The x2idn.32xlarge pairs 128 vCPU with 2048 GiB of RAM, the 16.00 GiB-per-vCPU ratio that marks a memory-optimized instance. In the name, "x" is the extra-memory line, "2" is the generation, "idn" marks Intel processors, local NVMe storage and enhanced networking and "32xlarge" is the size within the x2idn family. It runs on Intel Xeon Scalable (Icelake) at 3.5 GHz, and AWS introduced it in March 2022. AWS has since shipped x8i; x8i.32xlarge is the direct replacement at this size, scoring 3,045,263 CoreMark/s against this instance's 2,132,001 — roughly 43% more throughput for the same vCPU count. Spot's node ladder tops out below this size, so x2idn.32xlarge becomes 15 × mh.vs1.2xlarge (240 vCPU, 1920 GiB combined), $13.34/hr on AWS in US East (N. Virginia), from $0.315/hr on Rackspace Spot. Spot delivers compute as managed Kubernetes, so a node pool of that size is the native shape of the answer rather than a workaround — provided the workload scales horizontally.
Match qualityScale-out match
x2idn.32xlarge gives you 128 vCPU and 2,048 GiB in a single machine. The largest comparable Spot node, mh.vs1.2xlarge, is 16 vCPU and 128 GiB, so the equivalent here is 15 nodes rather than one.
Spot runs as managed Kubernetes, so a 15-node pool is the normal way to run things rather than a workaround. If your workload already sits behind a load balancer, a queue or a scheduler, it spreads across them without changes. If it needs all 128 cores and 2,048 GiB in one process, it does not.
No local diskx2idn.32xlarge comes with 2 × 1,900 GB NVMe SSD attached to the machine. Spot nodes use network storage only, so that disk doesn't come with it. If you were using it as fast scratch space, you'd move to a persistent volume or object storage, which are slower to read.
