AWS EC2 Instances›GPU Instance›p5›p5.48xlarge

Compute

vCPUs192
Memory2048 GiB
Mem / vCPU10.67 GiB
ProcessorEPYC 7R13 Processor
Clock speed2.95 GHz
Architecturex86_64
GPU8
GPU Wattage77 W
GPU ArchitectureHopper, Hopper, Hopper, Hopper, Hopper, Hopper, Hopper, Hopper
Video Memory640 GiB
FPGA0
ffmpeg FPSN/A
CoreMark / sec3,900,325

Storage

Type8 × 3,800 GB NVMe SSD
Bandwidth80,000 Mbps
Throughput10000 MB/s
IOPS260,000
Base Bandwidth80,000 Mbps
Base Throughput10000 MB/s
Base IOPS260,000
Devices8
NVMe DriveYes
SSDYes
Swap PartitionNo
Local Read IOPS4,400,000
Local Write IOPS2,200,000

Networking

Performance3200 Gigabit
Max ENIs64
IPv6Yes
GPUNVIDIA H100
Placement GroupSupported
TrunkingYes
Branch Interface120
IPs / ENI50
Baseline Bandwidth100 Gbps

NUMA Architecture

Uses NUMAYes
Node Count2
Max Distance32
Cores / Node48
Threads / Node96
Memory / Node1,023,979 MB
L3 Cache / Node192 MB
L3 Cache SharedYes

AWS Metadata

Instance typep5.48xlarge
Full nameP5 48xlarge
FamilyGPU instance
Variantp5
GenerationCurrent
Bare MetalNo
EMR supportYes
DedicatedN/A
Introduced2023-07-26
Family sizes
p5.4xlarge16 vCPU · 256 GiB
p5.48xlarge192 vCPU · 2048 GiB

p5.48xlarge

The p5.48xlarge is a GPU AWS EC2 instance with 192 vCPU and 2048 GiB of memory. Rackspace Spot has no equivalent to offer: 8 × NVIDIA H100 wired together behind a high-speed interconnect is one tightly coupled machine, not 8 independent nodes, and that interconnect is usually what the instance was bought for.

Rackspace Spot's largest VM tops out at 16 vCPU and 128 GiB, and even that is not the missing piece here. The table below prices what AWS charges; the honest answer on the Rackspace Spot side is that this workload stays where it is.

Server comparison caveats
No equivalent

The AWS p5.48xlarge is 8 Hopper, Hopper, Hopper, Hopper, Hopper, Hopper, Hopper, Hopper GPUs (640 GiB VRAM) wired together inside one chassis. The interconnect is what you bought (that is what makes it one accelerator with pooled memory rather than several machines), and Rackspace Spot has nothing that reproduces it. The Rackspace Spot GPU class is built on the NVIDIA A30 (Ampere, 24 GB VRAM), and separate nodes of it on ordinary datacentre networking are not a substitute for a single interconnect-joined node.

Split the job across independent single-GPU nodes only if it is data-parallel with no per-step synchronisation and the model fits one card. Model-parallel work, or anything synchronising gradients every step, does not survive the move. Rackspace Spot sells x86_64 servers up to 16 vCPU, plus one NVIDIA A30 GPU class.

Pricing by region

1 hr

Regions are Rackspace Spot locations.

RegionAWS on-demandAWS Spot1-yr Reserved3-yr ReservedRackspace on-demandRackspace SpotOn-demand price savingsSpot price savings
US East, N. Virginia$55.040$30.690not offered$23.777not offerednot offeredn/an/a
US Central, Chicagono AWS regionn/an/an/anot offerednot offeredn/an/a
US Central, Dallasno AWS regionn/an/an/anot offerednot offeredn/an/a
Hong Kongnot sold heren/an/an/anot offerednot offeredn/an/a
Sydney, Australia$71.552$56.478not offered$30.910not offerednot offeredn/an/a
London, UK$71.552$19.344not offered$30.910not offerednot offeredn/an/a
US West, San Jose$68.800$20.314not offered$29.722not offerednot offeredn/an/a
US Central, Dallas 2no AWS regionn/an/an/anot offerednot offeredn/an/a
US East, N. Virginia 2$55.040$30.690not offered$23.777not offerednot offeredn/an/a

Prices updated 18 Nov 2026.

When to use it

Good for
  • ●CUDA-accelerated batch jobs
  • ●Model inference and embedding generation
  • ●Video transcoding and image processing
  • ●Workloads that fit inside 2048 GiB of RAM
  • ●Jobs that want 8 × 3,800 GB of local NVMe scratch space
  • ●Network-heavy services (AWS rates this size at 3200 Gigabit)
Avoid if
  • ●The working set needs more than 2048 GiB of RAM
  • ●More than 10.67 GiB of RAM per vCPU is needed, since that ratio is fixed for the whole family
  • ●The work is single-threaded. Most of these 192 vCPU would sit idle.
  • ●Anything the CPU can do fast enough. GPU hours are the most expensive on the price list.
Rackspace Spot sells spare capacity at market prices. You pay what the market decides, not a list price.
  • Rackspace Spot sells spare capacity at market prices. It is the cheapest compute on the internet.
  • Prices are set by supply and demand. You pay what the market decides.
  • Compute is delivered as fully managed Kubernetes. All infrastructure is taken care of.
  • Egress costs $0.01/GB after the free allowance. AWS charges $0.09/GB.
Build your Cloudspace

Frequently asked questions

How much does p5.48xlarge cost?

The AWS p5.48xlarge costs $55.04/hr on-demand in US East (N. Virginia), about $40,179/mo running 24/7, or $30.69/hr at the AWS Spot price. Rackspace Spot also sells in Chicago and Dallas, where AWS operates no region at all, so those rows have no AWS price to compare against.

How much is p5.48xlarge per month?

At 730 hours a month, p5.48xlarge in US East (N. Virginia) is $40,179/mo on AWS on-demand, or $22,403/mo at the AWS Spot price.

Does Rackspace Spot have an equivalent to p5.48xlarge?

No, and it is worth saying so plainly: 8 × NVIDIA H100 wired together behind a high-speed interconnect is one tightly coupled machine, not 8 independent nodes, and that interconnect is usually what the instance was bought for. Rackspace Spot sells market-priced CPU and GPU capacity, which is a good fit for the batch, web and inference work around this instance, but not for p5.48xlarge itself.

What is a cheaper alternative to p5.48xlarge?

Not on Rackspace Spot, for the reason above. Within AWS the usual levers are a smaller size in the same family, a reserved or savings-plan commitment, or AWS Spot capacity if the work can tolerate interruption.