The way to make space transport cheaper is by more efficient rocket engines, lighter vehicles, better propellants, single single stage to orbit etc. i.e. new technology. One such project was the X33. That got canned:-(
A couple of TB? Er, the Pentium architecture has 36-bit addressing which is 4 bits + 32-bits which gives you 16 (4-bit's worth) banks of 4GB(32-bit's worth) of RAM. 16*4 is 64, so by my reckoning your "couple of terabytes" is in fact only 64 gigabytes, so it's a couple of orders of magnitude smaller. Anyway, we're back to the good old days of segmented memory:-) Oh joy! 64k of code, 64k of stack and 64k of data, and you'd better like it young man.
That's true, but most (not all) of the API's are already in there. A lot of Open Source/Free Software libraries have been migrated into the Solaris WOS (and are supported by Sun). I can't think of all of them off the top of my head but they're in/usr/sfw/lib on Solaris 9. Also, you get a lot more (but unsupported) if you install the Companion CD (which installs in/opt/sfw). So, for many applications, it's just a case of giving./configure the right paths, and stuff Just Works(TM). I agree that being LSB compliant would be good, and an improvement, but a lot of the (important) stuff is already there.
Mozilla, mozilla,....and Mozookie
(or should that be Mozuki?)
I'd have thought that would have been the unanimous first 300 suggestions, but maybe our cartoons were different in Scotland?
Whatever happened to LinearAerospike rocket engines, that were supposed to be more efficient than conventional ones with bell-shaped nozzles? In theory they could be made anular as well as linear. Wouldn't these make good replacements for the shuttles main engines, making it more efficient? What's holding back their development? I once read that they were considered for the space shuttle when it was being designed but they were considered too new and radical.
First of all, you need a Rotary Space Wave Generator, as commonly found on the NII's Winnebagos of doom. To make one, first you need two coat hangers, a 9V battery and two plastic plates. Then you need an anti-gravity source and amplifier consisting of at leas 125g of element 115 (ununpentium) obtainable in quantity from intel corp. For the gravity amplifier you will need an oak wardrobe and a flux capacitor. Connect the flux capacitor (in the wardrobe) with copper wire to the ununpentium. The rotary space wave generator may now be mated to the other end of the wardrobe. There will be a pleasant smell as you drill through the oak. Pick up a sencond hand anti-matter regeneration unit form your local scrap yard. This will transmit anti-waves to the rotary space wave generator. To control the device you will need a dual channel K6-2/Pentium III 500 system running Windez 98 under Linux Technology and a 28" plasma screen with military spec. joystick. Good luck, and have fun.
A company called nCube used to make a line of hyper-cube configuration transputer machines. Ten years ago when I was doing physics at university, there were some in the lab. I think they had specialised C and FORTRAN compilers for them as well as Occam.
Probably not quite as good as a plain cylinder/electric hybrid, but then there's the advantage of the substantial saving in weight and size. Horses for courses I suppose.
Cool! I had the Skywave Software multi-tasking FORTH ROM for my ZX81. It was a drop-in replacement for the 8k BASIC ROM and contained multi-tasking (real-time) FORTH. I had both ROMS installed on a daughterboard that slotted into the ROM socket and a switch to select between the two. It's still at my parents'...
Remember, US gallons are smaller than Imperial ones (as used in the UK). From the back of my diary, 1 US gallon = 0.8327 Imperial gallons, therefore 50mpg (US) is 60 mpg (UK/Imperial) which is pretty darn impressive for a petrol (gasoline) engine. A hybrid diesel would be even better, if you can put up with that nasty black soot they emit.
What would be really cool would be a hybrid Wankel-electric engine. You'd get the smoothness and high power to weight ratio of the wankel combined with the efficiency of the electric motor. Mazda, any plans for the RX-9?:-)
The coolest boat I've ever been on was the HSS Stena Explorer which goes from Harwich (England) to the Hook of Holand.
This thing is a 100 000 horsepower, 126 metre-long catamaran that cruises at about 40kts (IIRC). It takes 120 cars and 50 lorries.
