation copy of their software to prototype a recent project and even set up a conference with four of their top engineers when we ran into problems. Oracle also gives away free copies of their enterprise database for noncommercial use. I've worked with them since 1980 and while some of their HR practices may be unduly harsh, if you're a customer they're generally good to work with in my experience. It makes sense to let HP take the hit for R&D on the hardware side or at least acquire it from Intel and Hitachi. I think Oracle will spin off or kill most of Sun's business except for Java. Since WebSphere, IBM's cash cow, is dependent on Java, this kind of puts Oracle in the same position as Microsoft with Windows and Office. They can create a moving target hard for others to hit while skirting monopoly laws.
Yes, stop pricing software on CPU capacity
This has been a PITA for IBM mainframe customers for years. Furthermore it is exactly the kind of marketing model which led to customers simply shifting the jumper wire after the CE left the building and then fudging the metrics to hide it. Just sell the software for what it's worth. Hardware, too. This kind of stuff just alienates customers.
Part of this may be due to all the legal baggage associated with Sun's software which dates back to AT&T Unix in the 1970s. (The licensing issues predate Sun's founding.)
While I think it would be a win for IBM to gain control of Java given their WebSphere and Tivoli products that are so heavily rooted in Java tech, perhaps they are waiting for someone else to take the bait, break up and spin off Sun's assets, and then move in to pick up the pieces they want.
IBM doesn't need to buy Sun to get a piece of the server market, they already support both HP and Sun servers. IBM will sell and support anything that makes a profit even if it's produced by a competitor.
What they choose to buy, however, is a bit more selective. Sometimes you don't have to buy the whole box of candy just to get the piece with the cherry. Somehow I rather suspect this may be the motive here, not fear of monopoly status which IBM very nearly already has if you look at their market share.
Putting on my coat because IBM seems to be putting on theirs.
If this doesn't sink the Itanic...
I don't know what will. I've been desperately trying to talk HP into porting HP-UX to the little endian Nehalem because otherwise as I see it HP-UX is dead. Even with Tukwila, assuming it ever gets released, the Nehalem EX will blow it away before it's even released, and at much lower cost. Unless Intel has a hidden endian switch in Nehalem EX I don't really see how HP-UX can survive unless they port it to little endian Xeons. Even Solaris would run faster on an eight core Nehalem EX than HP-UX will on Tukwila.
Ha, the first people who will want to get their hands on this are the DEA so they can spot indoor pot growers. Just hook up to the utilities and you can spy on anyone. As for 1M resolution, anyone entering the structure had better watch out for the Claymores likely to be lurking behind those stud walls. Besides, I would think much of this can already be done much more readily using infrared and millimeter wave imaging and in a much more rapid fashion. As in real time. DARPA are like AIG. They once served a useful purpose but now it seems they are like NASA increasingly trying to find some way to justify their ongoing existence by coming up with increasingly wacky research proposals which are usually pointless. I think if I were a Marine about to enter a structure inhabited by hostiles I'd want something a bit more robust. I suppose it might be useful for targeting a JDAM but otherwise it seems pretty useless and redundant.
I'm putting on my coat because Elvis has obviously left the building. I'd use Paris but she has too much resolution.
Well HP also has had a tool on HP-UX for a long time that can do at least as much as dtrace if you know how to use it. As for Sun, many people have argued that the biggest problem they've had for more than a decade is Scott McNealy. There was a reason why Jobs was fired from Apple. And he destroyed NeXT. It remains to be seen if his charisma can save Apple. But maybe he learned some lessons during his exile. Sun just lost their vision. They started going off in all kinds of directions. SGI's failure was that they failed to diversify but Sun's is perhaps that they diversified too much. One has to wonder if IBM is paying attention to that. The idea of calling the entire world Tivoli or WebSphere sounds like the same litany they had in the heyday of the mainframe era. I'd like to see Sun survive because I think we need competition in the market but I think they need some new management with the vision to take them in a new direction. They've got a lot of intellectual property capital but they need some leadership to guide it in some productive direction. I'm not putting an emoticon on this post because it's just too depressing.
This begs the question...
...why the ISS was constructed in the first place. The Russians as I recall were quite opposed to it at the time because they felt they had already accomplished what any low orbit space station could do with MIR and indeed MIR could have been sustained past its mandated EOL had NASA not been so intent on this project. Now that NASA has sunk so much money into this low orbital version of AIG they now feel required to somehow justify its continuing existence. So I guess it's like that line from The Right Stuff, instead of exploring Mars we're exploring the planet Urine. One can only wonder when they will progress to Uranus. Maybe then they can make their own fertilizer in space. Of course in the eyes of many people they have already reached that objective.
Paris because she's smarter than most of the SES managers at NASA.
Hard to say...
...if this really fits IBM's business strategy. They do have a lot invested in Windows on their X series servers so I'm not sure they have anything to gain by running Windows on the mainframe unless Microsoft is willing to pony up a lot of partnership. Right now the IBM mainframe line is pretty much a layered SOA setup with Linux and evidently possibly Solaris running on peripheral servers which can be used to run WebSphere and other web apps while the central mainframe processors running z/OS are mainly focused on DB2 and similar backend high availability vertically scaled applications. Thus it's not clear to me that IBM has much to gain by running Windows on their mainframes. The can run it on Nehalem instead at much lower cost and WebSphere is horizontally scaled so having multiple four socket blades is not an issue there. It appears to me that IBM will probably keep Windows on X class blades though if there is evidence they're creating a new "reduced capability" processor like the ones they use for Linux on the mainframe then this would have more credence.
After all this debate...
...and not one mention of the "negative strangelet."
Grasping Defeat from the Jaws of Victory, or Deja Vu...
SGI's management have been more self destructive than Amy Winehouse on a bender since the mid 1990s. In every market segment (3-D graphics, HPC, storage, software) they have repeatedly squandered strong positions whilst off chasing unicorns.
The Elephant in the Room
Actually the Soviets reported a net positive energy output from one of the last of their Tokamak test devices in the 1980s. A seldom mentioned problem they found was that the magnetic field does little to contain the prodigious neutron flux created by the plasma which has the annoying habit of causing everything nearby to glow in the dark for a very long time as well as rapidly degrading the structural integrity of the reactor components, not a very desirable thing at all. This would seem to be the greatest obstacle to practical fusion power generation. Thus, fusion reactors are not so "clean" as their proponents frequently state. Even if one were to find some way of containing the neutron flux, what does one do with a few tons of waste neutrons?
Also, just to pick a nit, the majority of the yield of most thermonuclear weapons comes from fission of the depleted uranium "tamper" used to compress the thermonuclear fuel. This is also responsible for most of the radioactive fallout. The highly energetic neutrons from the thermonuclear reaction are capable of fissioning depleted uranium and this is used to dramatically boost the yield of the device.
Thalidomide, DDT, Asbestos
...are all examples of cases where extensive testing and widespread use did not initially reveal the harms eventually attributed to them. In most animal experiments Thalidomide did not cause developmental defects. This is why ongoing monitoring is important.