Dynamo scores GPUs by outstanding read work, not request size
toddhooper · x · 2026-07-27
The first term is outstanding read work
The thread says the first term in Dynamo's score is not the request's own size. It is the amount of reading the target GPU still owes:
- Everything it has already agreed to serve but has not finished.
- Your request added on top.
So the question is not "how big are you?" but "how long is the line you're about to join?" That is what makes the system behave like a load balancer rather than a lookup table.
Related event: Dynamo GPU Scheduler Scores by Outstanding Read Workload(2 posts)→
More from Infra
- Kimi K3 deployment estimate points to 2 B200 nodes or 1 B300 class node — HarveenChadha · 2026-07-28
- MLA-based KV cache costs 12 GB per million tokens, with KDA state at 230 MB BF16 — zephyr_z9 · 2026-07-28
- A 70B GGUF model stalls on an AMD R9700 as VRAM hits 30 GB but RAM stays flat — Developer-Y · 2026-07-28
- Reddit debates the cheapest practical way to run K3 locally — ylchao · 2026-07-28
- Netlify Says 60% of New Users Deploy Their First Site Through Drop — thisiskp_ · 2026-07-28
- K3 hit a usage-limit exploit on Higgsfield within a day of launch — Mediocre-Witness-778 · 2026-07-28