The present 32-bit ARMv7 application needs a lot of memory, so much that a Nvidia Jetson Nano with 2GB of RAM can't complete the tasks when doing four at a time -a 4GB model can barely finish them. A 4GB Odroid-N2+ that has six cores can't do six at a time, but has no problem doing two. As a 64-bit ARMv8 application could address much more memory it might speed up the calculations (with WEP-M+2 a 64-bit app halved the calculation time under Linux/ARM, under Android it was even brought back to a quarter).

Actually we could!
I do not know when the app binary gets compiled and how it ends up in the project. But it does compile on aarch64. This will be interesting.

Do you know the boinc platform string for aarch64 linux?

Actually we could!
I do not know when the app binary gets compiled and how it ends up in the project. But it does compile on aarch64. This will be interesting.

Do you know the boinc platform string for aarch64 linux?

I suppose you mean

aarch64-unknown-linux-gnu
for 64-bit ARM Linux

as opposed to
arm-unknown-linux-gnueabihf
for 32-bit ARM Linux

For even more BOINC platform strings: see https://boinc.berkeley.edu/trac/wiki/BoincPlatforms.

I have built the arm 64-bit app. It did not want to load at first, because the platform was missing.

I have built the arm 64-bit app. It did not want to load at first, because the platform was missing.

I got some back thank you very much

30 Jul 2022, 15:57:51 UTC 5 Aug 2022, 16:56:11 UTC En cours --- --- --- Symmetric Prime Tuples v3.00
aarch64-unknown-linux-gnu

30 Jul 2022, 15:57:51 UTC 5 Aug 2022, 16:56:11 UTC En cours --- --- --- Symmetric Prime Tuples v3.00
aarch64-unknown-linux-gnu