Scenarios
A solution library organized by compute platform × business workload. Every scenario names its measured platform and report IDs; cross-platform numbers are never mixed, and not-yet-measured combinations are honestly marked as scheduled.
Storage Acceleration for AMD MI308X Inference Clusters
An MI308X (192 GB HBM per GPU) inference cluster with external KV tiering: ROCm 7.2 + vLLM + LMCache + FX100 measured throughput +29–40% and TTFT −26–32% on a 480B model (signed reports R1–R4).
Model Loading and Training Storage Acceleration for Ascend 910B Clusters
The common bottleneck in Ascend 910B clusters is NFS storage: model loads take tens of minutes. Mingxin FX measured DeepSeek-32B/70B serving loads 6.2×/9.3× faster, with all three training workloads improved (R9).
HBM Efficiency and KV Tiering Practice on MetaX GPUs
MetaX and other domestic GPUs have limited HBM, making an external KV tier more valuable. R5 completed a seven-way comparison on a single N260, arguing that the external tier can substitute for 128 GB-class HBM residency.
Session-Recovery Acceleration for Agent and Code-Assistant Platforms
Agent and code-assistant workloads combine long contexts with frequent interrupt-resume cycles. An external KV tier turns recovery from full re-compute into read-back: measured on 480B, TTFT p50 fell from 149.5 s to 11.85 s.
GPU Rental Platforms: Raising Per-GPU Output with Storage
The hidden cost of GPU rental is switching idle time: GPUs wait on loading whenever models or tenants change. Measured with FX100, effective compute utilization at the 20-switches/hour level rose from 46.7% to 62.8%.
Checkpoint and Data-Read Acceleration for Training Clusters
Every training checkpoint is a bubble where everyone waits. Measured with FX100, an 8-GPU 32B LoRA full-model snapshot fell from 178 s to 94 s (1.9×), sustained write bandwidth 3.26 → 6.40 GB/s.
An AI Video-Generation Pipeline: ComfyUI + LTX-Video on Domestic Platforms
Video-generation stacks commonly hit dtype/operator/kernel errors on non-NVIDIA platforms. Mingxin completed end-to-end validation of all 7 ComfyUI + LTX-Video 2.3 models on MI308X / ROCm 7.2 with zero operator errors.
Inference Storage Planning for a Thousand-GPU AI Datacenter
The core of thousand-GPU datacenter storage planning is a three-tier architecture plus KV tiering: the measured one-FX100-per-8-nodes ratio, a fully reproducible Python TCO model, and gated joint tests to control build risk.
Inference Acceleration for RAG Knowledge-Base Q&A: Prefix Reuse's Best Scenario
Every RAG query stuffs retrieved document passages into the context, so the same popular documents get prefilled over and over — one of the highest-payoff workloads for prefix reuse. An external KV tier computes a hot document's KV once for the whole cluster.
An Inference Storage Solution for Enterprise Code-Assistant Platforms
Code assistants are the archetypal long-context, high-frequency-recovery workload: codebase contexts run to tens of thousands of tokens and developers close and reopen at will. External KV measured: 32B single-GPU recovery TTFT 37.97 s → 9.30 s; 480B 8.6–20× vs re-compute.
The Storage Substrate of a MaaS Platform: Turning Context Caching into a Billable Capability
A MaaS platform's three storage pains — fast model on/offboarding, tenant session retention, and context-cache pricing — map to measured anchors: loading 6.2–9.3×, switching utilization 46.7%→62.8%, KV read-back 12.6×.
Private Inference Appliances: Integrated Hardware/Software Delivery for Compliance Scenarios
A private inference appliance packages the validated stack (vLLM/LMCache + the FX storage tier) into out-of-the-box whole-machine delivery: data stays in-domain, the stack is auditable, and performance is backed by signed measurements (throughput +29–40%).