We considered having our IT staff wear the two-toned jumpsuits from STTNG, so we'd be easily identifiable.
That idea was voted down, but didn't involve physical violence, as a call for blue miniskirts on all staff would have.
Good for Intel; only ten years after Sun unveiled the Sparc IPX, several years after IBM/HP/De(u)ll managed to put Intel's own chips onto a machine the size of an add-in card, and several months after Apple.
On the other hand, since it's currently an empty box, it certainly won't catch any viruses, so it's at least as useful as current full-sized PCs.
Warning: I administered most of those machines at one point or another.
While I am sorry to see the Alpha go (a great missed opportunity, and still a reason to find Ken Olsen and slap him silly), and possibly the PA-Risc, it's hard to mourn the passing of the perennial dog Sparc "we have a balanced architecture; all of our components are equally slow!", and the bipolar SGI (Version 7.x of the compiler toolchain, otherwise known as the bug of the month club). Itanium, despite being razzed continually by 18yr old Doom players, is a good replacement for most of those processors, but thanks to Intel/HP's inept marketing, it's going to be the Alpha of the 21st century.
I have run, at one point or another, VAX/VMS, MIPS/Ultrix, Alpha/OSF-1, MIPS/Irix (3.0 -> 6.5.x), SunOS, Solaris, Power/AIX (3 -> 4), and several versions of Linux(x86/IA64/Alpha), and the only old architectures I'm sorry to see gone are the Cray vector processor and the Alpha. I can now commit more science in less time for less money than ever before, and am frankly encouraged to see the 4-8way Opterons with their Hypertransport interconnect appearing from major vendors. They'll replace nicely, over time, the SGI Origins, and similar NUMA boxes.
A mass extinction is always depressing to watch, but the only real loss in that list is the Alpha, of which pieces are showing up in x86 anyway.
Interestingly, if you remove the white "do not remove" stickers on the motherboard of their Opteron workstation, you'll find underneath three little letters, "IBM".
The question then starts to become, "why don't I buy an IBM/IBM workstation, or an HP-4way Opteron, and run Solaris-10 on it?"
Sun's had a problem for a long time. We stopped buying them when the Itanium-Is flogged the US-III for less money, then then the Itanium-2/Opterons came along and widened the gap. The slowest model of each of those was 2-3x faster for floating-point heavy simulation code, and in the case of the IA-64, would run HP-UX, if we didn't want to deal with Linux.
We told Sun about these numbers, they countered that the US-IV was on the way, and I pointed out that according to their specs, the US-IV would be no faster than the IA64s we could order then and there. The salesmen knew it, but management had made their decision "Sparc Then, Sparc Now, and Sparc Forever! (or until we have no customers, whichever comes first)"
"having helped drive a successful company into the ground, Ms. Fiorina will be expected to reimburse HPaq for her time spent with the company. Further, having been branded with the Corporate Mark of Cain(tm) she will be starting a new career as a Wayne Newton impersonator in Vegas."
Btw, according to the accounts in the business media (I know, less reliable than benchmarks) when she took over, she'd actually been a disaster at Lucent as well; just good at burying the bodies.
What you would expect; you perform the same functions, but they look prettier, then your machine is laboring to constantly defend itself against the outside world. On the other hand, if you're running simulations, you'll notice a massive improvement in computer power, so maybe this perceived "slowness" is instead just impatient users. (for those of you running WinXP on a P133, yes, you don't just think it's slow, it is slow.)
Kind of makes me miss Wordstar. On the other hand, think of how many fewer cycles would be left if AfterDark were still available.
I think the small fortran program around here is 20K lines, while the big ones are closing in on a million or so. Most Number crunching code is actually rather complex.
Note: these are Real Lines, not girlie-man C lines with one { on them.
You can do anything you'd ever want to with Fortran. I've seen OO F77, and I've seen CGI scripts written in it. (in a language where text processing was an extension from the 70s). The predecessor to Zork was written in Fortran as well.
It has achieved immortality, and shall never die. Where is Ada? Where is Algol? Whither Smalltalk? FORTRAN doth remain.
The bigger problem in the researcher pays to publish system is that you could easily blow most of a small research budget just publishing results, rather than generating them.
