You are currently browsing the monthly archive for July 2006.

A Japanese supercomputing center has laid claim to the holy grail of computing. They have developed and implemented the worlds first Petaflop computing system. Called MDGRAPE-3. (Molecular Dynamics GRAvity PipE.)I’ll drop some deeper, darker statistical and technical details later. But, the 10 second overview is that the system built at Riken in Yokohama Japan, will break the Petaflop barrier, or million billion floating point operations per second, and may even achieve as much as 1.4 Petaflops, according to one source. To put this in perspective, BlueGene/L, the US’s fastest system, is currently running at 280 Teraflops, or Trillion Floating Point operations per second, which means the fastest US based computer system will come in at about 1/6 the speed of the new Japanese system.

The system is built around the idea of taking standard computing systems and adding specialized extremely fast accelerator chips. These chips will only work well with specific types of codes, but that’s the whole idea for supercomputers, so this isn’t a downside really. In this case the special codes will be things dealing with N-Body problems, (explanation shortly), which is needed for things like galaxy simulation (how our world works), protein folding (picture HIV vaccine), Molecular Dynamics (Mach 5 commuter jets), etc etc. The idea of using specialized accelerator cards is also taking hold here in the US. There are several companies developing specialty accelerator cards, ClearSpeed, being one of them. There is also a lot of research and developement into developing usefull FPGA, or Field Programmable Gate Array accelerators. Cray is doing quite a bit of research into this along with some other new and upcoming companies. So, the US is behind, but at least we’re still in the race.

The new system will be built around 201 systems with 24 MDGRAPE-3 LSI chips. (This supposedly gives 4808 processors, my math says 4824), with another 64 4way Xeon servers (256 procs total) , and 37 2way Xeon servers( 74 procs total). In the scheme of things this it tiny for large HPC systems, very tiny.

Lets put this in perspective. BlueGene/L (US Leader) vs. the new MDGRAPE-3

Number of processors

BlueGene/L 130,000 Processors

MDGRAPE 4808 Processors (1/27 the number of processors)

Total Floating Point Operations per Second (FLOPS)

BlueGene/L 280,000,000,000 per second

MDGRAPE 1,400,000,000,000 per second (6 times faster)

Cost per flop

BlueGene/L $140 per gigaflop

MDGRAPE $15 per gigaflop (1/8 the cost per flop roughly)

Energy Efficiency

BlueGene/L 6 Watts per Gigaflop

MDGRAPE .1 Watt per Gigaflop (1/60th the power consumtion)

Total Cost

BlueGene/L Over $100 Million. (Research cost included with price)

MDGRAPE $9 Million (Lot’s of cost absorbed by NEC/Intel/SGI as chip developement costs)

Summation: Japanese 5, US 0. We’ve been spanked. At least for now.  

Now, what is an N-Body problem. Imagine modeling the evolution of the universe in a computer. To do this right you need to take ALL billion or so stars, (you can cheat and do a rough with only a few million representative stars) get their current position in relation to all the other stars in the universe, calculate the forces between them, then move forward one time step, calculate the forces of all stars in relation to all other stars, move forward one time step, repeat, repeat, etc. This basically means you have to do a calculation of N^2 for each time step. If you use a million stars that means 1,000,000,000,000 calculation per time step for a small representative number of stars. Ouch.

Now take the same idea and picture a few hundred million atoms in a protein, (say for HIV research, cancer, or my favorite, diabetes research, etc). As the protein folds, every atom will effect how all the other atoms in each time step move. So, you once again have an N-Body problem. Now imagine your trying to develope a protein that folds to a specific shape so that if fits the receptor of a cancer cell. You might have to try thousands or millions of different initial states to find the one that ends in the needed final state. Your looking at millions of calculations per trial and millions of trials. And let’s not forget that the bigger the protein the more the atoms the bigger (by a square) the calculations needed. Again, ouch.

 

There are several US companies working on systems similar to the MDGRAPE-3. I’m looking forward to seeing their response to this.

 

 

Advertisements

Well an interesting item has come up and now a decision needs to be made. Not in a rush mind you. But one needs to be made in the next bit of time. This may sound insane to most normal lower 48 folk. But would you give up sunny California for the cold miserable dark winters of interior Alaska?

Sunset over the Pacific                            For the Midnight sun?

Sunset over Monterey Bay   Midnight Fishing

How about this view                                 For this view?

Beach View   View from the Radar Site

 Or this Trail                                             For this one?

