You can't cheat Amdahl's law anymore than you can
give birth in one month with nine women. The law is a
rather simple idea similar to chemical kinetics, when you think about it. i.e. a rate limiting steps.
Proof of concepts are nice, but when the GPU is firmly planted in the CPU, this will make more sense. The PCI bus can be a bottleneck in these types of situations. AMD fusion is a great example of this idea.
That is the point of the free version. How else do you get people to read and talk about your book in the sea of crap. If it is worth reading it will get noticed because there is no cost barrier. Those, like me that want bound versions of good books will pay for the paper.
Many people first used Windows not by choice, but by mandate -- there was no other option and the Microsoft monopoly made sure it stayed that way. (unless you bought a Mac)
My guess is many people have found the MS experience frustrating and a general PITA, but there was never any other choice. They had to live with the shoddy time wasting experience Microsoft called computing.
Now given the option of having their "desktop experience" on their "phone" or "pad" I am sure many people are interested in real alternatives. My prediction is no matter how hard Microsoft tries to play the "we are the future of computing because we invented everything" song and dance, most users will chose iOS and Android for exactly that reason. Hi-tech karma at its best.
The reasons Linux is so successful on HPC clusters are numerous. Historically, Linux was the best solution for many reasons. In particular, HPC clearly demonstrates why "open" is superior to closed approaches to a markt.
I just wrote and article Low Cost/Power HPC using Atom processors and Mini-ITX boards. The first table is rather interesting (HPC = High Performance Computing):
The Nehalem Xeon runs 1.8 times faster, generates 7.3 times as much heat and costs 22 times as much as the D510 Atom. The performance is 7.7 times faster, but when you factor in the price-to-performance the Atom is 3 times better than the Xeon solution. Interestingly, the TDP/performance ratios are almost identical for both processors.
And Bill, I would say you were not the only company to go down because you believed:
... open source has been great for hobbyists to get involved...
Sun had the ball and dropped it, too bad because it was a good run.
I think most people have the wrong idea about the "Butterfly Effect." IIRC, the weather
scientists were talking about the precision with which they would need to know
air movement to make longer term predictions. i.e. the longer the forecast the more
digits of precision are needed in your measurement. They were referring to the
level of precision and not to butterflies causing a tornado or other such nonsense.
Unfortunately, the media like to report it otherwise.
When you are convinced that immersive liquid cooling is the future, read rule #1
For special and expensive systems such ideas may be
useful, but it is hard to beat air for commodity systems because
it is free, there is plenty of it, it cannot spill, it does not put excess load on the floor, it is inert
(to electronics), and it can be easily recycled.
This is one of the most intelligent posts I have read here in a long time. Good or new ideas are quickly dismissed by the mainstream because they are not, well, mainstream. Almost all good ideas have a champion behind them who works tirelessly to "make it happen."
If you look at the disruptive growth of Linux HPC clusters, you will find that there are no IP agreements. I actually wrote about this: Why Linux On Clusters?. The absence of IP agreements allowed the HPC community to work together and grow faster than anyone imagined. On the famed Beowulf mailing list (started by Don Becker BTW) their is a free exchange of ideas and no one claims ownership of any IP.
I call it a "Lawyer Free Zone" similar to what was proposed in Scotland back in the late 90's.
I entered "clustermonkey", which is a site about HPC Linux Clusters,
and it came back with references just fine. However, it had totally
out of context graphics like flowers and catalogs or something I cannot
really make out. I don't really think this "cute" feature helps at all,
as a matter of fact I think it reduces the quality of the search.
Clustermonkey.net is a cluster geek site, if you search and scan the thumbnails
(which is what attracts your eye) you will be seriously mis-led as to the results.
Obviously, it is try to grab "contextual" graphics that does not work quite right.
First, the Top500 list has plenty of value. What most people do not realize (or should realize) is
it is one data point on the HPC spectrum. If your HPC program does not perform the same or similar matrix operations as HPL then the ranking is meaningless to you. To some the list has become a public relations contest.
Second, performance is virtually independent of the OS (unless you are using TCP). Most big clusters use InfiniBand and run applications in "user space" by-passing the kernel. The rest of the code is crunching numbers.
Third, for the right cost, anyone can get a system on the Top500 list. It is a rather simple price/performance calculation, by the way. Breaking into the top 10 might be a little more difficult.
Finally, HPC and Linux are synergistic. Take a look at Why Linux on Clusters? to get the full story. The Windows model does not work very well in this space.
Not to burst any ones funding bubble, but this has been done.
Take a look at SiCortex.
They did it, they are shipping product, and it works quite well. And it runs Linux.
Slashdot Mirror
Take a look at Limulus systems from Basement Supercomputing: http://www.basement-supercompu...
These are fast low power/noise/heat systems with a fully installed open source HPC software stack.
You can't cheat Amdahl's law anymore than you can give birth in one month with nine women. The law is a rather simple idea similar to chemical kinetics, when you think about it. i.e. a rate limiting steps.