If you want the cost to institutions of journals to drop, the cheapest solution is drop everything by Reed-Elsevier, which tend to be low-circulation, low-impact, journals.
One could probably also argue that as the Gov't is already paying for a good deal of research, then it might as well shoulder the cost of publishing some of it. An institution or consortium of institutions would submit proposals to host a journal (as is now done for the National Lab management contracts), the most scientifically qualified would end up as the editors and be paid a stipend for running the journal, then the actual journal would be published on the servers from one of the national labs. In this system, all that's changed is that the middle-man of a commercial publisher has been cut out, and the excess profits they would have earned can instead be used for more server space or another journal or two.
Actually, I am a Linux user (and Solaris user, and AIX user) who switched. To quote His Steveness (cue chorale of relaxed sysadmins) It Just Works. OS-X is clean, efficient, unobtrusive, comes with a real CLI, X, and you can get IBM's Fortran for it. With minimal pain I've built a cluster from G5 Xserves, which are also more power efficient than their competition. What more do you want in a computer?
I would still run linux servers for specialized configurations (IBM Power5 or HP Itanium-2), but on the desktop there is no competition. All of the pain and randomness is just gone. The funky config files, uncertain power management, inconsistent interfaces, and "never quite finishedness" of linux, gone. My visualization and simulation apps, right there, and running clean and fast.
I'm not going back unless His Steveness is hit by a bus, and Gil Amelio takes over again.
We've actually have this problem already: PowerPoint presentations from NSF with video and voice-overs.
Personally, 5.1 sound makes them a less soporific experience. Besides, when you take the Windows Tour in your new copy of Longhorn, how are they going to simultaneously play the theme from "Rawhide", and lecture you in those soothing tones James Earl Jones brought to Vader?
You're comparing it to the wrong chips. The G5 competes with the Opterons and Itaniums. If you buy the cluster-node configuration (stripped down to the minimum on peripherals), it's less expensive than a major-vendor Opteron system, and you get IBM's XLF and XLC compilers in the bargain. This allows easy porting of code, often with identical optimizations, straight from AIX.
Dual-Xeons aren't all that cheap, and when the G5 outpaces them by a factor of two on floating-point intensive code, then the decision becomes clearer.
I just built a cluster of them, and for simulations that take weeks to run, Xeons/Athlons don't enter into it. Remember, that with 5x the number of nodes, you also get 5x the failures, presuming an identical MTBF between the G5 and the Xeon.
Maybe he meant "a helping of". One win-dose is the maximum quantity of windows you're supposed to be exposed to daily. Kind of like Rems for people who work with x-rays.
The interesting number I found was that for Itania we could afford, you're right about Opterons winning with scientifc (ahem, those are all CHEMSISTRY codes; declare you field with PRIDE!), with one exception: Coupled-cluster. (CCSD(T)) to be specific.
A 900 MHZ Itanium, consuming twice the power and running 2/3 as fast as a 2.0 GHZ Opteron on the rest of the applications, ran CCSD(T) in 1/2 the time. It's worth keeping them around just for that. I'll also note that the HP ZX6000 are very robust, reliable, environment-tolerant machines.
Such a lovely architecture, doomed to be the Alpha of the 21st century.
Come on, imagination, people...
You could put the entire Sloan Digital Sky survey on your desktop with only 100 of these...
You could start doing medical imaging
Or, you could go to #insert , buy a 6-8 MPel digital camera, and actually store some pictures.
500 G? Scans from 35mm take up 54M/frame, presuming you don't make backups or save multiple copies of work files.
Or, using your system, you could hire a million monkeys, buy a million typewriters, and save all of their output to read at your leisure later.
A truce:
We'll import your unpasteurized cheeses, and you'll stamp little (up, eh?, or down) thumb symbols on your wine that you export to us.
Maybe we can agree to start smoking again, and you'll stop trying to export Renaults.
Slashdot Mirror
We considered having our IT staff wear the two-toned jumpsuits from STTNG, so we'd be easily identifiable. That idea was voted down, but didn't involve physical violence, as a call for blue miniskirts on all staff would have.
You mean that there's an Apple-AIX as well? I've been dying to use SMIT on my home box!
