Instances›AWS EC2›g5g›g5g.16xlarge
Compute
vCPUs64
Memory128 GiB
Mem / vCPU2.00 GiB
ProcessorGraviton2 Processor
Clock speed2.5 GHz
Architecturearm64
Storage
TypeEBS only
Bandwidth19,000 Mbps
Throughput2375 MB/s
IOPS80,000
Networking
Performance25 Gigabit
Max ENIs15
IPv6Yes
GPUNVIDIA T4G Tensor Core
Family sizes
g5g.xlarge4 vCPU · 8 GiB
g5g.2xlarge8 vCPU · 16 GiB
g5g.4xlarge16 vCPU · 32 GiB
g5g.8xlarge32 vCPU · 64 GiB
g5g.16xlarge64 vCPU · 128 GiB
g5g.metal64 vCPU · 128 GiB
g5g.16xlarge

The g5g.16xlarge pairs 64 vCPU with 128 GiB of RAM, the 2.00 GiB-per-vCPU ratio that marks a GPU instance. In the name, "g" is graphics, "5" is the generation, "g" marks AWS Graviton (ARM) processors and "16xlarge" is the size within the g5g family. It runs on AWS Graviton2 Processor at 2.5 GHz, and AWS introduced it in November 2021. AWS has since shipped g6 and g7; g6.16xlarge is the direct replacement at this size, scoring 1,380,873 CoreMark/s against this instance's 1,283,422 — roughly 8% more throughput for the same vCPU count. Rackspace Spot runs one GPU class, gpu.vs2.xlargeplusplus — a single NVIDIA A30 with 24 vCPU and 64 GiB, in San Jose only. Against this instance's 2 × NVIDIA T4G Tensor Core that is a downgrade per card, honest for inference and not for the training this instance was bought for.

Match qualityWeaker GPU

g5g.16xlarge puts 2 Turing, Turing GPUs (32 GiB VRAM) in one machine. Spot has a single GPU class and no multi-GPU node, so there is no shape here that places 2 cards behind one host. Per card the NVIDIA A30 (Ampere, 24 GB VRAM) is a step down from what g5g.16xlarge carries, and the aggregate is not reachable at all.

That makes this honest for data-parallel inference and embarrassingly parallel batch work, where each card is independent, and dishonest for anything that shards a model across cards or synchronises every step. The catch is location rather than hardware: US West, San Jose is the only Spot region that offers it, so latency to your users and egress to your data are the things to check.

x86_64 onlyg5g.16xlarge is Graviton (arm64) and Rackspace Spot has no arm64 nodes, so you'd be moving to x86. You'll need x86 builds of your images, and per-core speed differs between the chips — worth testing before you decide how many nodes to run.
Pricing by region
1 hr
RegionAWS on-demandAWS Spot1-yr Reserved3-yr ReservedRackspace SpotYou save
US West, San Jose$0.191
When to use it
Good for
  • ●Machine learning experimentation and fine-tuning
  • ●Remote graphics workstations and rendering
  • ●CUDA-accelerated batch jobs
  • ●Workloads that fit inside 128 GiB of RAM
  • ●Parallel work that can occupy 64 vCPU
  • ●Network-heavy services — AWS rates this size at 25 Gigabit
Avoid if
  • ●Starting a new deployment. g6.16xlarge is the newer generation at this size.
  • ●The working set needs more than 128 GiB of RAM
  • ●You cannot rebuild x86-only binaries or vendor images for aarch64
  • ●The work is single-threaded. Most of these 64 vCPU would sit idle.
  • ●Local NVMe or instance-store disk is a hard requirement
  • ●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.
  • 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