Isn't it almost Summer? A couple of weeks to go. Just the normal severe storms here in Brisbane. Thirty degrees centigrade and ice falling from the sky to blanket the ground. Normal.
For the last two months, I've been paying $103/month (australian) for 3MB/512Kb (21GB) which has not even been connected. I'd just love to get an ACK at this stage. And I live within 3 kilometres of the second largest CBD in Australia.
Completely different architectures. Linux for one type, Unix for the other. I prefer the unix any day of the week. Many problems I work on cannot be done on the clusters. You are referring to different battles.
Man, you are one lucky bastard. I work with Australia's SGI machine (#73) and some others (not even worth mentioning). I'd love to be aware of shit like that. I'll continue with the apprenticeship and then go overseas.
I'm sure their accuracy is far greater than 330ft... they just don't want to alarm people [insert snigger here]
I believe this information would be useful to the signal processing to provide feedback to the algorithms determining the length of time traffic lights stay at a particular colour. This helps automate the process given that the edge detection algorithms used to process the images coming from traffic cameras is far more expensive and determinate of the watching control person [dull job]. The "turn left" green light may stay on and people will move to that lane. Then the algorithms can divert to another route.
Is this not how Internet pipes work? [conceptually]
I seriously doubt that their accuracy is 330ft. I've written code that triangulates my phone to within 2 metres and this is only knowing the location of some mobile towers.
This said, you can overlay the "bleeps" onto a road map and use some funky edge detection software to derive all.
Update regarding how truely cool SGI is. This is FYI to show how SGI FPGA's are able to provide services for Altix users. SGI has been working very closely with Mitrionics
Lifted from email received 5 minutes ago:
We have recently announced our Mitrion-C Open Bio Project, which will
include key bioinformatics applications such as BLAST, Smith-Waterman
and Hidden Markov Models - all open source (forget the black box
solutions!) and free to use and tweak to fit your projects. The first
application being released at sourceforge.net, the world's largest open
source community, is BLASTN for the Mitrion(tm) Virtual Processor and
SGI(r)
RASC(tm) RC100 Blade.
If you plan to visit SC06, we would like to give you a live
presentation, because this application is an impressing and real world
example of how different and superior the Mitrionics' software approach
to FPGA Supercomputing is compared to the hardware design approach.
And if you are not into bioinformatics, maybe we could entice you to
visit us by letting you know that if you respond to this email, we will
give you your own Mitrion SDK (Software Development Kit) Personal
Edition, so that you can experience the Mitrion advantage yourself.
*JUST RELEASED:
BLASTN FOR THE MITRION VIRTUAL PROCESSOR AND SGI^(r) RASC(tm) RC100
BLADE*
BLAST is used worldwide for similarity searches for genes and proteins
and is the main tool for data mining in large databases in molecular
biology.
The FPGA-accelerated Mitrion BLASTN program is based on the BLAST source
code from the National Center for Biotechnology Information (NCBI). NCBI
BLAST users will recognize the user interface, file formats and output
options. So if you like the BLAST you have used up to now - you will
love Mitrion BLAST.
You can access the Mitrion BLAST source code and start your own
development project, at www.sourceforge.net.
http://www.sourceforge.net/> BLAST for the Mitrion Virtual Processor and
SGI^(r) RASC(tm) RC100 Blade will be presented at the SGI booth (#905).
*ABOUT MITRION AND MVP*.
The revolutionary Mitrion(tm) Virtual Processor (MVP) and
software-centric
Mitrion-C programming language makes FPGA (Field Programmable Gate
Arrays) Supercomputing accessible to a broader market of scientists,
researchers and software developers than the traditional hardware
approach to FPGA computing.
Since the MVP was launched it has been tested and selected by many of
the world's leading supercomputing sites, such as National Center for
Supercomputing Applications (NCSA), Oak Ridge National Laboratory and
Naval Research Laboratory (NRL). For all our customers Mitrion unleashes
the massive performance benefits of hardware application acceleration
with FPGAs.
*WELCOME TO BOOTH NUMBER #2045 *
If you are going to SC06 in Tampa, Florida November 11 - 17, you don't
want to miss seeing the incredible software that unleashes the power and
promise of FPGA Supercomputing, and get your hands on a free Mitrion SDK
Personal Edition. Make sure you reply to this email, and we'll be sure
to reserve a copy for you. Come see us at SC06, booth # 2045, in lovely
Tampa, Florida!
I for one don't mind the Altix machines. For a shared memory system (yes, I have neither the time nor patience for distributed systems) price and support is way ahead of its competitors.
