some of us are going to rapidly run out of limbs!

I've already run out of saleable kidneys...

Rob, I’m not "Anti-BOINC," I’m "Pro-Science." The founding principle of BOINC was to use spare cycles effectively, but in 2026, the definition of "effective" has fundamentally changed. If a project accepts a month of crunching from an old heater when a modern $600 Mac Mini or a Snapdragon laptop can do that same work in an hour, we aren't just donating spare time—we’re donating a massive amount of avoidable carbon emissions for a negligible scientific return.

As for the "cost of limbs" argument from Dave: yes, RAM is expensive right now. But that is exactly my point. Integrated SoCs like the M4 or the Snapdragon X2 are actually the budget-friendly solution here. You don't need a $5,000 workstation anymore. These small, 15W-30W chips are outperforming legacy rigs that cost four times as much to run. It's about total cost of ownership, not just the sticker price.

We have to ask ourselves: Why are we so protective of the right to be inefficient? If we want research organizations to keep using BOINC, we have to show them that we aren't just a "hardware museum" but a high-throughput network that values results over nostalgia.

I’m not saying you have to throw your PC in the trash every year. I’m saying we should stop pretending that every "spare cycle" is equal. When the efficiency gap is 4000%, it's time to admit that some hardware is simply no longer suited for the job.

In reply to kasdashdfjsah's message of 5 May 2026:

These small, 15W-30W chips are outperforming legacy rigs that cost four times as much to run. It's about total cost of ownership, not just the sticker price.

prove it and actually use your system for more than 1-10 tasks on a project. so far, using apps built by other people for you, you have successfully completed 0 tasks on Einstein, 1 task on Primegrid, and 8 tasks on Asteroids. you do not have ANY credible data to make such claims. run 100+ tasks on each, then compare to other hardware running the same subprojects. I have a feeling you wont because you dont want to see that it's not as impressive as you're projecting.

Why are we so protective of the right to be inefficient? If we want research organizations to keep using BOINC, we have to show them that we aren't just a "hardware museum" but a high-throughput network that values results over nostalgia.

take a real look at the hardware most people are using. very very few people are using such old systems. some yes, but most recognize their own budget constraints vs output. most folks use hardware that's within the last 5 years or so.

When the efficiency gap is 4000%

prove it or shut up. this number is based on nothing.

---

And while we wait for kasdashdfjsah to prove his concept is the only true way forward to using BOINC for everyone, I will close this thread. He can always start a new thread later, to post his concept's proof or disproof.