It is totally awesome pounding across the north sea leaving every other vessel standing. It is powered by two small and two large gas turbines. The small ones are for low speed and the large for high speed. They drive several waterjets wich are steerable, and they're operated at maximum efficiency (hence small and large for different speeds). There is a plasma display with a map updated by GPS in realtime showing your progress, at the bow. The second time I crossed the sea in it, it was pretty rough, but I managed not to spill my pint, whereas all the scurvy land-lubbers were decidedly green:-)
How about UltraSPARC?
on
Going Itanium 2?
·
· Score: 3, Interesting
Have you thought about Sun UltraSPARC machines? You can get some pretty decent low-end boxen such as the Sun Blade 100 and Sun Blade 150 or go for more up-market workstations like the Sun Blade 2000 or servers. UltraSPARC is currently in it's 3rd incarnation (USIIICu), has been about for years, is well understood, and can run Solaris and various Linux distros. GCC runs on it as does most Free and Open Source software. UltraSPARC is the only proven high-end 64-bit CPU with a long-term roadmap.
Message Passing vs. Single System Image
on
Ask Donald Becker
·
· Score: 5, Interesting
Why do you think that message passing clusters are more popular than single system image clusters, and do you see the balance changing eventually? In other words, is there no compelling reason to choose single system image for most problems?
Also, when do you think that the 32-bit addressing limitations of x86 hardware will become a problem for doing Big Science on clusters?
You do realise that 64-bit workstations have been around for many years, and they didn't come from intel? Like this and this. They're pretty affordable too.
I had to for my final year project at uni. I wasn't allowed to use C, and I had to write standards-compliant F77 i.e. with none of the convenient extensions. The program was to look for structure in simulated collapsing gas clouds in the intertellar medium, i.e. to look for protostars. I ended up having to implement a doubly linked list and binary trees inside (static) arrays. It was a miracle it worked. After learning C, going to FORTRAN was hell. You can see where early 80s microcomputer BASICs came from when you are reduced to writing FORTRAN-77, only the source formatting is worse...
I hear a lot of negative comments from people here saying that you must be some kind of loser to start with, but I disagree strongly.
It's easy to feel "burned-out" and stressed working for the wrong sort of company i.e. two-bit operations and places where they expect you to do absolutely everything, work in the evenings at home, weekends on unrealistic projects, unrealistic deadlines etc. I've been there myself, and here's what I did.
My degree is in Astrophysics, but I've been a computer geek since I was 8 years old.
After my degree I got some losy jobs contracting with some small IT companies that were going no where and burning me out too. So, I exited stage left and got a job as a Reactor Physics Engineer at a nucelar powerstation.
It was very different. Sensible, intelligent people, realistic deadlines and workload, reasonable pay, long holidays, sensible overtime. I stayed there for nearly 5 years, being trained in all maner of technical subjects (reactor physics theory) right down to occupational safety. In the meantime I was running Linux at home, improving my coding, learning C++, etc. Then, when a better job came up at a big computer systems company I was all set.
I have no regrets at all. It's not just about you, it's about who you're working for.
Re:address computations for 4GB of memory
on
Itanium Problems
·
· Score: 2
You're right, it is utterly straight-forward, and that's why AMD has done it. You can see this from AMD's documentation about the Hammer architecture at www.x86-64.org. Intel can't do the same because it has bet the farm on itanium and it needs to persuade the corporate world to buy millions of itanium machines to recoup its R&D spending. They may have backed the wrong horse because 10 years ago when they started the project it seemed like the best thing to do. They will try every marketing, slaes and PR trick in the book to noble AMD's architecture simply because the cannot afford for Hammer to succeed.
Re:Intel relies on compiler, Turing says it's fool
on
Itanium Problems
·
· Score: 2
You only have to look at it's (itanic) performance under gcc to see the answer to this question. I seem to remember performance figures 20% of that obtainable with intel's compiler.
