LOL, I wrote it as a joke, I'm pretty confident you only read the first part and went instantly into "superhero out to protect innocent civilian mode".

Of course HYIPs are a hazard to ones coins.

But they are such an obvious hazard that you should also make a large sign to place on pavements, something like:


If someone gets burned on a HYIP then as far as I am concerned it is a cheap lesson to learn and it may, just may, slap a bit of the stupid out of them and slap a little mathematical insight into them.

I tried it...

I was lucky enough to martingale a faucet drop into 10BTC over the last 13 months playing BTC dice.

Anyway... I invested that 10 BTC in the HYIP and 37.332 hours later I received 80 BTC, I'm still shaking as I eat my loaded 12" pizza that I just ordered for a little BTC, what a ride! I have calculated that I can now eat callout pizza every night for the rest of my life, any toppings I crave AND fizzy drink! Prior to this I was only able to callout once a week and I had to choose my toppings carefully and certainly no fizzy drink.

Initially there were some problems with my payment not going through, there was no foul play involved, it was explained to me that the auto transaction HYIP system was a very complex recursive system, and when set running even the programmers did not understand it fully, crazy times we live in.

Anyway... the site admin got on the phone with a "REAL company" manager and everything went pretty quickly after that, thank god the "REAL company" was in the loop or I shudder to think what may have transpired.

Well I'm off to re-invest, I have calculated that by Autumn I will own all the bitcoins ever mined to date, and will then have to wait until more are mined.

And it's all happening because I was fortunate enough to hit a perfect martingale streak and then "Get in on the ground floor" on a HYIP.

Brilliant underclock data and presentation. 

@Marto wrote that the temp sensor hardware is in place but it needs extra firmware code to get the temps out (to cg-miner I presume?)

Out of my 48 coolers, some of the bases look a little suspect and will probably need lapping a little, knowing the die temp would be super handy for thermal debug.

@2GOOD, thanks for posting this great review!

@Marto
I'm about to make up a power loom for HEX4M boards.

Are there 3 * 6-PIN PCIE power connectors next to each other on the bottom of the board? (Like shown on the old 8 chip CAD render)

@loshia, no worries, thanks for your initial info

@scrappy do

No worries, good luck with it, it sounds messy, but I've read a dishwasher works wonders when you pull boards out.

So your plan is: Immerse 5 boards with passive finned blocks on chips in mineral oil and then have 2 pipes with pump leading to the tank. Pipes going through an auto radiator/fan?



About watercooling: if you buy square copper waterblocks, they are universal, can fit on any chip now and in the future. The blocks are the only difficult part. The rest is Vinyl tube -> solvent welded PVC pipe manifold -> rubber auto hose -> radiator/fan.

Is someone planning to mineral oil cool a HEX4M?

That would not be sensible, the only advantage of mineral oil cooling is that you don't need to make cooler blocks, you can stick whole board in because oil remains non conductive. I think most people do it for gimmick aesthetic reasons ( fish tank setup with lights etc )

With constant 400W for each board you would need a large oil volume or you would need to employ a radiator/fan/pump setup and so are back into plumbing, may as well employ blocks then and just use water.

A decent hires top down photo with camera close to 90 degree above board, and a ruler laid over two asics could do it. Camera as high as possible and use zoom to reduce parallax.

All board dimensions could then be derived to ~mm accuracy in a paint program.

P.S. I'm impressed with your ( get it done ) watercool plan. If I didn't have all my fans on hand I would consider doing the same, I would probably buy $10 copper waterblocks on ebay though ($480  ) as I only have a hand drill. And then hit up a junk yard for radiator and hoses.

With thick hoses through a window (or multiple thin through wall holes) you could run the radiator outside in shade in the summer (completely silent rig) and bring it inside in the winter, hmmm very nice!

No worries, It'll all become second nature to you soon, you have no choice if you want BTC

@sorehammer,

I will downconvert from linux_speak for you:


Connect to the TPLINK by connecting via SSH using a terminal client.
( I use puTTY for windows ): http://www.chiark.greenend.org.uk/~sgtatham/putty/

You will always need to connect as a root user, the root username and password will be supplied with the tplink.

The TPLink is running a reduced linux OS, it is a full computer in its own right that can run any programs you place on it. In this case it is loaded with CG Miner.

