• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Computer modeling and simulation

alienentity said:
I agree the hardware alone wouldn't be expensive. No idea what the software costs....and the programming? Not free either.
Click to expand...


Oh, I agree. The hardware was all I was referring to - you would obviously still need people who are experienced with LS-DYNA to run the simulation, etc.
 
deep44 said:
Oh, I agree. The hardware was all I was referring to - you would obviously still need people who are experienced with LS-DYNA to run the simulation, etc.
Click to expand...
The hardest part is understanding the results and Gage, you and a bunch of fringe CTers don't understand NIST and other independent studies so you most likely will fail to understand your own efforts. Hard to get software to show you evidence for thermite and explosives when fire did the deed. At least you have your conclusions in place no matter what you do in the future.
 
deep44 said:
I'm telling you that a regular Linux cluster with comparable hardware specs could be put together for less than $20,000, using commodity hardware. What custom cards are you referring to?
Click to expand...

Add maybe $8k for a low-latency switch and the node adapters that go with it.
 
BigAl said:
Add maybe $8k for a low-latency switch and the node adapters that go with it.
Click to expand...


The cluster would be tied together with gigabit ethernet, so the switch would cost no more than $250.
 
Justin39640 said:
apparently hes never heard the term "government spec"
not cheap
im sure the NIST doesnt shop at newegg or microcenter lmao
Click to expand...


Actually, I'm very familiar with the requirements of government agencies when it comes to technology products. That's why I've specified "commodity hardware" four separate times now, so you would know that I was talking about hardware that you can order from newegg, etc.
 
deep44 said:
Actually, I'm very familiar with the requirements of government agencies when it comes to technology products. That's why I've specified "commodity hardware" four separate times now, so you would know that I was talking about hardware that you can order from newegg, etc.
Click to expand...

Complex nonlinear cluster computation on commodity hardware. Uh huh.

Just what are you planning on computing for your validation/burnin/test runs now? SETI at Home doesn't count for this, ya know.
 
Mr.D said:
Complex nonlinear cluster computation on commodity hardware. Uh huh.

Just what are you planning on computing for your validation/burnin/test runs now? SETI at Home doesn't count for this, ya know.
Click to expand...


Are you claiming that it wouldn't work at all? I don't understand what you're trying to say. If so, please explain.

As for testing, I would leave it up to the person or persons with LS-DYNA experience. Personally, I would at least want to see them run a few of the benchmarks that are posted at Top Crunch - car2car, etc.
 
deep44 said:
Are you claiming that it wouldn't work at all? I don't understand what you're trying to say. If so, please explain.

As for testing, I would leave it up to the person or persons with LS-DYNA experience. Personally, I would at least want to see them run a few of the benchmarks that are posted at Top Crunch - car2car, etc.
Click to expand...

The commodity PCs are fine but you don't get your money's worth out of them without a low-latency cut-through switch and node adapters and these aren't cheap.
 
deep44 said:
Are you claiming that it wouldn't work at all? I don't understand what you're trying to say. If so, please explain.

As for testing, I would leave it up to the person or persons with LS-DYNA experience. Personally, I would at least want to see them run a few of the benchmarks that are posted at Top Crunch - car2car, etc.
Click to expand...

example:
man A: "im gonna get into drag racing"
man B: "great, what kinda car you got?"
man A: "a Hyundai Elantra :)"
 
deep44 said:
Are you claiming that it wouldn't work at all? I don't understand what you're trying to say. If so, please explain.
Click to expand...

I'm saying that you don't just throw together a bunch of off the shelf hardware and call it a high-performance cluster. Do you know what the typical memory error rates are for desktop PC's? How do single bit errors propagate through highly nonlinear calculations that take days or even weeks?
 
These last, saay, 10 posts sounded like this to me.


shdgashgi 9 wr oihgah ro 983 4t 98rugh 9reruiurhgg9834 4teurrhiuerhg9838t eurguhrgugiuig urhh ifdfhgoih943 gfh398hrrguege



WHAT THE ??????

I thought that my Dell Inspiron b130 with 1 gig, plus the 512 that came installed was up to date. Aam I wrong???? <~~~ Slight tongue in cheek here.
 
deep44 said:
Are you claiming that it wouldn't work at all? I don't understand what you're trying to say. If so, please explain.

As for testing, I would leave it up to the person or persons with LS-DYNA experience. Personally, I would at least want to see them run a few of the benchmarks that are posted at Top Crunch - car2car, etc.
Click to expand...

OK, deep! Got it - it's a snap, easy as falling off a log and cheap too.
You might even get free financing for 6 months to boot.

Let us know when you've got your results. But remember this little tip: the more science you get, the less CD you get....funny how that works.

I can just see the headlines on USA Today: '9/11 Truthers accidentally confirm NIST results on WTC collapses'
National Inquirer 'Jim Fetzer fakes suicide and is spotted in Bolivia with Ace Baker'
 
alienentity said:
Let us know when you've got your results. But remember this little tip: the more science you get, the less CD you get....funny how that works.
Click to expand...


How could I get results if they haven't released the model? We're talking about a hypothetical situation.
 
deep44 said:
How could I get results if they haven't released the model? We're talking about a hypothetical situation.
Click to expand...

you build your own
you think NISTs data is flawed anyway
what would you need those numbers for

build your own cluster for your 20 grand (LMAO)
and build your own model

let us know how it came out when its done in 2014 lol
 
Justin39640 said:
you build your own
you think NISTs data is flawed anyway
what would you need those numbers for

build your own cluster for your 20 grand (LMAO)
and build your own model

let us know how it came out when its done in 2014 lol
Click to expand...

Well, assuming they had the time to reprogram some of the code, they could possibly get something together using X58 boards and Tri-SLI.

Assuming they tested mounds of GPUs for as few memory errors as possible, they could probably jury-rig a farm of i7's on X58 boards in Tri-SLI using 3xGTX 295's. A single Tri-SLI rig alone could probably crank out 6000 GFLOPs.

Of course, you could probably cook an egg on the heat that a farm of rigs would create.

And you'd also need the programming expertise to convert code to something such a rig could easily utilize.

But you'd still need more than $20k in parts. You'd probably need a few grand alone just to get an adequate cooling setup.
 
You must log in or register to reply here.

Back
Top Bottom