DAOS Storage Stack (client libraries, storage engine, control plane)
-
Updated
Oct 15, 2026 - C
DAOS Storage Stack (client libraries, storage engine, control plane)
Autoscaling Azure SQL Database Hyperscale with Azure Functions
The open-source brain for personal AI agents. One agent for every user of your app, with memory, schedules, tools and approvals, on a hyperscale serverless architecture where idle agents cost only storage.
📝 Examples of how to use Neptune for different use cases and with various MLOps tools
A curated collection of papers describing real-world production systems from top systems conferences (2020+).
This workshop helps you understand Real-time data with Azure Database for PostgreSQL Hyperscale (Citus data)
The Hyperscale programming language
⚙️📡 High-throughput system drift detection.
⚖️ Enable decentralized voting on proposals with a Governance Token in this Ethereum-based DAO Governance Portal.
Public mission-critical field-discipline notes for electrical, controls, and infrastructure review boundaries.
SSD Hyperscale Simulation Test Framework with Behave, HTML Report, CI/CD and Docker
Robotics Test Driven Development (TDD) Framework (Simulation-Only) with Dockers and CI/CD
Robotics Behavior-Driven Development (BDD) Framework (Simulation-Only) with Docker, Kubernestes, Allure Report, and CI/CD
Hyperscale event-driven platform on AWS using EKS, Kafka, Terraform, and GitOps with autoscaling, observability, and production-grade infrastructure automation
DayOne — independent third-party profile of a public API surface, by API Evangelist. DayOne is a hyperscale data center company delivering next-generation digital infrastructure for cloud, AI, and edge computing workloads. Spun out of the GDS International business and backed by SoftBank Vision Fund 2, DayOne develops and operates strategically loc
A (failed) attempt to train the Clay AI Foundation model to detect data centers
The NEW home of butrelinux
A production-style Python prototype that treats a global AI data centre fleet as a living system: it fuses macro climate stress signals with asset telemetry, predicts next-hour thermal efficiency degradation with machine learning, and prescribes sustainability-aware workload redistribution — inside a mission-control Streamlit experience.
Cloud, storage, account, artifact and platform services.
To associate your repository with the hyperscale topic, visit your repo's landing page and select "manage topics."