Alibaba Cloud vs Canonical

Alibaba Cloud Named a Leader in Cloud AI Infrastructure Report

Alibaba Cloud offers comprehensive full-stack AI upgrade for the agentic era.

Memoria: Self‑Evolving Personal AI with Memory

Memoria is a production-ready personal AI MemoryAgent built for the Qwen Cloud Hackathon that implements human-like long-term memory: extraction, prioritisation, decay, consolidation, conflict resolution and reflection. It organises knowledge into three tiers (Session Memory in Redis, Personal Memory in PostgreSQL 16 + pgvector with text-embedding-v3, and a Context Archive for full transcripts). The system uses Qwen models (qwen-plus and qwen-max) for extraction and consolidation, a Python FastAPI backend, Celery workers with Redis broker, and a React frontend. Memoria was deployed on Alibaba Cloud (ECS, ApsaraDB, Redis) and provisioned via Terraform; the author reports a benchmarked 77.6% improvement in decision accuracy across 12 scenarios. Planned next steps include voice input, multi-agent collaboration (MCP), a mobile companion, and fine-tuning Qwen for memory tasks.

• Memoria was built as a production-ready MemoryAgent for the Qwen Cloud Hackathon (Track 1).
• Memory is organised in three tiers: Session Memory (Redis), Personal Memory (PostgreSQL 16 + pgvector with text-embedding-v3), and a Context Archive (full transcripts).
• The stack includes Python FastAPI backend, SQLAlchemy async, Celery background workers (Redis broker), and a React + Vite frontend; deployment on Alibaba Cloud ECS with ApsaraDB and Redis via Terraform.

Migrate Cloud TPU API Workloads to Compute Engine

This technical migration guide explains moving TPU workloads from Google Cloud's deprecated Cloud TPU API to Compute Engine instances. The Cloud TPU API is no longer under active development and future TPU hardware generations (starting with TPU7x) are supported only through Compute Engine or Google Kubernetes Engine. Migration requires flag and command mapping (e.g., accelerator-type -> machine-type, tpu-vm ssh -> compute ssh), checking different quota metrics (preemptible vs family quota) and provisioning models (FLEX_START, SPOT, STANDARD, RESERVATION_BOUND), and adjusting startup scripts and images (some Compute Engine accelerator images lack tools like docker). The guide documents practical troubleshooting: using SPOT to probe capacity, checking both quota metrics via the Cloud Quotas API, handling silent failures where RUNNING != ready, and other pitfalls encountered during real migrations.

• Google's Cloud TPU API is no longer under active development; new hardware generations starting with TPU7x are supported only via Compute Engine or GKE.
• Compute Engine uses different flags and flows (e.g., --machine-type=ct6e-standard-1t, --image-family, --request-valid-for-duration, --provisioning-model=FLEX_START) compared with the Cloud TPU API.
• Flex-start provisioning on Compute Engine consumes preemptible quota (PREEMPTIBLE-TPU-V6E-per-project-region) and falls back to the family quota; quota and capacity are separate and reported by different APIs.