Bike Trail   Trail In

 

 

Well today is the day John heads back to Iraq. His third tour there. After being home for a little over six months he is headed back to teach the Iraqi military how to do artillery. Go figure. Good luck my friend and I’ll keep the beer cold til you get back.

 Rich

 

Well,

     I have finally started my dive class. I ended up going with Bamboo Reef in Monterey. It’s actually on the edge of Monterey and Pacific Grove. Not far at all from the house. Bamboo Reef is an SSI dive shop. I was a little concerned about this because I’ve only ever been aware of PADI. But, there seems to be multiple certification agencies now. PADI, SSI, and NAUI just to name a few. I also verified that the PADI dive shop, Aquarius, would accept the SSI Open Water Cert for entry into their multiple specialty dive Certs. They do.

     One thing that annoyed the crud out of me is the fact that the first chapter in the SSI Open Water book was more of an infomercial than anything else. I finally go so tired of seeing ‘Your Local SSI Dealer’ will help you purchase “fill in the blank” that I got out the good old Sharpie and started crossing them out. 26 times in the first chapter. I was getting frustrated enough that I started calling other dive shops to see what PADI and NAUI were like. I found out they’re damn near as bad. What a crock. Oh well. I just have to put up with the marketing crap for two weeks and then I can do what ever I like.

      Another Rant. Because I’m a type 1 diabetic I have to fill out a special medical form and have my doctor say I can dive. Okay, I can see that, it makes sense. However, I also have 4 pages of “SSI Dive Protocol” I have to fill out and follow. Why this is such a load of bull I shall shortly explain. It seems that the “Dive Protocol” was developed by a doctor, Duke H. Scott,  back in 1995, updated (some) in 2001. This protocol is designed for a diabetic on the MDI program, or multiple daily injections. Which means they take a long acting insuling for their basal rates and a short acting for thier quick acting. What they do NOT take into acount are diabetics on an Insulin Pump, like me, or those using some of the newer ultra fast acting insulins. I tried to explain this to the Dive Shop owner. Didn’t care. Needs the stupid protocol followed. I explained it won’t work for me. Didn’t care. So I told him I’d have it all filled out by the time it came to my pool and open water sessions. In other words, I’ll treat it like a theoretical exercise. I’ll make up some numbers that look good, and then do what I need to do to be safe. Life sure would be easier if people would just have a little flexibility and not be so stuck in the box. Grrrrrr….

     I know, I know. Shut up and go diving. At least I get to go diving. By the way. I would include pictures, but my digital camera has gone to the big toaster in the sky. The lense motor died. Took it apart, but, it’s a gonner. Oh well. I’ll pick up another camera again soon so that I can take pics of my diving experiences.

 

Well, I haven’t posted much in the last week or so. Been somewhat brain burned. Still reading the “Federal Plan for High-End Computing.”  I’ll post some high lights on that when I get somewhere with it.

 

 I also plan to do some research into the Global Warming hype and see if I can’t point out some debunking information. I’ll get to that when I can.

 

 But of greatest importance. The girls and I ended up getting masks and snorkels to go with our wet suits. So now we can not only go surfing and boogie boarding we can also go snorkeling in the fridged waters of Monterey Bay. In fact we had our first outing on Sunday. We went down to the beach at Lovers Point here in Pacific Grove.

 

Waters were a touch chilly, mid 50’s I’d say. Waves were pretty calm, but the water was still pretty murky. The ladies stayed pretty close to shore only adventuring out to about armpit depth on them. I decided to go a touch further and see what I could see. I did see two fishies, pretty small ones. One looked like a perch and the other like an oversized sardine. Not too exciting, but hey, it was my first outing.

 

What was really strange was to be floating along the top with the sea grass waving 4-5 feet down while being surrounded by kelp, all this while in murky water. Makes ones imagination get going. ie, gee, where’s the shark? hehe… I’m hoping to hit different locations around  the Pacific Grove area. Just snorkel around and have some fun. A co-worker and I will be taking our dive classes starting on the 18th of this month at the bamboo reef dive shop. I have all the paper work signed by the doc, I had to get a bunch of forms signed because of my diabetes.

 

I did ask the question of Minimed (the maker of my insulin pump) what information they had about pressure differences effecting a pump. I know soaking them in salt water is bad for their functionality, but I figure if I wore a dry suit then the pump wouldn’t get wet, so why couldn’t I just keep wearing it. They have NO information on what pressure differences would do to a pumps performance. So, I’ll just take it off. Oh well, at least I get to dive, so I won’t complain. At least not too loudly. (-:

 

Anyhow. That’s what’s up for now. I’ll try to get some pictures of the girls snorkeling for fun. When I do I’ll put em up.

