In radioastronomy correlations and FFT's is a large part of the computations we do. Some tests we've done on Nvidia 7900's shows those could be sped up by a factor of 300 over a general CPU. The big problem is bandwidth. The amount of processing and bandwidth needed for these new telescopes comming online is staggering. For LOFAR we're using a IBM BlueGene with 12.000 Cores (Stella, nr.12 in the top500), using a 144Gb/s connection, and for SKA the nummbers are going to be orders of magnitude larger. The possibilities of these GPGPUs are studied with great interest in the radioastronomy community.
This is about PDA's and and TAblet PCs that need to be operated with a pen. I have one experience with this: When I was looking for a PDA in 2000, I looked at my friends Palm V, which where the scrollbars could be flipped to the left, and the pen holder on the right was identical to the holder of the flipcover on the left, so they could be interchanged. But the Palm was only 32Mhz, back&white, while the new PocketPCs were fast enough to run mp3, have CF readers and 16bit color displays.
So I bought a PocketPC (Cassiopeia E125).
I can tell you these are VERY awkward to use for lefthanded people. You are constantly moving your hand in front of the screen, the pen holder is on the wrong side, all buttons are on the left side, INCLUDING the scroll wheel, which I _can't_ use, because I'm holding a pen with my left hand.
I have only used my (very expensive) PocketPC for a few months, then I didn't use a PDA for years, but since some time I am now using a Nokia smart phone, and realy liking it. I heard that the newest model PocketPCs have become somewhat better, as the one I had 6 years ago, but I'll never buy another one.
I don't know where you live, but that's not true overhere, we hard a reasonably cold winter, and a very hot spring and summer. (average temperature of 23 instead of 17 degrees and 15 instead of 12) http://www.knmi.nl/klimatologie/maand_en_seizoenso verzichten/index.html#jaar rainfall has been 15 instead of 70 mm, so I don't know where you got that from either.
Only after the first of august have temperatures gotten more in the usual range again.
The advantage of the Blue Gene is that it is relatively simple to indeed go low level and manually tune your application at the assembly level. But I agree that a BG isn't the best solution to everything, and there are still some issues IBM needs to work on.
I think about 75% of the programming of the comercial Dutch TV stations (7) I recieve and about 25% on the public Dutch and Belgian TV stations is comming from the USA, UK, Australia, about 5% from Germany. Nothing gets dubbed everything has sub-titles.
A random example: 06.00 Batman 06.20 Duck Dodgers 07.00 Batman 07.25 Roseanne 07.55 Tel Sell 08.25 Children's ward 08.55 TV Shop 09.55 Astro TV 10.30 TV Shop 11.30 Phone party 12.35 Providence 13.35 Children's ward 14.05 Name the word 15.15 Disney feest: Aladdin 16.00 Duck Dodgers 16.25 Word quest 17.05 ER 18.05 Providence 19.00 Strong medicine 20.00 Friends 20.30 Possession 22.30 Extreme makeover Vlaanderen 23.35 Friends 00.05 Queen's supreme
I would think that the combination of: Novell (SuSE, eDirectory, lot's of enterprise tools and credibility) Corel (WordPerfect Office, CorelDraw/Photopaint, Paintshop Pro) Borland (Kylix, jbuilder) if combined, could make a difference.
But all of them tried and failed on their own, (on Novell te jury is still out), because of different timelines. The main thing that would be lacking is a good game developement platform.
If substract all firearms murders (0.027) from the total (0.042), you get almost the same number as for the Netherlands (0.011), where owning a firearm is mostly illegal. I know you can easily lie with statistics, but to me it suggests that a country with few people owning firearms, is a lot saver as one where almost everyone owns one.
I just thought to add a link: http://www.lofar.org/p/systems.htm Those numbers aren't exactly the same as the ones I quoted, it might be the website is out of date with the latest info, or my memory is failing me. They're still the same order of magnitude though. It migh just depend if you quote raw data rate, correlated data rate, or the rate at which the scientific results come out.
Actually, most astronomers use FORTRAN there days. Packages like AIPS and MIRIAD are completely written in them. The newer stuff like AIPS++ uses C++.
I'm working on one of these next-generation telescopes, it LOFAR, we hope to have it operational in 2008. All software is written in C++, except for some user interfaces in Java.
The telescope in the topic is only a dream at this point, they have nowhere near the funding to start yet. LOFAR on the other hand is already being build. Our software correlator is already running on our IBM BlueGene, making it the 9th fastest computer in the world. Our 144 GBit/s links to the sub-stations are operational, and the first full substation (of 77) will be operational next month.
