g4ad.8xlarge as an EKS node
32 vCPU128 GiBx86_64AMD EPYC 7R322 GPUAWS us-east-1G4AD Eight Extra Large
Where the g4ad.8xlarge goes
Default EKS setup: EKS, VPC CNI default (ENI limit).
CPU31.65 of 32 vCPU left for your pods
Memory126.74 GiB of 128 GiB left for your pods
your podsDaemonSets (200m, 300 MiB assumed)kubelet reservedeviction
58max pods per node
$1,265.82on demand / month · $1.73/h
$541.22spot / month (varies)
g4ad.8xlarge EKS max pods and allocatable memory — quick answers
- How many pods can a g4ad.8xlarge run as a node with EKS, VPC CNI default (ENI limit)? Up to 58; kubelet keeps 993 MiB of memory and 150m CPU, leaving 127.03 GiB and 31.85 vCPU for pods.
- How many pods can a g4ad.8xlarge run as a node with EKS, prefix delegation? Up to 250; kubelet keeps 3105 MiB of memory and 150m CPU, leaving 124.97 GiB and 31.85 vCPU for pods.
- How many 500m / 1 GiB pods fit? 55 per node after DaemonSets; 1 nodes for 10 replicas, about $1,265.82/month on demand.
What kubelet reserves on g4ad.8xlarge
| Setup | Max pods | kube-reserved CPU | Memory reserved + eviction | Allocatable CPU | Allocatable memory | Memory usable |
|---|---|---|---|---|---|---|
| EKS, VPC CNI default (ENI limit) secondary IPs per ENI | 58 | 150m | 893 Mi + 100 Mi | 31.85 | 127.03 GiB | 99.2% |
| EKS, prefix delegation ENABLE_PREFIX_DELEGATION=true | 250 | 150m | 3005 Mi + 100 Mi | 31.85 | 124.97 GiB | 97.6% |
How these numbers are computed: reservation formulas. How many subnet IPs a g4ad.8xlarge takes with the VPC CNI: EKS max pods and IP calculator. Check a live node with kubectl describe node → Allocatable.
How many pods fit on one g4ad.8xlarge
Assumes DaemonSets (CNI, kube-proxy, log and metrics agents) request 200m CPU, 300 Mi and 3 pods per node. Try your own numbers in the node calculator.
| Pod requests (CPU / memory) | EKS, VPC CNI default (ENI limit) | EKS, prefix delegation | $ per pod / month |
|---|---|---|---|
| 100m / 128 Mi | 55 | 247 | $23.01 |
| 250m / 256 Mi | 55 | 126 | $23.01 |
| 250m / 512 Mi | 55 | 126 | $23.01 |
| 500m / 512 Mi | 55 | 63 | $23.01 |
| 500m / 1024 Mi | 55 | 63 | $23.01 |
| 1000m / 1024 Mi | 31 | 31 | $40.83 |
| 1000m / 2048 Mi | 31 | 31 | $40.83 |
| 2000m / 4096 Mi | 15 | 15 | $84.39 |
| 4000m / 8192 Mi | 7 | 7 | $180.83 |
Similar EKS node types
| Type | vCPU | GiB | Alloc CPU | Alloc memory | Max pods | $/month | $ per alloc GiB |
|---|---|---|---|---|---|---|---|
| g6.8xlarge | 32 | 128 | 31.85 | 125.14 GiB | 234 | $1,470.51 | $11.75 |
| g4dn.8xlarge | 32 | 128 | 31.85 | 127.03 GiB | 58 | $1,588.48 | $12.50 |
| g5.8xlarge | 32 | 128 | 31.85 | 125.14 GiB | 234 | $1,787.04 | $14.28 |
| g7.8xlarge | 32 | 128 | 31.85 | 120.86 GiB | 632 | $2,982.96 | $24.68 |
| g3.8xlarge | 32 | 244 | 31.85 | 240.97 GiB | 250 | $1,664.40 | $6.91 |
| p3.8xlarge | 32 | 244 | 31.85 | 241.14 GiB | 234 | $8,935.20 | $37.05 |
| gr6f.4xlarge | 16 | 128 | 15.89 | 125.14 GiB | 234 | $778.18 | $6.22 |
| g5g.8xlarge | 32 | 64 | 31.85 | 61.14 GiB | 234 | $1,001.56 | $16.38 |
Closest match on other clouds
| Type | vCPU | GiB | Alloc CPU | Alloc memory | Max pods | $/month | $ per alloc GiB |
|---|---|---|---|---|---|---|---|
| g2-standard-32 | 32 | 128 | 31.85 | 118.58 GiB | 110 | $857.27 | $7.23 |
| a2-highgpu-2g | 24 | 170 | 23.87 | 159.74 GiB | 110 | $1,079.64 | $6.76 |
| g2-standard-24 | 24 | 96 | 23.87 | 88.50 GiB | 110 | $642.95 | $7.26 |
| a3-highgpu-1g | 26 | 234 | 25.86 | 222.46 GiB | 110 | $923.23 | $4.15 |
| g2-standard-48 | 48 | 192 | 47.81 | 181.30 GiB | 110 | $1,285.90 | $7.09 |
| a2-ultragpu-1g | 12 | 170 | 11.90 | 159.74 GiB | 110 | $832.72 | $5.21 |