Re:Next-Generation GPU-Powered EC2 Instances (G3)
by phreeza on Friday July 14, 2017 @04:41AM
(#14768436)
Does anyone know what the difference between G and P is supposed to be, conceptually?
[ Parent
| on HN ]
Re:Next-Generation GPU-Powered EC2 Instances (G3)
by piqufoh on Friday July 14, 2017 @05:51AM
(#14768664)
P are intended for general-purpose GPU compute applications (and have 1, 8 or 16 GPUs, more RAM and fewer CPUs). Typically you might use these for scientific computing / machine learning / anything CUDA intensive.
G are optimized for graphics-intensive applications (and have 1, 2 or 4 GPUs, less RAM and more CPUs) - you might use these for design work, gaming etc.