These guys are talking 30 TByte/day, we're talking a raw datarate of 1.5 Petabyte/day at the end of 2008. This is going to be the largest radio-telescope in the world, at 300km (200 mi.), at least until SKA gets build (if it gets build)
I'm mainly rehashing what I learned when I visited JET a couple of years ago. I'm more kind of an astrophysics engineer myself.
As far as I understand the whole thing is mostly an engineering issue. With something like an Apollo Moon project style approach it could maybe be done in 10 years. It would have to be a massive national or super national effort.
I think history has shown us that especially where efficiency is the main concern, like it seems to be with the whole fussion problem, it will just take time for people to come up with bright engineering solutions.
And you're right that there's also a lot of politics involved, as always with large sums of money.
Short explanation: Try building a heating system or oven entirely out of ice, without melting the ice in the process.
Long explanation: The problem is threefold: 1) It's very hot (millions of degrees (Celsius)). So you need very good isolation and confinement, or it will boil your facility and cool down itself to the point where it's useless. You have something which is generating energy, it is very, very hot, and therefore contains very fast particles, that want to travel in a straight line, "filling" the vacuum you are trying to use to confine them, and then hitting your device and vaporizing it.
This is the main reason why in the fussion world big is beautiful, because the volume/surface area ratio gets more favorable the bigger you get, so it becomes easier to keep the heat inside.
2) So on the one hand you're trying to create the perfect thermosflask, but then you also want to extract some of that energy in a controlled way. Remind you, this stuff is HOT. It will vaporize anything it comes into contact with, including any heat exchanger you can think of. Basically the only known way is how the sun does it: using (infrared) light. You have the light that gets generated travel accross the vacuum you are using to contain the actual plasma and hit something, like the sunlight can heat something. But you want to control what get's heated and light can not be guided with magnetic fields as plasma particles can. So you need the most fancy mirrors in existence.
3) once you have all of this figured out (and they hope to get close with ITER) you'll need to have a sustained reaction, so you'll also need a way to keep feeding it new "fuel", without destabilizing the entire process. THink of it as chucking new logs on a fire, but these logs are very wet and frozen, because the're millions of degrees colder as your fire. It's very easy to exstinguish the fire.
I find the difference in reception quality between my old Sony and my new Nokia, quite big. I work in an official no RFI zone, in my country one of the few places with very poor reception (GSM). The Nokia does not give these "Under water sounds" that the Sony gives when they're at almost no signal. (same location same provider, only different phone) They both sometimes drop the entire conversation, if I move while on the phone, so the actual ability to receive the signal probably is about the same.
I,ve only used Nokia and SonyEricsson, but my experience exactlymatches yours. The Nokia just works, and keeps having a clear signal up to the maximum range (no "under water noise") The Sony also sometimes failed to ring, had much poorer sound when the signal became less, and a much les understandable interface, but it was also a lot cheaper.
I realy like my new Nokia N70, but also had other ones in the past that performed fine.
I don't like Siemens either, but have little experience with them.
Try the Volkswagen Transporter. The 2.5 Liter Tdi engine will give you 200 hoursepower, run 100 Mph fully loaded (with 6 windsurfboards on the roof), and still get you about 20km/Liter or 50 Mpg. With the extra seats it will fit 6 people.
A friend of mine has one, because he also has his own company and we did a 4500km (3000mile) trip trough Europe with it last year. The above is the consumption we get while cruising german and french highways at around 130-140 kph (85-90 mph)
I am not sure there is a market for Unix workstations.
Yes there is, I see it at my employer, but it needs a combination the traditional unices lack: - Run all the old Unix stuff (CLI or AWK/Motif/Athena in Fortran60 and what have you), example: AIPS - Run most of the stuff Windows runs (MS Office predominantly) - Look nice (you're paying top dollar) and be easy to use, with any of the shelve PC compatible Printer/Scanner/USB stick/camera/whatever.
Sun is somewhat 'getting' it by providing StarOffice, OOo and the like, but the only remaining Unix Workstation vendor that realy seems on the right track, is the one that's not a traditional Unix vendor at all: Apple with Mac OSX.
I see a lot of the traditional Unix workstation market (We used to have HP, Sun, DEC and SGI) moving exclusively to Apples.
And I too regret SGI passing away, I use OpenGL and C++/STL quite often, and also remember the Indy 166Mhz workstations with 17" screen, nice audio and camera, sitting next to the newest and latest Windows 3.1 486DX-33Mhz machines.
