More: The $8 million award

[Art Carlson]

I can't figure out why a .8T pulsed machine is going to cost $8 mil. Unless it is very big. Say 2 m or larger coils.

This is all about giving pB11 a realistic chance...

Throwing money at it won't solve that one. Try repealing the laws of physics.

[BenTC]

Alternately, if they have a computer guru on board, and wanted to setup a internet based massively parallel network like SETI and others have done, I'd sign up. For that matter, how do home made super computers ( like parallel Playstation 3's) compare to the best current supercomputers, or the supercomputers of the early 90's ?

Wish that it wasn't so, but a SETI-like network isn't suitable since the problem can't be divided up into independent parts.

The Playstation 3 is around 200gigaFLOPS (1E9).
Hereis a comparison of PC systems.

The HPC petaFLOPS (1E15) mark was passed in 2008.
Hereis a comparison of HPC systems (page 3).

This other discussion pegged useful simulation requirement at 1E34.

[bcglorf]

I can't figure out why a .8T pulsed machine is going to cost $8 mil. Unless it is very big. Say 2 m or larger coils.

This is all about giving pB11 a realistic chance...

Throwing money at it won't solve that one. Try repealing the laws of physics.

Aren't you overstating the case just a little? Is our understanding of plasma physics really so in depth and complete that net power from pB11 has been proven to be contrary to the laws of physics? I'm inclined to believe that if it was that fundamentally proven there wouldn't be guys like Nebel even contemplating the idea.

Your understanding of this is way beyond what I can grasp, but it seems from what I can follow that it has only been demonstrated that there are many situations in which net power from PB11 is impossible. It does NOT look like it has been fully disproven, so as long as the Polywell could potentially operate outside the conditions that have been proven as impossible, it still has a chance, however very remote that chance might be.

I guess I'm mostly making a long winded statement of the difference between impossible due to the basic laws of physics, and just extremely difficult if it is even possible at all.

[bcglorf]

Or perhaps more to my point, there seems to be at least a single counter example that should produce net power from pB11. Getting together a solar mass worth of pB11 soup and placing it in close proximity somewhere out in deep space. It's ridiculously impractical, but the basic laws of physics should still allow that method to generate net power from pB11.

[Aero]

I remembered! Google is my friend! Here is an example of cost breakout that I found on the Internet.

Company A Budget for 2003
Total Direct Cost = $60K
Direct Labor = $50K
Direct Materials = $10K
Total Indirect Cost = $75K
Fringe Pool = $25K
Overhead Pool = $30K
Material Overhead Pool = $5K
G&A Pool = $15K

Fringe: $25K / $50K = 50%
Labor Overhead: $30K / ($50K + $25K) = 40%
Material Overhead: $5K / $10K = 50%
G&A: $15K / (50+10+25+30+5) = 12.5%

Using this example, bid Labor might look like this:
$50K direct labor * 1.50 fringe * 1.40 O/H * 1.125 G&A = $118,125 (before profit)

And bid ODCs might look like this:
$10K ODCs * 1.50 Material O/H *1.125 G&A = $16,875 (before profit)

Direct Labor is for people working to directly advance the contract objectives, and Direct Materials are magnets, ion guns, most test equipment, materials (but not supplies) needed in order to perform the contract.

In the above example, profit comes out at the end, but in EMC2's contract it is a fixed fee, so profit can be considered to come off the top. In the above example, the contract bid would be
$118,125 + $16,875 + 10% profit = $148,500 for the $60,000 project.
If I were to use these ratios on EMC2's projected $12.3 M of contract money, then I calculate $4.97 M available for the actual project in direct labor and equipment. Put another way, $4.97 M is the money that MSimon would have to pay his own salary and to do the project, if he did it alone without any help whatsoever. He could use independent contractors to help him, but if he did then he would need to do his own payroll and liability insurance, which costs money and time away from directly working the project.