Good for Intel; only ten years after Sun unveiled the Sparc IPX, several years after IBM/HP/De(u)ll managed to put Intel's own chips onto a machine the size of an add-in card, and several months after Apple.
On the other hand, since it's currently an empty box, it certainly won't catch any viruses, so it's at least as useful as current full-sized PCs.
Warning: I administered most of those machines at one point or another.
While I am sorry to see the Alpha go (a great missed opportunity, and still a reason to find Ken Olsen and slap him silly), and possibly the PA-Risc, it's hard to mourn the passing of the perennial dog Sparc "we have a balanced architecture; all of our components are equally slow!", and the bipolar SGI (Version 7.x of the compiler toolchain, otherwise known as the bug of the month club). Itanium, despite being razzed continually by 18yr old Doom players, is a good replacement for most of those processors, but thanks to Intel/HP's inept marketing, it's going to be the Alpha of the 21st century.
I have run, at one point or another, VAX/VMS, MIPS/Ultrix, Alpha/OSF-1, MIPS/Irix (3.0 -> 6.5.x), SunOS, Solaris, Power/AIX (3 -> 4), and several versions of Linux(x86/IA64/Alpha), and the only old architectures I'm sorry to see gone are the Cray vector processor and the Alpha. I can now commit more science in less time for less money than ever before, and am frankly encouraged to see the 4-8way Opterons with their Hypertransport interconnect appearing from major vendors. They'll replace nicely, over time, the SGI Origins, and similar NUMA boxes.
A mass extinction is always depressing to watch, but the only real loss in that list is the Alpha, of which pieces are showing up in x86 anyway.
Interestingly, if you remove the white "do not remove" stickers on the motherboard of their Opteron workstation, you'll find underneath three little letters, "IBM".
The question then starts to become, "why don't I buy an IBM/IBM workstation, or an HP-4way Opteron, and run Solaris-10 on it?"
Sun's had a problem for a long time. We stopped buying them when the Itanium-Is flogged the US-III for less money, then then the Itanium-2/Opterons came along and widened the gap. The slowest model of each of those was 2-3x faster for floating-point heavy simulation code, and in the case of the IA-64, would run HP-UX, if we didn't want to deal with Linux.
We told Sun about these numbers, they countered that the US-IV was on the way, and I pointed out that according to their specs, the US-IV would be no faster than the IA64s we could order then and there. The salesmen knew it, but management had made their decision "Sparc Then, Sparc Now, and Sparc Forever! (or until we have no customers, whichever comes first)"
Actually, AS/400s from IBM have implemented that feature quite nicely for over a decade.
Maybe that's the answer; they're putting a Windows front-end onto OS/400. Your new copy of Windows, now shipping with XEDIT!
wouldn't you like to see the announcement read:
"having helped drive a successful company into the ground, Ms. Fiorina will be expected to reimburse HPaq for her time spent with the company. Further, having been branded with the Corporate Mark of Cain(tm) she will be starting a new career as a Wayne Newton impersonator in Vegas."
Btw, according to the accounts in the business media (I know, less reliable than benchmarks) when she took over, she'd actually been a disaster at Lucent as well; just good at burying the bodies.
to pay someone to beat B&B with a halibut instead?
You don't get more episodes, but you do get a greater sense of personal accomplishment.
What you would expect; you perform the same functions, but they look prettier, then your machine is laboring to constantly defend itself against the outside world. On the other hand, if you're running simulations, you'll notice a massive improvement in computer power, so maybe this perceived "slowness" is instead just impatient users. (for those of you running WinXP on a P133, yes, you don't just think it's slow, it is slow.) Kind of makes me miss Wordstar. On the other hand, think of how many fewer cycles would be left if AfterDark were still available.
True. You should be learning Chinese, so you'll have a useful skill and be able to talk to your new, well-educated, overlords in a few years.
I think the small fortran program around here is 20K lines, while the big ones are closing in on a million or so. Most Number crunching code is actually rather complex. Note: these are Real Lines, not girlie-man C lines with one { on them. You can do anything you'd ever want to with Fortran. I've seen OO F77, and I've seen CGI scripts written in it. (in a language where text processing was an extension from the 70s). The predecessor to Zork was written in Fortran as well. It has achieved immortality, and shall never die. Where is Ada? Where is Algol? Whither Smalltalk? FORTRAN doth remain.