Was playing with a Prism for a while and it gave me a pretty much constant boner.
I for one am glad that SGI has emerged from chapter 11. I think you need to better understand their full product line before you jump on the band-wagon, blend into the crowd and slay 'em.
If I were releasing scientific visualisations regarding some simulation science (my guess regarding the images), I'd be making sure that those images looked right. The second is a little off, making me question the model they're using.
When I studied hard, I studied hard from a book. That is the way my brain works. When I research, I research with a starting point from Google. While these days (sound like an old fart but am twenty nine) I tend not do much in the way of solid theory, I do do a lot of application work. In my area, application trends can last a maximum of three months before they are no longer *new*. This is too long for printed media. The boom is over before it is printed. Keep in mind that I live in "the arse end of the earth", Australia.
This (public) university is not censored. The librarian will always be around. You need people who study research trends and make those trends available.
Could not agree more. The stats quoted are typical of trash journalism. The 25% of energy consumed probably does not take into account exported goods (and the like). I imagine that quite a bit of energy goes into the production of automobiles (for example) which are then exported. This is not exactly consumed by the American people.
Look, as a non-American, I'd like to see America (the Government, that is) clean up its act a little, but crappy reporting like this is going to have a negative impact (or rather, it should).
I'm not sure about my budget being able to afford them... but that is not the point. The company I'm employed with would definately pay for a whole bunch of 'em. If there were some way some logic part of this CPU could flag the specific memory structures used in the algorithm to know how to farm the job, my company would pretty much pay anything. The time savings would more than pay for the chip.
Finally finished having a little play. Did a bitonic sort of 4.19x10^6 floats in 0.09s. LU decomp with partial pivot (n = 2000) in 1.2s. Does beg the question, why don't they put this power into ordinary CPUs? I understand only special algorithms will work, but, damn!
Slashdot Mirror
50 percent of college students are below average intelligence.
Isn't it almost Summer? A couple of weeks to go. Just the normal severe storms here in Brisbane. Thirty degrees centigrade and ice falling from the sky to blanket the ground. Normal.
I was not aware that tone, timbre, and harmony were orthogonal to each other.
"Smoking is not addictive"
For the last two months, I've been paying $103/month (australian) for 3MB/512Kb (21GB) which has not even been connected. I'd just love to get an ACK at this stage. And I live within 3 kilometres of the second largest CBD in Australia.
Awesome!
Thanks for the effort! Good job.
Completely different architectures. Linux for one type, Unix for the other. I prefer the unix any day of the week. Many problems I work on cannot be done on the clusters. You are referring to different battles.
Man, you are one lucky bastard. I work with Australia's SGI machine (#73) and some others (not even worth mentioning). I'd love to be aware of shit like that. I'll continue with the apprenticeship and then go overseas.
Sorry to hear that dude ... a really, REALLY tough break. I'll think twice about taking a wizz in the future. I used to enjoy the outdoors.
But my window does not have a sign "Please feel free to enter" next to it
Not a bad analogy ... but from what I've seen, anything but *random* movement.
I'm sure their accuracy is far greater than 330ft ... they just don't want to alarm people [insert snigger here]
I believe this information would be useful to the signal processing to provide feedback to the algorithms determining the length of time traffic lights stay at a particular colour. This helps automate the process given that the edge detection algorithms used to process the images coming from traffic cameras is far more expensive and determinate of the watching control person [dull job]. The "turn left" green light may stay on and people will move to that lane. Then the algorithms can divert to another route.
Is this not how Internet pipes work? [conceptually]
I seriously doubt that their accuracy is 330ft. I've written code that triangulates my phone to within 2 metres and this is only knowing the location of some mobile towers.
This said, you can overlay the "bleeps" onto a road map and use some funky edge detection software to derive all.
Those ads are right up there with actors with spotless faces promoting [insert brandname here] acne treatment. Bastards
Update regarding how truely cool SGI is. This is FYI to show how SGI FPGA's are able to provide services for Altix users. SGI has been working very closely with Mitrionics
Lifted from email received 5 minutes ago:
We have recently announced our Mitrion-C Open Bio Project, which will include key bioinformatics applications such as BLAST, Smith-Waterman and Hidden Markov Models - all open source (forget the black box solutions!) and free to use and tweak to fit your projects. The first application being released at sourceforge.net, the world's largest open source community, is BLASTN for the Mitrion(tm) Virtual Processor and SGI(r) RASC(tm) RC100 Blade.