What I find annoying on my PocketPC, is that as long as you only use US english, it performs reasonably well in recognizing my writing and guessing words, but my native language is Dutch. This gives 2 problems: - It tries to guess Dutch words using an US English dictionairy, which is so much of a PITA that I switch off the entire dictionairy function. - Dutch has a few characters that aren't in the standard US character set, this leaves me "international" as the only other option, but this also contains a lot of characters I will never use, and only cause confusion for the OCR system. - Next to that I don't like that it forces you to learn it's alphabet instead of it learning yours.
In short I am very disappointed about my PocketPC, also because of some other limitations I was unaware of when I bought it. (remove battery and it forgets everything, coupled with an ActiveSync backup that doesn't work; I'm lefthanded, which makes the user interface very akward), I now have a Nokia Series 60 phone and prefer that.
That's a significant part of the 'solution', they seem to say that they didn't have the knowledge inhouse to design an infrastructure that met their increasing demands, and they had 'MS consultants' to turn to for a solution. I think there might just be to few 'Linux Consultants' that are capable of providing this kind of service, even if the tools might be out there. Pitching them to management in the right way is a skill in itself.
The biggest problem that faces India is that it's economy is growing, but it's population is growing faster. So everything you've heard about India's booming businesses is true, but still also the number of poor people in the country keeps increasing, which keeps labor cheap.
They did win the case, but went out of business long bfore that.
Re:The future isn't Open Office
on
Office Delayed, Too
·
· Score: 2, Insightful
It is almost like the whole business world learned one piece of software and decided they would do _everything_ with it.
Emacs !!!
I live in a world with a lot of people who think that Emacs is good for everything. Similar to your rant on Office.
In the end most people use office for a lot of things it wasn't meant to do because of the costs associated with buying the likes of Photoshop, QuarkXpress, Matlab, etc. AND the time needed to learn to use those tools.
It took me 4 months at my previous job to get my manager to agree on buying Matlab, you don't know how much of a pain it is to analyse Gbyte datasets in Excel...
Slashdot Mirror
In radioastronomy correlations and FFT's is a large part of the computations we do. Some tests we've done on Nvidia 7900's shows those could be sped up by a factor of 300 over a general CPU. The big problem is bandwidth.
The amount of processing and bandwidth needed for these new telescopes comming online is staggering. For LOFAR we're using a IBM BlueGene with 12.000 Cores (Stella, nr.12 in the top500), using a 144Gb/s connection, and for SKA the nummbers are going to be orders of magnitude larger. The possibilities of these GPGPUs are studied with great interest in the radioastronomy community.
I was never ever bored as a child, I had paper and pencil, lego and an imagination. I would either draw of build.
I still think those made me into the engineer I am today.
This is about PDA's and and TAblet PCs that need to be operated with a pen. I have one experience with this:
When I was looking for a PDA in 2000, I looked at my friends Palm V, which where the scrollbars could be flipped to the left, and the pen holder on the right was identical to the holder of the flipcover on the left, so they could be interchanged.
But the Palm was only 32Mhz, back&white, while the new PocketPCs were fast enough to run mp3, have CF readers and 16bit color displays.
So I bought a PocketPC (Cassiopeia E125).
I can tell you these are VERY awkward to use for lefthanded people. You are constantly moving your hand in front of the screen, the pen holder is on the wrong side, all buttons are on the left side, INCLUDING the scroll wheel, which I _can't_ use, because I'm holding a pen with my left hand.
I have only used my (very expensive) PocketPC for a few months, then I didn't use a PDA for years, but since some time I am now using a Nokia smart phone, and realy liking it. I heard that the newest model PocketPCs have become somewhat better, as the one I had 6 years ago, but I'll never buy another one.
I don't know where you live, but that's not true overhere, we hard a reasonably cold winter, and a very hot spring and summer. (average temperature of 23 instead of 17 degrees and 15 instead of 12)o verzichten/index.html#jaar
http://www.knmi.nl/klimatologie/maand_en_seizoens
rainfall has been 15 instead of 70 mm, so I don't know where you got that from either.
Only after the first of august have temperatures gotten more in the usual range again.
The advantage of the Blue Gene is that it is relatively simple to indeed go low level and manually tune your application at the assembly level. But I agree that a BG isn't the best solution to everything, and there are still some issues IBM needs to work on.
I think about 75% of the programming of the comercial Dutch TV stations (7) I recieve and about 25% on the public Dutch and Belgian TV stations is comming from the USA, UK, Australia, about 5% from Germany. Nothing gets dubbed everything has sub-titles.