 

 

 

I was going over the 2007 budget supplement for the “Network and Information Technology Research and Developement Program“, Ya I need a life I know, anyhow, I found a few intereresting items mixed in with all the jargon and double talk.

 As a quick overview. The NITRD program is a blanket organization over a lot of the high performance computing groups in the government. In fact the NITRD program controls the budgets, etc of eight major “Program Component Area’s” or groups as it were. These are.

  1. High End Computing infrastructure and applications.
  2. High-end computing research and developement
  3. Cyber Security and information assurance
  4. Human-Computer interaction and information management
  5. Large-Scale networking
  6. High-Confidence Software and systems
  7. Social, economic and workforce implications of IT and IT workforce developement
  8. Software design and productivity

 To give you an idea of the government agencies that are effected by this here’s a short list. NSF (National Science Foundation), NIH (National Institute of Health), DARPA (Defence Advanced Research Projects Agency), NASA, NSA, DoD, DOE, NOAA, EPA, OMB, etc etc etc. We’re talking some big hitters here. ALL are effected by this supplemental budget.

Some of the highlights of the new budget are.

  • Chartering of a High End Computing working group. This new group will report to the NITRD Subcommittee.
  • Chartering of the Cyber Security and Information Assurance Working Group.
  • Reporting from the High Performance Computing Modernization Program Office (HPCMPO, whom I used to work with)
  • A doubling over the next 10 years of the budgets that support basic research in the physical sciences and engineeering.

Highlights that I personally found interesting:

     In the High End Computing Infrastructure and Applications Working Group.

  1. NSF: Petascale resources by 2010. (That’s almost a factor of 10 increase in performance over our current large scale systems)
  2. NERSC to aquire a 100-150 Tera Flop computing system.
  3. ORNL to upgrade current systems to over 250 TeraFlops. (Here’s a plug for an ex co-worker of mine. Go Richard Barrett! The man is a Nuclear Hero. Truly, he has the plaque and all.)
  4. ANL to get a 100+ TeraFlop Blue Gene system.
  5. NSF to have a new Office of Cyberinfrastructure.

This list goes on and on. This is truly a good sign for us. Getting High End computing systems out to the different agencies and departments means more research is done, which means new and exciting items get created, and it means I get to keep playing with some really cool toys. (-:

 

Now one VERY interesting item in the budget supplement was the complete lack of mention of any specific computer company or computer company project, outside of purchases/upgrades that is. With one very glaring exception. Three Cray  projects are mentioned by name. THREE, and not another computer company mentioned AT ALL. Hmmmm………

High End Computing (HEC) Research and Developement (R&D) Working Group

  • Vector processor system: Continue cooperative developement- NSA, with other NITRD agencies. (As far as I know Cray is still the ONLY U.S. company making/designing vector machines.)
  • Black Widow performance reviews: Assess progress on developmental milestones- NSA, with DARPA, DOE/NNSA, DOE/SC, NASA, NSF, OSD. (Why do I say this  is Cray? Well, the Black Widow project is the follow on project to the Cray X1e supercomputer no less. I think this is about the only U.S. based vector system currently under serious development, and there are a couple of three letter inteligence agencies that just love love love! the way vector systems run their code.)
  • NSA: Eldorado– work with vendor on XT3 modifications, fully funding development in 2005-2006, available in 2006-2007. (The Eldorado project is a modification on the Cray XT3 system which is the commercialized version of the Red Storm system at Sandia National Labs. By the way, did you catch that fully funding part?)

So, Cray is looking pretty good considering they were darn near ready to go belly up not too long ago. In fact things were so ugly they put an ex IBM’r in charge. Someone I’ve actually met no less. Go Peter GO! Peter Ungaro left IBM to be a VP at Cray and before you know it , WHAM!, he’s the CEO and President. I’ve heard many rumours that the Govt. wouldn’t let an asset like Cray go poof, and gee, what do you know……

 Well, enough for now. I need to get back to making money and paying bills. By the way, next on my reading list is the “Federal Plan for High-End Computing.” I’ll do a partial review of that shortly. Till then, Ta ta….

 

Blog Stats

  • 60,514 hits
July 2006
S M T W T F S
« Jun   Aug »
 1
2345678
9101112131415
16171819202122
23242526272829
3031  
Advertisements