If you are interested in a non-mathematical description of Amdahl's law have a look at http://www.clustermonkey.net/Parallel-Programming/parallel-computing-101-the-lawnmower-law.html
For those who would like a gentle introduction to quantum computing take a look at: A Smidgen of Quantum Computing
quod erat demonstrandum
You may be thinking of this (Limulus Project)
Have a look at The Limulus Project. You will be able to buy one of these real soon.
Been there done that, maybe want to consult Linux Cluster Urban Legends before you continue down this path
Proof of concepts are nice, but when the GPU is firmly planted in the CPU, this will make more sense. The PCI bus can be a bottleneck in these types of situations. AMD fusion is a great example of this idea.
That is the point of the free version. How else do you get people to read and talk about your book in the sea of crap. If it is worth reading it will get noticed because there is no cost barrier. Those, like me that want bound versions of good books will pay for the paper.
Many people first used Windows not by choice, but by mandate -- there was no other option and the Microsoft monopoly made sure it stayed that way. (unless you bought a Mac) My guess is many people have found the MS experience frustrating and a general PITA, but there was never any other choice. They had to live with the shoddy time wasting experience Microsoft called computing.
Now given the option of having their "desktop experience" on their "phone" or "pad" I am sure many people are interested in real alternatives. My prediction is no matter how hard Microsoft tries to play the "we are the future of computing because we invented everything" song and dance, most users will chose iOS and Android for exactly that reason. Hi-tech karma at its best.
Check out the Limulus Project
Forgot this link! Why Linux On Clusters?
The reasons Linux is so successful on HPC clusters are numerous. Historically, Linux was the best solution for many reasons. In particular, HPC clearly demonstrates why "open" is superior to closed approaches to a markt.
Here are some numbers from a post I made last week.
The Nehalem Xeon runs 1.8 times faster, generates 7.3 times as much heat and costs 22 times as much as the D510 Atom. The performance is 7.7 times faster, but when you factor in the price-to-performance the Atom is 3 times better than the Xeon solution. Interestingly, the TDP/performance ratios are almost identical for both processors.
And Bill, I would say you were not the only company to go down because you believed: ... open source has been great for hobbyists to get involved ...
Sun had the ball and dropped it, too bad because it was a good run.
I think most people have the wrong idea about the "Butterfly Effect." IIRC, the weather scientists were talking about the precision with which they would need to know air movement to make longer term predictions. i.e. the longer the forecast the more digits of precision are needed in your measurement. They were referring to the level of precision and not to butterflies causing a tornado or other such nonsense. Unfortunately, the media like to report it otherwise.
For special and expensive systems such ideas may be useful, but it is hard to beat air for commodity systems because it is free, there is plenty of it, it cannot spill, it does not put excess load on the floor, it is inert (to electronics), and it can be easily recycled.
Right on, 5 digits and proud
I interviewed these guys at SC09 for Linux Magazine. There are some close up shots of the servers in the oil.
This is one of the most intelligent posts I have read here in a long time. Good or new ideas are quickly dismissed by the mainstream because they are not, well, mainstream. Almost all good ideas have a champion behind them who works tirelessly to "make it happen."
If you look at the disruptive growth of Linux HPC clusters, you will find that there are no IP agreements. I actually wrote about this: Why Linux On Clusters?. The absence of IP agreements allowed the HPC community to work together and grow faster than anyone imagined. On the famed Beowulf mailing list (started by Don Becker BTW) their is a free exchange of ideas and no one claims ownership of any IP. I call it a "Lawyer Free Zone" similar to what was proposed in Scotland back in the late 90's.
I entered "clustermonkey", which is a site about HPC Linux Clusters, and it came back with references just fine. However, it had totally out of context graphics like flowers and catalogs or something I cannot really make out. I don't really think this "cute" feature helps at all, as a matter of fact I think it reduces the quality of the search. Clustermonkey.net is a cluster geek site, if you search and scan the thumbnails (which is what attracts your eye) you will be seriously mis-led as to the results. Obviously, it is try to grab "contextual" graphics that does not work quite right.
First, the Top500 list has plenty of value. What most people do not realize (or should realize) is it is one data point on the HPC spectrum. If your HPC program does not perform the same or similar matrix operations as HPL then the ranking is meaningless to you. To some the list has become a public relations contest.
Second, performance is virtually independent of the OS (unless you are using TCP). Most big clusters use InfiniBand and run applications in "user space" by-passing the kernel. The rest of the code is crunching numbers.
Third, for the right cost, anyone can get a system on the Top500 list. It is a rather simple price/performance calculation, by the way. Breaking into the top 10 might be a little more difficult.
Finally, HPC and Linux are synergistic. Take a look at Why Linux on Clusters? to get the full story. The Windows model does not work very well in this space.
Not to burst any ones funding bubble, but this has been done. Take a look at SiCortex. They did it, they are shipping product, and it works quite well. And it runs Linux.