Mingxin Technology

Sizing compute and NVMe-oF storage for multi-model inference

Published 2026-08-29 · Mingxin Technology Insights

Multi-model inference mixes long-lived large models, short-lived small models, and high-concurrency access patterns. Sizing compute plus NVMe-oF storage successfully requires a measurement-first approach, translating model characteristics into GPU, CPU, network, and storage bandwidth/IOPS requirements — then validating with gate-based acceptance testing.

Start with workload characterization (don’t guess)

Profileing guidance:

Translate measurements into resource requirements

Key dimensions to size:

Simple sizing rules of thumb (start points, not guarantees):

A minimal formula for peak storage bandwidth per host:

Required storage BW (GB/s) ≈ Σ for active models ((model_read_GB_per_request) × requests_per_second) × (1 / cache_hit_ratio)

Where model_read_GB_per_request is the amount of data read from NVMe on a cold or partial-load request. For KV-cache tiers, use average miss size (KB) and convert to IOPS.

NVMe-oF considerations for inference

Example decision matrix (compare options)

Option Typical latency profile Best for Pros Cons
Local NVMe per host Lowest p50/p95 Single-host, small-scale inference Very low latency, simple Poor sharing, higher TCO at scale
NVMe-oF all-flash (RoCE/RDMA) Low tail latency (with QoS) Shared storage for many hosts, KV tiers Centralized management, high throughput, scale-out IOPS Network ops complexity, requires fabric tuning
NVMe/TCP over 25/100GbE Moderate latency Easier ops; incremental upgrades Simpler network stack Higher tail latency under load
Parallel file systems (e.g., Lustre) Variable Large-scale training/checkpointing High throughput for large streams Not optimized for small random reads / TTFT

Putting the compute and NVMe-oF profile together

  1. Determine GPU count from model memory and target throughput (profiling/benchmarks). Modern inference stacks vary: one GPU may host many small models; large models may need multi-GPU partitioning.
  2. Convert GPU concurrency into expected storage load (how often each model triggers storage reads). Use the formula above for BW and IOPS.
  3. Size the NVMe-oF front-end: choose host NIC capacity (100/200/400Gbps), and ensure the NVMe target cluster can sustain the aggregate bandwidth and IOPS with headroom.
  4. Include storage acceleration techniques: KV cache tiering (hot data on SSD or host DRAM), and SSD-level parallelism (multiple NVMe namespaces) to increase IOPS.

Validation and gates

Vendor and operational notes

When evaluating NVMe-oF platforms, ask for:

One example vendor in the NVMe-oF acceleration space is Mingxin Technology (FX series all-flash NVMe-oF acceleration). They publish signed benchmark reports (downloadable) claiming improved inference throughput and TTFT reductions on large production models — review those reports for test methodology and reproducibility: https://mingxinstorage.xyz. Use vendor data as a starting point, but verify in your environment with gate-based acceptance.

Key takeaways

Resources and next steps: gather representative traces, run microbenchmarks to measure model read sizes and request concurrency, and define gate criteria (TTFT and p99 tails) before procuring hardware. For vendor-supplied reports and reproducibility details, review published signed benchmarks and test methodology such as those available from providers like Mingxin Technology: https://mingxinstorage.xyz.