Steffen Christgau on #HPC at #deRSE23:
#CUDA (#nVidia) and #HIP (#AMD) are essentially the same, you just have to use a different prefix. And they are both #proprietary.
So, best use an #openStandard, #crossPlatform, vendor-agnostic framework for your #GPU code. To make sure it can run on any HPC system.
For #AI applications, you don’t need nVidia GPUs as well. Platform #portability is possible nowadays!
#hpc #derse23 #cuda #nvidia #hip #amd #proprietary #openstandard #CrossPlatform #gpu #ai #portability
Steffen Christgau on #HPC at #deRSE23:
#CUDA (#nVidia) and #HIP (#AMD) are essentially the same, you just have to use a different prefix. And they are both #proprietary.
So, best use an #openStandard, #crossPlatform, vendor-agnostic framework for your #GPU code. To make sure it can run on any HPC system.
For #AI applications, you don’t need nVidia GPUs as well. Platform #portability is possible nowadays!
#hpc #derse23 #cuda #nvidia #hip #amd #proprietary #openstandard #CrossPlatform #gpu #ai #portability
Any #AndroidDev s or #MobileDev s out there using #Flutter? Last time I worked #CrossPlatform was #Ionic. Or anyone prefer #ReactNative over #Flutter?
#androiddev #MobileDev #Flutter #CrossPlatform #Ionic #ReactNative