Using GPT‑4o, researchers cut the average bins needed for 2D bin‑packing from 16 to 15 and raised space utilisation to about 0.83 after just two evolutionary iterations. https://getnews.me/evaluating-llms-on-2d-bin-packing-one-phase-and-two-phase-heuristics/ #llm #binpacking #evolutionary

Pondering optimal binpacking of people into coffeeshop seating areas while fielding oncall. http://www.google.com/buzz/106912596786226524817/Svqg7AYyCRt/Pondering-optimal-binpacking-of-people-into

If you’ve gotten this far and passed each of these in sequence, you may have noticed: each of these is exactly the acceptable length of a post on bluesky. It’s an attempt to maximize symbolic compression, binpacking a cohesive worldview into the shortest form possible. Each is a prompt for the next.

A new framework merging LLMs with evolutionary algorithms lowered the average bin count for 2‑D packing from 16 to 15 and lifted utilization to 0.83, achieving optimal solutions in two iterations. https://getnews.me/llms-boost-2d-bin-packing-efficiency-with-evolutionary-heuristics/ #llm #binpacking

for the most part, I don't regret dropping out of CS in college but damn if I really wished I'd taken a class on binpacking and every trick in that book of dark magicks

did you read the vercel blog about how they’re binpacking jobs onto lambdas

If you’ve gotten this far and passed each of these in sequence, you may have noticed: each of these is exactly the acceptable length of a post on bluesky. It’s an attempt to maximize symbolic compression, binpacking a cohesive worldview into the shortest form possible. Each is a prompt for the next.

Binpacking multiple unrelated containers onto a single machine is not a priority due to concerns about noisy neighbors; many teams are using entire single machines and optimized for it. #SystemsAtScale

Show HN: 3D Binpacking Algorithm Visualized (skusavvy.com)

Main Link | HN Post

🚛💡 Revolutionising Logistics with 2L-VRLP💡🚛

An innovative two-level vehicle routing and loading model.

2L-VRLP integrates #BinPacking and #VehicleRouting to optimise space utilisation, costs, and operational efficiency ➡️ decisionlab.co.uk/2024/09/19/i...

#Logistics #SupplyChain

Binpacking: an underused algorithm in the market research industry
allthingsdatascience.blogspot.com/2024/12/binp...
#mrx #datascience

Going to #KubeCon? Check out #HannahTaub's talk "The Node Tetris Rabbit Hole: Why Your #Binpacking Might Be Underperforming" You'll learn some #Kubernetes #utilization & #costefficiency techniques, tips & tricks. I'm lucky enough to work w/her & she's impressive!
kccncna2024.sched.com/event/1i7rY

Constant-factor approximation algorithms for maximin-share fairness in bin-packing were presented for ordinal models. The paper was submitted on 6 October 2025 and appears at WINE conference. https://getnews.me/new-algorithms-advance-maximin-share-fairness-in-bin-packing/ #maximinshare #binpacking

Have you tried asking ChatGPT to invent a quantum binpacking algorithm?

If you're at Kubecon this week, come join me on Friday to hear about my team's journey with improving our Kubernetes cluster utilization. I'll also share some tips and tricks for better binpacking for cost efficiency!

kccncna2024.sched.com/event/0f582c...

Binpacking the court

As a further example to the point: none of the three examples in your original post are unique to k8s.

Most services have a consistent API.

Service Discovery has existed for yonks via multiple methods.

Docker Swarm et al can do binpacking.

You want to know what makes k8s unique? I'm telling you.

another useful feature is Binpacking. by default, kuberenetes uses LeastRequestedPriority which prioritizes to schedule the Pod to least utilized hosts, and try best to spread the work load evenly.

unfortunately, when comes with GPU, this is likely to happen:

For example: they have ECS tasks reserving the equivalent of 60 vCPU and are needing 120 vCPU of servers to run/deploy comfortably even with a majority of the services using binpacking.