The p5.48xlarge instance is part of the p5 series, featuring 192 vCPUs and 3200 Gigabit of RAM, with Gpu Instances. It is available at a rate of $98.3200/hour.
Amazon EC2 P5 instances represent the newest GPU-based generation, offering top-tier performance for deep learning and high-performance computing (HPC) tasks on Amazon EC2.
Equipped with 3rd Gen AMD EPYC processors (AMD EPYC 7R13)
Features up to 8 NVIDIA H100 Tensor Core GPUs
Offers up to 3,200 Gbps of network bandwidth, supporting Elastic Fabric Adapter (EFA) and NVIDIA GPUDirect RDMA (remote direct memory access)
Delivers 900 GB/s peer-to-peer GPU communication via NVIDIA NVSwitch
Utilized in Amazon EC2 UltraClusters that include up to 20,000 NVIDIA H100 Tensor Core GPUs, petabit-scale networking, and expandable low-latency storage using Amazon FSx for Lustre
Credits: AWS Resources, Updated At: 2024-08-23
Applications in generative AI such as answering questions, generating code, creating videos and images, and recognizing speech
Instances | vCPUs | Memory (GiB) |
---|---|---|
p5.48xlarge | 192 | 2048 GiB |
General | |
---|---|
Type | p5.48xlarge |
Region | N. Virginia |
Family Group | P5 |
Family Category | Gpu Instances |
Instance Generation | Current |
Term Type | On-demand |
Pricing (USD/hr) | $ 98.3200 |
Compute | |
---|---|
vCPU | 192 |
Memory | 2048 GiB |
Storage | 8 x 3840 GB SSD |
Processor Architecture | 64-bit |
Operating System | Linux |
EBS Optimized | NA |
Tenancy | Shared |
Clock Speed | 2.95 GHz |
GPU | 8.000000 |
ECU | 0.0000 |
Virtualization | HVM |
Networking | |
---|---|
Network Performance | 3200 Gigabit |
ENA Support | True |
Region | Price / Unit | Monthly Pricing |
---|---|---|
Ohio | $98.3200 | $71773.60 |
Oregon | $98.3200 | $71773.60 |
Stockholm | $105.2024 | $76797.75 |
Central | $113.0680 | $82539.64 |
Mumbai | $117.9840 | $86128.32 |
Tokyo | $123.1950 | $89932.32 |
Sydney | $127.8160 | $93305.68 |
London | $127.8160 | $93305.68 |
Sao Paulo | $165.1776 | $120579.65 |