Amazon EC2 - AshokBhat/notes GitHub Wiki

Amazon Elastic Compute Cloud (EC2)

Instances

• General Purpose - A, T and M Series
• Compute Optimized - C Series
• Memory Optimized - R and X series
• Accelerated Computing - P, G and F series
• Storage Optimized - I, D and H series

EC2 Accelerated Computing series

Type	CPU	GPU	Interconnect	Use-case
P3	[Intel]] Xeon E5-2686 v4 (Broadwell) ](/AshokBhat/notes/wiki/[[NVIDIA-Tesla) V100	[NVLink]]](/AshokBhat/notes/wiki/[[GPGPU)
P3.24dn	Intel Xeon Platinum 8175M	[NVIDIA Tesla]] V100](/AshokBhat/notes/wiki/[[NVLink)	GPGPU
P2	Intel Xeon E5-2686 v4 (Broadwell)	NVIDIA Tesla K80 GPU	GPGPU compute
Inf1	Intel Xeon (2nd Gen) SP	[AWS Inferentia]]](/AshokBhat/notes/wiki/GigaBit-[[Ethernet)	ML inference
G4	Intel Xeon (Cascade Lake)	[NVIDIA Tesla]] T4 ](/AshokBhat/notes/wiki/GigaBit-[[Ethernet)	ML inference and Graphics
G3	Intel Xeon E5-2686 v4 (Broadwell)	NVIDIA Tesla M60		Graphics

See also

• AWS
• [AWS Graviton]] ](/AshokBhat/notes/wiki/[[AWS-Inferentia)