That's an awful lot of wasted energy. Just one hash? An RTX 4090 can do smth like 16 million hashes per a single joule.

Since I am, like everyone else around here, an expert in arctic wild life, let's clear things up.

Estimating the Energy Needed for a Penguin’s Single Wing Flap

Introduction:
To roughly estimate the energy required for a penguin to flap one wing once, we can use a simplified physical model. In this model, we treat one wing as a rigid rod rotating about an axis at the shoulder. The energy needed to accelerate this wing is given by the rotational kinetic energy formula:


where:

I is the moment of inertia of the wing,

ω is the final angular velocity.

1. Estimating the Parameters

For a medium-sized penguin, we can assume:

• Mass of the wing: Approximately 0.2 kilograms.
  
• Length of the wing: About 0.3 meters.
  
• Moment of Inertia: Approximating the wing as a rod rotating about one end, the moment of inertia is estimated as:
  
  

• Angular Displacement and Time: Suppose the wing moves through an angle of 90° (approximately 1.57 radians) in about 0.2 seconds. The average angular velocity is then:
  
  
  Code:
  ω ≈ 1.57 0.2 ≈ 7.85 rad/s
  
  2. Calculating the Rotational Energy
  
  Plugging the estimated values into the formula:
  
  
  Code:
  E_rot ≈ (1/2) * 0.006 * (7.85)²
  
  Carrying out the calculation:
  *² ≈ 61.6
  • Thus, E_rot ≈ 0.5 * 0.006 * 61.6 ≈ 0.1848 Joules (approximately 0.2 Joules)
  This value represents the mechanical energy needed to accelerate the wing, without considering energy losses such as friction, air resistance, or the biological inefficiencies in converting metabolic energy to mechanical work.
  
  3. Accounting for Muscle Efficiency
  
  Since muscle efficiency in converting metabolic energy to mechanical work is typically around 20–25%, the actual metabolic energy required could be significantly higher. Assuming an efficiency of about 25%:
  
  
  Code:
  Metabolic Energy ≈ 0.2 J 0.25 ≈ 0.8 J
  
  4. Conclusion
  
  In this simplified model, a medium-sized penguin might require roughly between 0.2 Joules (the calculated mechanical energy) and 1 Joule (when accounting for inefficiencies) for one wing flap. Note that these calculations are based on several approximations. The actual energy could vary with the wing’s mass, flap speed, angular range, and muscle efficiency. Also, while penguins use their wings in a “flying” motion underwater, the dynamics in water differ from those in air.
  
  LE: oh shit, my bad. I confused seagulls with penguins. Guess we need experts from that area to redo all the calculations. Or you might be right. Still, I'd wire up the windmill to an RTX card rather than to a Pi. Free energy + high tech

Kowala
@Kowala24731
** Announcement **
The Kowala team will start working on BTC69 April 10th. Funding round 1 is ongoing, restricted to BTC67 and BTC68 participants. If there is room for extra capital (i.e the cap of 400k USD is not reached by April 10th) we will open funding to external parties.
Expected time to resolution : 2 to 4 months.
We are letting the world know so that people considering taking part in this brute force competition can decide to participate or not knowing all the facts. We estimate competition’s odds of success to be sub 1%.

Not a profitable investment, for a chance of less than 1%

Bad wording. If I understand correctly, it refers to the "others" (the external competitors) odds of success.

No one would be crazy to invest $ 400k with a risk of > 99% to lose the invested funds.

You forget what thread you're in. People twist things around 180, if you put a comma in a wrong place here.

But it might be understood literally as "the competition which was just mentioned odds of success".

I think it's because you called the puzzle a "competition" and then right after referred to the "competition’s odds" and not, like, "our competitors' odds"...


So, yeah, slightly ambiguous use of the same word that could mean either the thing or the people participating in the thing:


p.s. @kTimesG replied while I was writing this, but I already wrote it so I'm posting it anyway

edit: also BTW this is really English's fault, not yours

So many things are off in this message I won’t take the time to respond I’m sorry.

I apologize in advance, I don’t want to offend anyone.
I just want to make one thing clear: it doesn’t matter whether someone has over 2,000 GPUs or just one old GT card — the chances are almost the same for everyone. The only difference is that the one with more computing power pays more — they take on more risk, and yes, they have a slightly higher chance. But for the rest, there’s practically no risk at all.

The whole idea behind this "game" was to show just how "perfectly" the network functions. And if someone finds a vulnerability, the goal is to fix it as fast as possible for the sake of security. So, in a way, we’re bug hunters — and I’m genuinely impressed by RC. He went much further to tackle more complex ranges.

Everyone has their own approach to searching for private keys in this game. I, for example, don’t use a GPU. I based my work on keyhunt with heavy modifications, and now I’ve reached 400 Pkeys/s on an old CPU, even though initially I couldn’t get above 60 Pkeys/s.

P.S. Please don’t ask for the code just yet — I’m still working on improvements, and once it’s ready, I’ll share it with you.
P.S.2. RC, your SOTA method is very impressive and flashy, but not very efficient (sorry). It’s a good method though, and I did start using it myself — until I found a new solution.
P.S.3. Bram, there’s so much buzz around you, but I still have some doubts.