If you plan to visit SC06, we would like to give you a live presentation, because this application is an impressing and real world example of how different and superior the Mitrionics' software approach to FPGA Supercomputing is compared to the hardware design approach.
And if you are not into bioinformatics, maybe we could entice you to visit us by letting you know that if you respond to this email, we will give you your own Mitrion SDK (Software Development Kit) Personal Edition, so that you can experience the Mitrion advantage yourself.
*JUST RELEASED: BLASTN FOR THE MITRION VIRTUAL PROCESSOR AND SGI^(r) RASC(tm) RC100 BLADE* BLAST is used worldwide for similarity searches for genes and proteins and is the main tool for data mining in large databases in molecular biology.
The FPGA-accelerated Mitrion BLASTN program is based on the BLAST source code from the National Center for Biotechnology Information (NCBI). NCBI BLAST users will recognize the user interface, file formats and output options. So if you like the BLAST you have used up to now - you will love Mitrion BLAST.
You can access the Mitrion BLAST source code and start your own development project, at www.sourceforge.net. http://www.sourceforge.net/> BLAST for the Mitrion Virtual Processor and SGI^(r) RASC(tm) RC100 Blade will be presented at the SGI booth (#905).
*ABOUT MITRION AND MVP*. The revolutionary Mitrion(tm) Virtual Processor (MVP) and software-centric Mitrion-C programming language makes FPGA (Field Programmable Gate Arrays) Supercomputing accessible to a broader market of scientists, researchers and software developers than the traditional hardware approach to FPGA computing.
Since the MVP was launched it has been tested and selected by many of the world's leading supercomputing sites, such as National Center for Supercomputing Applications (NCSA), Oak Ridge National Laboratory and Naval Research Laboratory (NRL). For all our customers Mitrion unleashes the massive performance benefits of hardware application acceleration with FPGAs.
*WELCOME TO BOOTH NUMBER #2045 * If you are going to SC06 in Tampa, Florida November 11 - 17, you don't want to miss seeing the incredible software that unleashes the power and promise of FPGA Supercomputing, and get your hands on a free Mitrion SDK Personal Edition. Make sure you reply to this email, and we'll be sure to reserve a copy for you. Come see us at SC06, booth # 2045, in lovely Tampa, Florida!
Awesome! Well said. I agree whole heartedly.
I for one don't mind the Altix machines. For a shared memory system (yes, I have neither the time nor patience for distributed systems) price and support is way ahead of its competitors.
Was playing with a Prism for a while and it gave me a pretty much constant boner.
I for one am glad that SGI has emerged from chapter 11. I think you need to better understand their full product line before you jump on the band-wagon, blend into the crowd and slay 'em.
If I were releasing scientific visualisations regarding some simulation science (my guess regarding the images), I'd be making sure that those images looked right. The second is a little off, making me question the model they're using.
... ?
I know, a bit of a tough bastard, but
Hmmmm ....
When I studied hard, I studied hard from a book. That is the way my brain works. When I research, I research with a starting point from Google. While these days (sound like an old fart but am twenty nine) I tend not do much in the way of solid theory, I do do a lot of application work. In my area, application trends can last a maximum of three months before they are no longer *new*. This is too long for printed media. The boom is over before it is printed. Keep in mind that I live in "the arse end of the earth", Australia.
This (public) university is not censored. The librarian will always be around. You need people who study research trends and make those trends available.
More snobs ... great, that's the solution.
The net gain calculation is incorrect. I get a net gain every 9.6 seconds. Perhaps they used a GPU to keep up?
Could not agree more. The stats quoted are typical of trash journalism. The 25% of energy consumed probably does not take into account exported goods (and the like). I imagine that quite a bit of energy goes into the production of automobiles (for example) which are then exported. This is not exactly consumed by the American people.
Look, as a non-American, I'd like to see America (the Government, that is) clean up its act a little, but crappy reporting like this is going to have a negative impact (or rather, it should).
I'm not sure about my budget being able to afford them ... but that is not the point. The company I'm employed with would definately pay for a whole bunch of 'em. If there were some way some logic part of this CPU could flag the specific memory structures used in the algorithm to know how to farm the job, my company would pretty much pay anything. The time savings would more than pay for the chip.
Finally finished having a little play. Did a bitonic sort of 4.19x10^6 floats in 0.09s. LU decomp with partial pivot (n = 2000) in 1.2s. Does beg the question, why don't they put this power into ordinary CPUs? I understand only special algorithms will work, but, damn!