Read the announcement
Read the announcement
Get started
Get started
October 21, 2026
From data center to cloud: Crusoe announces new capabilities across its vertically integrated infrastructure
October 16, 2026
Quantized weights, full precision: The math behind 75% cheaper fine-tuning
October 14, 2026
Crusoe Spark AI factories achieve NVIDIA Exemplar-level performance
October 13, 2026
Crusoe expands NVIDIA DSX collaboration with open-source NVIDIA Run:ai

Our
Blog

Latest stories

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
All
Cloud
Engineering
General
Eight isometric tiles with open model logos, including DeepSeek, Meta, and OpenAI, linked in a loop on dark green
October 21, 2026
Open models, end to end: Introducing Crusoe Managed RL and Managed MLflow
Erwan Menard
All
Data center
General
Cloud
Four panels: a lit server rack, an aerial view of a data center campus, a worker grinding metal, and a wind turbine
October 21, 2026
From data center to cloud: Crusoe announces new capabilities across its vertically integrated infrastructure
Cully Cavness
All
Cloud
Engineering
Surf report image
October 19, 2026
AI infrastructure spend: a framework for evaluating cost, risk, and vendor choice
Omar Lari
Rohan Mathews
Michael YenChi Ho
All
Engineering
Cloud
Wireframe illustration of stacked cube blocks, representing model weights kept in compact 4-bit form during LoRA fine-tuning
October 16, 2026
Quantized weights, full precision: The math behind 75% cheaper fine-tuning
Daniel Shen
Suman Debnath
Lei Ma
All
Engineering
Cloud
Isometric illustration of modular blocks fitting into a larger structure, representing Crusoe Spark modular AI factories
October 14, 2026
Crusoe Spark AI factories achieve NVIDIA Exemplar-level performance
Sanchit Pathak
Young Jeong
Bill Stube
Alex Akesson
All
Cloud
Engineering
Isometric illustration of a cube dropping into a matching slot, representing NVIDIA Run:ai scheduling workloads on Crusoe CMK
October 13, 2026
Crusoe expands NVIDIA DSX collaboration with open-source NVIDIA Run:ai
Avinash Mandava
Michael YenChi Ho
All
Engineering
Cloud
Isometric illustration of a cloud over server racks, representing Crusoe's inference stack tuned for energy efficiency
October 9, 2026
How green is Crusoe's inference engine?
Daria Jammal
Tavor Lavi
Megha Jain
Hui Wen Chan
All
Data center
General
Thirteen people, most in green Crusoe T-shirts and name tags, stand in the Claude ISD gym during a community open house
October 9, 2026
Building community in the Lone Star State
Erica Currie
All
Engineering
Cloud
Isometric cube grid with one block lifted out, representing how GPU work is split into blocks of threads
October 7, 2026
GPU programming with Triton, part 1
Suman Debnath
Daniel Shen
Janaki Ram Goteti
Connor Guerrero
Lei Ma
All
Cloud
Engineering
Isometric illustration of a CPU chip surrounded by performance charts, representing fastokens tokenizer benchmarks.
September 30, 2026
White glove tokenization for frontier models
Alon Kejzman
Omri Berkovitch
Omer Landau
Michael YenChi Ho
All
Data center
Energy
Aerial view of the Crusoe-developed Google data center campus under construction in Armstrong County, Texas, with cranes and steel framing at sunset.
September 28, 2026
Optimizing and expanding American energy infrastructure through co-located power for AI data centers
No items found.
All
Cloud
Engineering
Isometric illustration of server racks and a metrics dashboard representing NVIDIA HGX H200 systems on Crusoe Cloud
September 28, 2026
10,195 AI clips in one weekend, powered by clean energy
Rohit Kalmankar
No results found

Scaling inference doesn’t need faster GPUs

Learn how to skip the inference bottleneck.

Are you ready to build something amazing?

Contact us