A random example:
06.00 Batman
06.20 Duck Dodgers
07.00 Batman
07.25 Roseanne
07.55 Tel Sell
08.25 Children's ward
08.55 TV Shop
09.55 Astro TV
10.30 TV Shop
11.30 Phone party
12.35 Providence
13.35 Children's ward
14.05 Name the word
15.15 Disney feest: Aladdin
16.00 Duck Dodgers
16.25 Word quest
17.05 ER
18.05 Providence
19.00 Strong medicine
20.00 Friends
20.30 Possession
22.30 Extreme makeover Vlaanderen
23.35 Friends
00.05 Queen's supreme
I would think that the combination of:
Novell (SuSE, eDirectory, lot's of enterprise tools and credibility)
Corel (WordPerfect Office, CorelDraw/Photopaint, Paintshop Pro)
Borland (Kylix, jbuilder)
if combined, could make a difference.
But all of them tried and failed on their own, (on Novell te jury is still out), because of different timelines.
The main thing that would be lacking is a good game developement platform.
It might be an interesting statistic for you to know that in the USA three times more people per capita get murdered a year by firearms, as the total number of people murdered in my country per capita (Netherlands).i r_percapp
http://www.nationmaster.com/graph-T/cri_mur_wit_f
http://www.nationmaster.com/graph-T/cri_mur_perca
If substract all firearms murders (0.027) from the total (0.042), you get almost the same number as for the Netherlands (0.011), where owning a firearm is mostly illegal. I know you can easily lie with statistics, but to me it suggests that a country with few people owning firearms, is a lot saver as one where almost everyone owns one.
I just thought to add a link: http://www.lofar.org/p/systems.htm
Those numbers aren't exactly the same as the ones I quoted, it might be the website is out of date with the latest info, or my memory is failing me. They're still the same order of magnitude though. It migh just depend if you quote raw data rate, correlated data rate, or the rate at which the scientific results come out.
Actually, most astronomers use FORTRAN there days. Packages like AIPS and MIRIAD are completely written in them.
:-)
The newer stuff like AIPS++ uses C++.
I'm working on one of these next-generation telescopes, it LOFAR, we hope to have it operational in 2008. All software is written in C++, except for some user interfaces in Java.
The telescope in the topic is only a dream at this point, they have nowhere near the funding to start yet. LOFAR on the other hand is already being build. Our software correlator is already running on our IBM BlueGene, making it the 9th fastest computer in the world. Our 144 GBit/s links to the sub-stations are operational, and the first full substation (of 77) will be operational next month.
These guys are talking 30 TByte/day, we're talking a raw datarate of 1.5 Petabyte/day at the end of 2008. This is going to be the largest radio-telescope in the world, at 300km (200 mi.), at least until SKA gets build (if it gets build)
It's a realy cool project
I'm mainly rehashing what I learned when I visited JET a couple of years ago. I'm more kind of an astrophysics engineer myself.
As far as I understand the whole thing is mostly an engineering issue. With something like an Apollo Moon project style approach it could maybe be done in 10 years. It would have to be a massive national or super national effort.
I think history has shown us that especially where efficiency is the main concern, like it seems to be with the whole fussion problem, it will just take time for people to come up with bright engineering solutions.
And you're right that there's also a lot of politics involved, as always with large sums of money.
American (I think) Should be about 3.8 liters.
USB harddrive
Short explanation:
Try building a heating system or oven entirely out of ice, without melting the ice in the process.
Long explanation:
The problem is threefold:
1) It's very hot (millions of degrees (Celsius)). So you need very good isolation and confinement, or it will boil your facility and cool down itself to the point where it's useless. You have something which is generating energy, it is very, very hot, and therefore contains very fast particles, that want to travel in a straight line, "filling" the vacuum you are trying to use to confine them, and then hitting your device and vaporizing it.
This is the main reason why in the fussion world big is beautiful, because the volume/surface area ratio gets more favorable the bigger you get, so it becomes easier to keep the heat inside.
2) So on the one hand you're trying to create the perfect thermosflask, but then you also want to extract some of that energy in a controlled way. Remind you, this stuff is HOT. It will vaporize anything it comes into contact with, including any heat exchanger you can think of. Basically the only known way is how the sun does it: using (infrared) light. You have the light that gets generated travel accross the vacuum you are using to contain the actual plasma and hit something, like the sunlight can heat something. But you want to control what get's heated and light can not be guided with magnetic fields as plasma particles can. So you need the most fancy mirrors in existence.
3) once you have all of this figured out (and they hope to get close with ITER) you'll need to have a sustained reaction, so you'll also need a way to keep feeding it new "fuel", without destabilizing the entire process. THink of it as chucking new logs on a fire, but these logs are very wet and frozen, because the're millions of degrees colder as your fire. It's very easy to exstinguish the fire.