The bigger problem in the researcher pays to publish system is that you could easily blow most of a small research budget just publishing results, rather than generating them. If you want the cost to institutions of journals to drop, the cheapest solution is drop everything by Reed-Elsevier, which tend to be low-circulation, low-impact, journals. One could probably also argue that as the Gov't is already paying for a good deal of research, then it might as well shoulder the cost of publishing some of it. An institution or consortium of institutions would submit proposals to host a journal (as is now done for the National Lab management contracts), the most scientifically qualified would end up as the editors and be paid a stipend for running the journal, then the actual journal would be published on the servers from one of the national labs. In this system, all that's changed is that the middle-man of a commercial publisher has been cut out, and the excess profits they would have earned can instead be used for more server space or another journal or two.
Actually, I am a Linux user (and Solaris user, and AIX user) who switched. To quote His Steveness (cue chorale of relaxed sysadmins) It Just Works. OS-X is clean, efficient, unobtrusive, comes with a real CLI, X, and you can get IBM's Fortran for it. With minimal pain I've built a cluster from G5 Xserves, which are also more power efficient than their competition. What more do you want in a computer? I would still run linux servers for specialized configurations (IBM Power5 or HP Itanium-2), but on the desktop there is no competition. All of the pain and randomness is just gone. The funky config files, uncertain power management, inconsistent interfaces, and "never quite finishedness" of linux, gone. My visualization and simulation apps, right there, and running clean and fast. I'm not going back unless His Steveness is hit by a bus, and Gil Amelio takes over again.
We've actually have this problem already: PowerPoint presentations from NSF with video and voice-overs. Personally, 5.1 sound makes them a less soporific experience. Besides, when you take the Windows Tour in your new copy of Longhorn, how are they going to simultaneously play the theme from "Rawhide", and lecture you in those soothing tones James Earl Jones brought to Vader?
You're comparing it to the wrong chips. The G5 competes with the Opterons and Itaniums. If you buy the cluster-node configuration (stripped down to the minimum on peripherals), it's less expensive than a major-vendor Opteron system, and you get IBM's XLF and XLC compilers in the bargain. This allows easy porting of code, often with identical optimizations, straight from AIX. Dual-Xeons aren't all that cheap, and when the G5 outpaces them by a factor of two on floating-point intensive code, then the decision becomes clearer. I just built a cluster of them, and for simulations that take weeks to run, Xeons/Athlons don't enter into it. Remember, that with 5x the number of nodes, you also get 5x the failures, presuming an identical MTBF between the G5 and the Xeon.
Maybe he meant "a helping of". One win-dose is the maximum quantity of windows you're supposed to be exposed to daily. Kind of like Rems for people who work with x-rays.
The interesting number I found was that for Itania we could afford, you're right about Opterons winning with scientifc (ahem, those are all CHEMSISTRY codes; declare you field with PRIDE!), with one exception: Coupled-cluster. (CCSD(T)) to be specific. A 900 MHZ Itanium, consuming twice the power and running 2/3 as fast as a 2.0 GHZ Opteron on the rest of the applications, ran CCSD(T) in 1/2 the time. It's worth keeping them around just for that. I'll also note that the HP ZX6000 are very robust, reliable, environment-tolerant machines. Such a lovely architecture, doomed to be the Alpha of the 21st century.
Come on, imagination, people... You could put the entire Sloan Digital Sky survey on your desktop with only 100 of these ...
You could start doing medical imaging
Or, you could go to #insert , buy a 6-8 MPel digital camera, and actually store some pictures.
500 G? Scans from 35mm take up 54M/frame, presuming you don't make backups or save multiple copies of work files.
Or, using your system, you could hire a million monkeys, buy a million typewriters, and save all of their output to read at your leisure later.
A truce: We'll import your unpasteurized cheeses, and you'll stamp little (up, eh?, or down) thumb symbols on your wine that you export to us. Maybe we can agree to start smoking again, and you'll stop trying to export Renaults.
Actually, that's why we installed OS/2... and he could have had Visual Cobol for it as well. (now there's a GNU project that needs to get started)