Once connected you have a command line interface (like DOS) to talk to the TPLINK. From here you can do everything.

Then set the timezone for the linux on the TPLINK:
https://linuxacademy.com/blog/linux/changing-the-time-zone-in-linux-command-line/

---

run the "fix_rc" script.
this script will disable alot of un needed crap.
http://www.cyberciti.biz/faq/howto-run-a-script-in-linux/

---

In linux the "cron system" always ticks when the OS is running every one minute, "cron jobs" are specified in the "cron tab", you can specify to do stuff every minute, every X minutes, every X hours etc.

So you check all cron jobs and remove any that may be daily restsarting cg-miner etc.

http://blog.fayland.org/2011/10/removeadd-job-to-crontab-by-commandline.html

---

Self explanatory.

---

Good info that I did not know before.

You will login via SSH to your router and then issue cg-miner commands to do this.

A public address X will map to a private Key Y

RIPEMD-160(SHA-256(Y))  =  X

Some addresses (X) will have multiple (Y) that will work due to "collisions" in the hash algorithm.

RIPEMD-160(SHA-256(Y  ))  =  X
RIPEMD-160(SHA-256(Y1))  =  X
RIPEMD-160(SHA-256(Y2))  =  X
... maybe more

The chances of finding/generating these collisions is astranomically small, maybe some other member might know the actual chances?

The BEST strategy.


Dell.com now accepts bitcoin, I assume the CEO of Dell, having a personal net worth of ~15 billion, is rather well "connected".

So I am assuming bitcoin IS being allowed to take that next leap.

Good old Copper beats it, but gold is favorable for switches and connectors or extremely thin wires due to it's higher corrosion resistance.

Resistivity (Ω across faces of 1m³ cube)

Carbon (graphene)    1×10−8
Silver    1.59×10−8
Copper    1.68×10−8
Gold    2.44×10−8
Aluminium 2.82×10−8
Calcium 3.36×10−8
Tungsten 5.60×10−8


I mean measurement noise.

A tungsten loaded bar can be thought of as a short stubby wire. A tungsten core surrounded by quite a thick gold sheath.

If you take any high impedance resistance measurement (like a multimeter) then the difference in resistance will be buried deep within the noise, you could amplify the voltage signal as much as you like but no good relative measurements could be made, you would only find noise.

I don't think a resistance test could be accomplished using simple push contact probes.

The cross sectional area of a 400 Oz bar is ~0.0036 M²
The length of a 400 Oz bar is ~0.2 M

So its resistance end to end is ~=  ( 2.44*10-8  * 0.2 ) /  0.0036 ~= 5.6 * 10-7 Ω  that's 560 nΩ

A tungsten loaded bar would be on the order of 700 nΩ

To bring the resistance reading into a range where one could detect a tungsten core you would have to push significant current through that bar, in the order of 10A or more if you had the most sensitive galvanometer on a Wheatstone bridge. Plus the apparatus would need good environment control ( whole apparatus as same temperature ).

You would need a known interface layer at each end. Otherwise you could take a measurement and see a relativley higher resistance but that might just be a relativley poor connection on one end.

Machining flat areas is needed to give a known electrical interface when dealing below µΩ, the flatness reuired would not be super flat optical though as you could apply pressure to the contact blocks and take advantage of golds high malleability.

For a simple resistance check, the amount of current you would have to push through to get out of the noise would require very good large surface area electrical connections at each end of the bar. This would require more machining of the bar than a simple small hole or two that could be refilled with molten gold after heating the bar to accept the melt.

More preferable would be non invasive ultrasonic scanning:

http://www.olympus-ims.com/en/applications/ut-testing-gold-bars/

edit: I think one non invasive way to measure resistance would be a VERY high field induction test over a second or less while analysing the thermal response of the bar using an IR camera. I have not seen reference to this online. It would take more energy than ultrasonic scanning.

It's a no brainer, I would choose BTC.

If I wanted a little gold also from the gift then I could order it online quickly using some fraction of the BTC I received.

It would be more of a hassle if I received gold and then also wanted a little BTC, especially if the gold received was one piece.

Mr. Bitty, got so much style!

377 Mr. Bitty.