Double-double and quad-double arithmetic have even been implemented for the GPU as research prototype libraries (gpuprec [0,1] and campary [2,3] among other).

Double-double and quad-double arithmetic have even been implemented for the GPU as research prototype libraries (gpuprec [0,1] and campary [2,3] among other). [0] https://github.com/lumianph/gpuprec https://github.com/lumianph/gpuprec [1] https://event.cwi.nl/damon2010/gpuprecision.pdf https://event.cwi.nl/damon2010/gpuprecision.pdf [2] https://homepages.laas.fr/mmjoldes/campary/ https://homepages.laas.fr/mmjoldes/campary/ [3] https://hal.archives-ouvertes.fr/hal-01312858/document https://hal.archives-ouvertes.fr/hal-01312858/document