I find the difference in reception quality between my old Sony and my new Nokia, quite big. I work in an official no RFI zone, in my country one of the few places with very poor reception (GSM). The Nokia does not give these "Under water sounds" that the Sony gives when they're at almost no signal. (same location same provider, only different phone) They both sometimes drop the entire conversation, if I move while on the phone, so the actual ability to receive the signal probably is about the same.
I,ve only used Nokia and SonyEricsson, but my experience exactlymatches yours.
The Nokia just works, and keeps having a clear signal up to the maximum range (no "under water noise")
The Sony also sometimes failed to ring, had much poorer sound when the signal became less, and a much les understandable interface, but it was also a lot cheaper.
I realy like my new Nokia N70, but also had other ones in the past that performed fine.
I don't like Siemens either, but have little experience with them.
Hmmm, your description sounds a lot like what Apple could become for FreeBSD?
Try the Volkswagen Transporter.
The 2.5 Liter Tdi engine will give you 200 hoursepower, run 100 Mph fully loaded (with 6 windsurfboards on the roof), and still get you about 20km/Liter or 50 Mpg. With the extra seats it will fit 6 people.
A friend of mine has one, because he also has his own company and we did a 4500km (3000mile) trip trough Europe with it last year. The above is the consumption we get while cruising german and french highways at around 130-140 kph (85-90 mph)
I am not sure there is a market for Unix workstations.
Yes there is, I see it at my employer, but it needs a combination the traditional unices lack:
- Run all the old Unix stuff (CLI or AWK/Motif/Athena in Fortran60 and what have you), example: AIPS
- Run most of the stuff Windows runs (MS Office predominantly)
- Look nice (you're paying top dollar) and be easy to use, with any of the shelve PC compatible Printer/Scanner/USB stick/camera/whatever.
Sun is somewhat 'getting' it by providing StarOffice, OOo and the like, but the only remaining Unix Workstation vendor that realy seems on the right track, is the one that's not a traditional Unix vendor at all:
Apple with Mac OSX.
I see a lot of the traditional Unix workstation market (We used to have HP, Sun, DEC and SGI) moving exclusively to Apples.
And I too regret SGI passing away, I use OpenGL and C++/STL quite often, and also remember the Indy 166Mhz workstations with 17" screen, nice audio and camera, sitting next to the newest and latest Windows 3.1 486DX-33Mhz machines.
What I find annoying on my PocketPC, is that as long as you only use US english, it performs reasonably well in recognizing my writing and guessing words, but my native language is Dutch. This gives 2 problems:
- It tries to guess Dutch words using an US English dictionairy, which is so much of a PITA that I switch off the entire dictionairy function.
- Dutch has a few characters that aren't in the standard US character set, this leaves me "international" as the only other option, but this also contains a lot of characters I will never use, and only cause confusion for the OCR system.
- Next to that I don't like that it forces you to learn it's alphabet instead of it learning yours.
In short I am very disappointed about my PocketPC, also because of some other limitations I was unaware of when I bought it. (remove battery and it forgets everything, coupled with an ActiveSync backup that doesn't work; I'm lefthanded, which makes the user interface very akward), I now have a Nokia Series 60 phone and prefer that.
Working with Microsoft consultants
That's a significant part of the 'solution', they seem to say that they didn't have the knowledge inhouse to design an infrastructure that met their increasing demands, and they had 'MS consultants' to turn to for a solution.
I think there might just be to few 'Linux Consultants' that are capable of providing this kind of service, even if the tools might be out there. Pitching them to management in the right way is a skill in itself.
The biggest problem that faces India is that it's economy is growing, but it's population is growing faster.
So everything you've heard about India's booming businesses is true, but still also the number of poor people in the country keeps increasing, which keeps labor cheap.
Try looking at Python + PyLab + Matplotlib, unless you're doing 3d graphs, it might be what you need, and in that case I'd suggest looking at ROOT.
Anybody remember Stacker ?
They did win the case, but went out of business long bfore that.
It is almost like the whole business world learned one piece of software and decided they would do _everything_ with it.
Emacs !!!
I live in a world with a lot of people who think that Emacs is good for everything. Similar to your rant on Office.
In the end most people use office for a lot of things it wasn't meant to do because of the costs associated with buying the likes of Photoshop, QuarkXpress, Matlab, etc. AND the time needed to learn to use those tools.
It took me 4 months at my previous job to get my manager to agree on buying Matlab, you don't know how much of a pain it is to analyse Gbyte datasets in Excel...