p5.48xlarge
Amazon EC2 · P5 family
The p5.48xlarge is a gpu & accelerated Amazon EC2 instance type with 192 vCPUs and 2 TiB of memory, 8 × NVIDIA H100, running on AMD EPYC 7R13 Processor with 3200 Gigabit networking, starting at $55.04 per hour on demand.
- vCPUs
- 192
- memory
- 2 TiB
- gpu
- 8× H100
- on-demand
- $55.04
- spot
- $57.76
- regions
- 15
dedicated
10.67 GiB per vCPU
640 GiB total
$40,179/mo · us-east-1
−-5% vs on-demand
across 5 areas
p5.48xlarge price by region
15 regions · hourly list prices in USD · On-demand · Spot · 1-yr Reserved · 3-yr Reserved
| US East (N. Virginia)us-east-1cheapestreferenceNorth America | $55.04 | $57.76 | $23.78 |
| US East (Ohio)us-east-2cheapestNorth America | $55.04 | $35.72 | $23.78 |
| US West (Oregon)us-west-2cheapestNorth America | $55.04 | $31.30 | $23.78 |
| EU (Stockholm)eu-north-1Europe | $58.89 | – | $25.44 |
| Canada (Central)ca-central-1North America | $63.30 | – | $27.34 |
| Asia Pacific (Mumbai)ap-south-1Asia | $66.05 | – | $28.53 |
| AWS GovCloud (US)us-gov-west-1North America | $66.05 | – | $28.53 |
| Asia Pacific (Tokyo)ap-northeast-1Asia | $68.80 | – | $29.72 |
| EU (Frankfurt)eu-central-1Europe | $68.80 | – | – |
| US West (N. California)us-west-1North America | $68.80 | – | $29.72 |
| Asia Pacific (Sydney)ap-southeast-2Oceania | $71.55 | – | $30.91 |
| EU (London)eu-west-2Europe | $71.55 | – | $30.91 |
| Asia Pacific (Jakarta)ap-southeast-3Asia | $75.96 | – | $32.81 |
| Asia Pacific (Seoul)ap-northeast-2Asia | $75.96 | – | – |
| South America (Sao Paulo)sa-east-1South America | $92.47 | – | $39.95 |
Specifications
compute
- vCPUs
- 192
- Memory
- 2 TiB
- Memory per vCPU
- 10.67 GiB
- Processor
- AMD EPYC 7R13 Processor
- CPU vendor
- AMD
- Architecture
- x86-64
- Clock speed
- 2.95 GHz
- Burstable CPU
- No
storage & network
- Local storage
- 30.7 TB SSD (8 disks)
- Storage detail
- 8 x 3840 GB SSD
- Network
- 3200 Gigabit
- Peak bandwidth
- 3,200 Gbps
accelerators
- Model
- NVIDIA H100
- Count
- 8
- Memory (total)
- 640 GiB
- Architecture
- Hopper
- Vendor
- NVIDIA
aws details
- API name
p5.48xlarge- Family
- P5
- Category
- GPU & accelerated
- Generation
- Current
- Regions
- 15 (Asia, Europe, North America, Oceania, South America)
- Cheapest region
- US East (N. Virginia) · $55.04/hr
- Dedicated EBS throughput
- 80 Gbps
- Enhanced networking
- Yes
- Normalization size factor
- 384
- Processor features
- AMD Turbo; AVX; AVX2
- Spot interruption rate (us-east-1)
- <5%
- Spot savings vs on-demand (us-east-1)
- 57%
p5.48xlarge equivalents on other clouds
closest match by vCPU, memory, CPU architecture, accelerator and class · Linux on-demand in each provider's reference region
| Provider | Instance | vCPU | Memory | Processor | On-demand | vs p5.48xlarge |
|---|---|---|---|---|---|---|
| AWS | p5.48xlarge | 192 | 2 TiB | 8 × NVIDIA H100 | $55.04 | – |
| Azure | ND96isr H100 v596 vCPU (2× less) · 1900 GiB (1.1× less) | 96 | 1,900 GiB | 8 × NVIDIA H100 | $98.32 | +79% |
| Google Cloud | a3-highgpu-8g208 vCPU (1.08×) · 1872 GiB (1.1× less) | 208 | 1,872 GiB | 8 × NVIDIA H100 | $88.49 | +61% |
| a3-megagpu-8g208 vCPU (1.08×) · 1872 GiB (1.1× less) | 208 | 1,872 GiB | 8 × NVIDIA H100 | $92.94 | +69% |
P5 family sizes
| Size | vCPU | Memory | Accelerator | Local storage | On-demand | Monthly | Spot |
|---|---|---|---|---|---|---|---|
| p5.4xlarge | 16 | 256 GiB | 1 × NVIDIA H100 | 3.8 TB SSD | $6.88 | $5,022 | – |
| p5.48xlarge | 192 | 2 TiB | 8 × NVIDIA H100 | 30.7 TB SSD (8 disks) | $55.04 | $40,179 | $57.76 |
p5.48xlarge FAQ
How much does p5.48xlarge cost?
p5.48xlarge costs $55.04 per hour on demand with Linux in US East (N. Virginia) (us-east-1), about $40,179 per month. Across 15 regions the Linux on-demand price ranges from $55.04 in US East (N. Virginia) to $92.47 in South America (Sao Paulo). Spot capacity costs $57.76 per hour (-5% less), but can be reclaimed.
How many vCPUs and how much memory does p5.48xlarge have?
p5.48xlarge has 192 vCPUs and 2 TiB of memory (10.67 GiB per vCPU), running on AMD EPYC 7R13 Processor. It is a gpu & accelerated instance type in the P5 family.
What GPU does p5.48xlarge have?
p5.48xlarge has 8 × NVIDIA H100 with 640 GiB of accelerator memory in total.
What are the equivalents of p5.48xlarge on other clouds?
The closest matches by vCPU, memory, CPU architecture, accelerator and instance class are on Azure, ND96isr H100 v5 (96 vCPU, 1,900 GiB, $98.32/hour in East US); on Google Cloud, a3-highgpu-8g (208 vCPU, 1,872 GiB, $88.49/hour in Iowa).