Slashdot Mirror
The way to make space transport cheaper is by more efficient rocket engines, lighter vehicles, better propellants, single single stage to orbit etc. i.e. new technology. One such project was the X33. That got canned :-(
A couple of TB? Er, the Pentium architecture has 36-bit addressing which is 4 bits + 32-bits which gives you 16 (4-bit's worth) banks of 4GB(32-bit's worth) of RAM. 16*4 is 64, so by my reckoning your "couple of terabytes" is in fact only 64 gigabytes, so it's a couple of orders of magnitude smaller. Anyway, we're back to the good old days of segmented memory :-) Oh joy! 64k of code, 64k of stack and 64k of data, and you'd better like it young man.
That's true, but most (not all) of the API's are already in there. A lot of Open Source/Free Software libraries have been migrated into the Solaris WOS (and are supported by Sun). I can't think of all of them off the top of my head but they're in /usr/sfw/lib on Solaris 9. Also, you get a lot more (but unsupported) if you install the Companion CD (which installs in /opt/sfw). So, for many applications, it's just a case of giving ./configure the right paths, and stuff Just Works(TM). I agree that being LSB compliant would be good, and an improvement, but a lot of the (important) stuff is already there.
RPM is provided on the Solaris Companion CD so you can already use source RPMs with Solaris.
Mozilla, mozilla,.. ..and Mozookie
(or should that be Mozuki?)
I'd have thought that would have been the unanimous first 300 suggestions, but maybe our cartoons were different in Scotland?
Whatever happened to Linear Aerospike rocket engines, that were supposed to be more efficient than conventional ones with bell-shaped nozzles? In theory they could be made anular as well as linear. Wouldn't these make good replacements for the shuttles main engines, making it more efficient? What's holding back their development? I once read that they were considered for the space shuttle when it was being designed but they were considered too new and radical.
First of all, you need a Rotary Space Wave Generator, as commonly found on the NII's Winnebagos of doom. To make one, first you need two coat hangers, a 9V battery and two plastic plates. Then you need an anti-gravity source and amplifier consisting of at leas 125g of element 115 (ununpentium) obtainable in quantity from intel corp. For the gravity amplifier you will need an oak wardrobe and a flux capacitor. Connect the flux capacitor (in the wardrobe) with copper wire to the ununpentium. The rotary space wave generator may now be mated to the other end of the wardrobe. There will be a pleasant smell as you drill through the oak. Pick up a sencond hand anti-matter regeneration unit form your local scrap yard. This will transmit anti-waves to the rotary space wave generator. To control the device you will need a dual channel K6-2/Pentium III 500 system running Windez 98 under Linux Technology and a 28" plasma screen with military spec. joystick. Good luck, and have fun.
A company called nCube used to make a line of hyper-cube configuration transputer machines. Ten years ago when I was doing physics at university, there were some in the lab. I think they had specialised C and FORTRAN compilers for them as well as Occam.
I'll be keeping my eye out for one in the 2nd hand shops...
Yes, Tesla turbines look cool. How come we don't see more of them in real world applications?
In which case my father's 1988 Vauxhall Cavalier equals your hybrid's fuel economy.
Probably not quite as good as a plain cylinder/electric hybrid, but then there's the advantage of the substantial saving in weight and size. Horses for courses I suppose.
Cool! I had the Skywave Software multi-tasking FORTH ROM for my ZX81. It was a drop-in replacement for the 8k BASIC ROM and contained multi-tasking (real-time) FORTH. I had both ROMS installed on a daughterboard that slotted into the ROM socket and a switch to select between the two. It's still at my parents'...
Remember, US gallons are smaller than Imperial ones (as used in the UK). From the back of my diary, 1 US gallon = 0.8327 Imperial gallons, therefore 50mpg (US) is 60 mpg (UK/Imperial) which is pretty darn impressive for a petrol (gasoline) engine. A hybrid diesel would be even better, if you can put up with that nasty black soot they emit.
