Blog

Einblicke, Tutorials und Neuigkeiten von unserem Team

CETC Cloud CNCF Case Study

Fallstudie

CETC Cloud Builds a Domestic GPU Sharing Foundation for a Portable Knowledge Base with HAMi

CETC Cloud needed to bring its intelligent knowledge base to project sites, running text generation, embedding, rerank, knowledge processing, and R&D debugging on a single set of portable devices with domestic GPUs. With Kubernetes and HAMi, per-device dev environment capacity grew from 2 to 30 concurrent Pods, and a single card freed 56 GB of memory and 80% of compute.

HAMi Composable Scheduling Policies

Technischer Deep Dive

HAMi 2.10 Deep Dive (Part 3): Exclusive, Binpacked, and NUMA-Aligned — Scheduling Policies Are Now Composable

Inference replicas want tight binpacking, multi-GPU Pods want the same NUMA node for bandwidth, and latency-sensitive workloads want whole GPUs to themselves — three policies in one sentence, and you used to have to pick just one. In the third article of our HAMi 2.10 deep-dive series, we explain the filter-then-sort evaluation model and verify each step with a real selection path across four T4s.

Volcano + HAMi vNPU Soft Slicing Tested

Technischer Deep Dive

HAMi 2.10 Deep Dive (Part 2): One Ascend Card, Two Pods — Testing Volcano + HAMi-core vNPU Soft Slicing

GPUs have HAMi for hard memory isolation — but what about Ascend NPUs? In the second article of our HAMi 2.10 deep-dive series, we validate Volcano scheduling + HAMi-core soft slicing on a real Ascend 310P3: the container sees only its 8192 MiB slice, two Pods binpack onto the same physical card, and monitoring metrics report faithfully.

KAI + HAMi Memory Isolation Tested

Technischer Deep Dive

HAMi 2.10 Deep Dive (Part 1): Tested — KAI Scheduler + HAMi GPU Memory Hard Isolation, Not a Single MiB Extra

The biggest anxiety around GPU sharing is "we agreed to split it fifty-fifty — so how can you oversubscribe?" This is the first article in our HAMi 2.10 deep-dive series: with a reproducible test on GKE, we answer whether HAMi-core can truly pin each Pod's GPU memory to its quota after KAI Scheduler co-schedules two Pods onto the same NVIDIA T4. Requesting 3 GiB succeeds; accumulating 5 GiB fails with an immediate out-of-memory.

HAMi v2.10.0 Release Deep Dive

Produktversion

HAMi v2.10.0 Deep Dive: Flexible MIG, AMD vGPU, and a Full Scheduling Ecosystem Upgrade

HAMi v2.10.0 is officially released! Flexible MIG creates and reclaims MIG instances on demand via NVML, AMD Instinct MI300X gains software vGPU support, the mutex policy and composable scheduling policy chains cover real-cluster needs, and the KAI Resource Isolator plus Volcano vNPU integration carry HAMi-core's isolation into a broader scheduling ecosystem.

Subscribe to our blog

Get the latest technical articles, tutorials, and HAMi community updates.

Blog - Dynamia AI | GPU-Virtualisierung & HAMi