How to earn from excess electricity or computing power

Because I live in a truck, in summer I have the interesting problem of excess (practically free) photovoltaics electricity. The same can happen in an off-grid home or a grid-connected home in a location with zero-export regulations.

This is a small overview of the current options to earn from utilizing your unused or underused computing resources and / or electricity.

By “recommendability”, judged purely subjectively by myself:

• SONM. Blockchain project where you rent out systems by time, like VPS hosts on a cloud platform. The system just went live (yesterday). Looks like well-done tech, worth a try. Of course, nobody knows what you can earn with this (yet), but it should not get lower than the price of electricity. So if you have excess (“free”) electricity available, it’s always a benefit for you.
• Golem. Blockchain project for various special-purpose computation tasks (Blender rendering, later machine learning etc.). Already live since a few months, see here for reported earnings.
• iExec. Blockchain project where you rent out your CPU resources and earn tokens. For a comparison to SONM, see here.
• Primecoin. The first “meaningful mining” coin ever created. Coins are mined by securing transactions with a prime number chain called “Cunningham chain“. This is for the most part basic research, but has some uses: “Cunningham chains are now considered useful in cryptographic systems since “they provide two concurrent suitable settings for the ElGamal cryptosystem … [which] can be implemented in any field where the discrete logarithm problem is difficult.” (source). For results of the prime number chains it found, see the records and  the details. The coin is “naturally scarce” due to the scarcity of prime numbers, just that the upper limit of coins that will exist is not known beforehand (nice feature :D).
• Gridcoin. One of the first “useful mining” coins, started in October 2013. Uses an interesting concept called “proof of research” that combines proof of stake and proof of BOINC (contributions to the BOINC platform for distributed scientific computing). You are not paid by BOINC projects but donate your CPU resources to them; instead you are paid in newly minted Gridcoins. Since this (together with the 1.5% inflation from teh proof-of-stake) sets Gridcoins on a path of continuous inflation and there is no immediate use value for Gridcoin (except speculation), this is a rather poor design for a currency. I once tested this about 1-2 years ago and calculated what I could make when running my i7 notebook on excess solar power (4-6 hours a day), and it was only 1-2 USD a year.
• EFF prizes for large primes. You can participate in GIMPS (a collaborative effort hunting these primes) but this is more for sportsmanship and not for the money, as it seems there are no regular “mining pool style” payouts or shares of a future payout in case of an eventual, collaborative success. GIMPS will distribute a small fraction to the person actually finding it on their computer (3000 USD of 150k USD? compare here and here). You could instead hunt these primes solo, but the chances of success are of course slim. Good for those who like playing lottery and have free electricity around, so it does not cost them anything …
• Proof-of-work mining. There are lots of cryptocurrencies you can mine with proof-of-work, including Bitcoin of course (but that’s only meaningful with GPUs and ASIC miners these days) and others that are designed to be economically CPU mineable. However, I don’t recommend this, as all these calculations are used for nothing beyond securing transactions – which can also be done with proof-of-stake instead of burning all that electricity. All mineable coins where mining serves a meaningful purpose beyond this have been included in the list above.

And some not yet or no longer functional projects:

• DCP. Very similar to Gridcoin, as rewards are again earned from BOINC calculations. But seems to provide a more modern tech stack that could potentially do other tasks in the future. Not released yet.
• Curecoin. Similar to Gridcoin, but limited to only one of the BOINC tasks (protein folding). Also, only half of the energy is used for these computations while the other half goes for proof-of-work. Gridcoin does not have that issue, as proof-of-stake uses only negligible CPU resources. This applies to the previous version. The coin seems to undergo a rewrite / relaunch currently.

There are other (blockchain based) projects that reward people for data storage, data transmission (CDN, video streaming), attention (“voluntary ads viewing”) and sharing personal data. We focused on CPU / GPU intensive tasks here only, as that is the best use in case you have to “burn” free electricity as meaningfully as possible.