Skip to content

🖥️ Index — Hardware

Hardware dictates what you can load in VRAM, what memory throughput you can reach during decoding, and whether you can run tensor parallelism or offloading.


Path by scenario

BlueprintPriority hardware pages
A — Dev LabAPU & unified memory, Multi-GPU workstations (offloading)
B — SMB ApplianceAPU & unified memory, GPU rack servers (1–2 GPU)
C — Office clusterMulti-GPU workstations, AI networking (Thunderbolt 5 + RDMA)
D — DatacenterGPU rack servers, RoCE / InfiniBand

Chapter pages

  1. 🧠 APU & Unified Memory — Apple M5, AMD Ryzen AI Max PRO 400, NVIDIA DGX Spark 64/128 GB
  2. 🧩 Multi-GPU Workstations — PCIe workstations (no NVLink since Ada), PCIe switches, L40S / A100 servers
  3. 🏭 GPU rack servers — 1U/2U/4U, HGX, RTX vs datacenter choice
  4. 🌐 AI networking — RoCE, InfiniBand, Thunderbolt

Useful hardware lexicon

VRAM · HBM · NVLink · PCIe · Multi-GPU · Unified memory

Give my opinion

Everything is optional: one answer is enough. This site sends nothing: you send it, the way you choose.

Your noteNo note
Your opinion is about
About what? (several choices possible)
What will be sent
Send by e-mail

Private, no account: opens your mail program with the message ready, for ia-on-prem@damien.becherini.fr.

Open a GitHub issue ↗

Public, with a GitHub account. You read the issue again before you publish it.

To paste it wherever you want.

Give at least a note, a subject or a message to be able to send.