What would be really cool would be a hybrid Wankel-electric engine. You'd get the smoothness and high power to weight ratio of the wankel combined with the efficiency of the electric motor. Mazda, any plans for the RX-9? :-)
The coolest boat I've ever been on was the HSS Stena Explorer which goes from Harwich (England) to the Hook of Holand. :-)
This thing is a 100 000 horsepower, 126 metre-long catamaran that cruises at about 40kts (IIRC). It takes 120 cars and 50 lorries.
It is totally awesome pounding across the north sea leaving every other vessel standing.
It is powered by two small and two large gas turbines. The small ones are for low speed and the large for high speed. They drive several waterjets wich are steerable, and they're operated at maximum efficiency (hence small and large for different speeds).
There is a plasma display with a map updated by GPS in realtime showing your progress, at the bow.
The second time I crossed the sea in it, it was pretty rough, but I managed not to spill my pint, whereas all the scurvy land-lubbers were decidedly green
Have you thought about Sun UltraSPARC machines? You can get some pretty decent low-end boxen such as the Sun Blade 100 and Sun Blade 150 or go for more up-market workstations like the Sun Blade 2000 or servers. UltraSPARC is currently in it's 3rd incarnation (USIIICu), has been about for years, is well understood, and can run Solaris and various Linux distros. GCC runs on it as does most Free and Open Source software. UltraSPARC is the only proven high-end 64-bit CPU with a long-term roadmap.
Why do you think that message passing clusters are more popular than single system image clusters, and do you see the balance changing eventually? In other words, is there no compelling reason to choose single system image for most problems? Also, when do you think that the 32-bit addressing limitations of x86 hardware will become a problem for doing Big Science on clusters?
Yes, they rule :-) Seen them live 3 times, at Reading 2001, Ozzfest 2002 and headlining at Brixton.
You do realise that 64-bit workstations have been around for many years, and they didn't come from intel? Like this and this. They're pretty affordable too.
I had to for my final year project at uni. I wasn't allowed to use C, and I had to write standards-compliant F77 i.e. with none of the convenient extensions. The program was to look for structure in simulated collapsing gas clouds in the intertellar medium, i.e. to look for protostars. I ended up having to implement a doubly linked list and binary trees inside (static) arrays. It was a miracle it worked. After learning C, going to FORTRAN was hell. You can see where early 80s microcomputer BASICs came from when you are reduced to writing FORTRAN-77, only the source formatting is worse...
C already went in that direction, twice. It's called C++ and Objective-C.
I hear a lot of negative comments from people here saying that you must be some kind of loser to start with, but I disagree strongly.
It's easy to feel "burned-out" and stressed working for the wrong sort of company i.e. two-bit operations and places where they expect you to do absolutely everything, work in the evenings at home, weekends on unrealistic projects, unrealistic deadlines etc. I've been there myself, and here's what I did.
My degree is in Astrophysics, but I've been a computer geek since I was 8 years old.
After my degree I got some losy jobs contracting with some small IT companies that were going no where and burning me out too. So, I exited stage left and got a job as a Reactor Physics Engineer at a nucelar powerstation.
It was very different. Sensible, intelligent people, realistic deadlines and workload, reasonable pay, long holidays, sensible overtime. I stayed there for nearly 5 years, being trained in all maner of technical subjects (reactor physics theory) right down to occupational safety.
In the meantime I was running Linux at home, improving my coding, learning C++, etc. Then, when a better job came up at a big computer systems company I was all set.
I have no regrets at all. It's not just about you, it's about who you're working for.
You're right, it is utterly straight-forward, and that's why AMD has done it. You can see this from AMD's documentation about the Hammer architecture at www.x86-64.org. Intel can't do the same because it has bet the farm on itanium and it needs to persuade the corporate world to buy millions of itanium machines to recoup its R&D spending. They may have backed the wrong horse because 10 years ago when they started the project it seemed like the best thing to do. They will try every marketing, slaes and PR trick in the book to noble AMD's architecture simply because the cannot afford for Hammer to succeed.
You only have to look at it's (itanic) performance under gcc to see the answer to this question. I seem to remember performance figures 20% of that obtainable with intel's compiler.