But I think the Company A example above is for a pretty cut and dried operation. High tech companies have higher rates (more vacation, days off, better medical) to attract the best people, higher operating costs (security), and higher materials risk, hence materials overhead, and higher R&D, which I doubt the Company A example even considered.

[Art Carlson]

This is all about giving pB11 a realistic chance...

Throwing money at it won't solve that one. Try repealing the laws of physics.

Aren't you overstating the case just a little? ...

I guess I'm mostly making a long winded statement of the difference between impossible due to the basic laws of physics, and just extremely difficult if it is even possible at all.

Well, if you insist. I just got ticked off at the suggestion that p-B11 has a "realistic" chance in any meaningful sense of the word. The lowest bar - and that is plenty high enough - is producing net energy with D-T in any machine less expensive than ITER. That's "what this is all about".

[TallDave]

Well, if you insist. I just got ticked off at the suggestion that p-B11 has a "realistic" chance in any meaningful sense of the word.

Rick seems to think the problems are manageable. So do the FRC guys. And the Navy reviewers were sufficiently intrigued to make the concept an add-on to WB-8/9.

It certainly looks impossible in a tokamak. I'm tempted to say "sour grapes," but the challenges are obviously formidable in any scheme.

[alexjrgreen]

The lowest bar - and that is plenty high enough - is producing net energy with D-T in any machine less expensive than ITER. That's "what this is all about".

Los Alamos beat that a long, long time ago.

[TallDave]

Interestingly, except for the tiny bit of data from the Russian picosecond laser experimeent, all the p-B11 stufff is purely theoretical.

WB-8.1 will apparently flesh this out more than any experiment done to date.

[jmc]

10 million or so is a nice number. Its enough to get sufficient data to show whether there's anything in the approach or not, but not too much that money gets needlessly thrown down the drain for a high risk project.

[Aero]

It seems we are hanging our hat on projected benefits of scaling up in size. A question.
"How big is a reactor still small enough to benefit the Navy?"
I guess I'm asking, "How far up can we scale the polywell radius and still hope for support from the current sponsor, the Navy?" I'd guess no more than five meters radius but that is a guess.

[jmc]

That's no big deal. If the reactor produces neutrons but is too big to fit on a ship then the navy will release the data and someone else (DOE?) will pick up the torch.

IMHO the size of the Polywell isn't the main concern, the main concern is the giagantic quantity of heating power ( 100s of MW) it will take to heat a high density beta=1 lossy line cusp machine, even one of modest size.

[D Tibbets]

Alternately, if they have a computer guru on board, and wanted to setup a internet based massively parallel network like SETI and others have done, I'd sign up. For that matter, how do home made super computers ( like parallel Playstation 3's) compare to the best current supercomputers, or the supercomputers of the early 90's ?

Wish that it wasn't so, but a SETI-like network isn't suitable since the problem can't be divided up into independent parts.

The Playstation 3 is around 200gigaFLOPS (1E9).
Hereis a comparison of PC systems.

The HPC petaFLOPS (1E15) mark was passed in 2008.
Hereis a comparison of HPC systems (page 3).

This other discussion pegged useful simulation requirement at 1E34.

Thanks.
Well, if 10^34 flops are needed to do a full up simulation in ,say 1 minute, then with current systems it would only take ~ 10^19 minutes, or ~ 5 * 10^14 years, or ~ 500 trillion years. Even ITER may be finished by then !


Dan Tibbets

[alexjrgreen]

IMHO the size of the Polywell isn't the main concern, the main concern is the giagantic quantity of heating power ( 100s of MW) it will take to heat a high density beta=1 lossy line cusp machine, even one of modest size.

You don't need to heat the plasma....

[Maui]

What I do know is that EMC2 got $8 mil in TARP funds and $4 mil more promised.

I thought Obama promised no more bank bailouts? Personally, I think Dr. Neber should focus on fusion instead of derivative markets, but who am I to say?