it could be fixed on our side by aggregating all p...
# arrow
p
it could be fixed on our side by aggregating all processors in a mega-processor and running them sequentially, although I’m unsure whether that’s allowed