# Memory Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/memory
> Markdown URL: https://aitinkerers.org/technologies/memory.md
> Technology record last updated: 2026-02-27T04:09:06Z
> Generated: 2026-09-22T09:45:31Z

Memory is the core semiconductor technology (RAM, ROM, Flash) that provides rapid, temporary, or permanent storage for a computer's working data and machine code.

Memory technology is the critical link for data access, categorized primarily by volatility and speed. Volatile memory, like high-speed SRAM (used in CPU caches) and high-density DRAM (Dynamic Random-Access Memory), holds data only while powered: DRAM access times are typically around 10 nanoseconds. Non-volatile memory, such as NAND flash (used in SSDs), retains data without power, offering massive storage capacity at a lower cost per gigabyte. Emerging technologies like MRAM (Magnetoresistive RAM) and PCM (Phase-Change Memory) are actively bridging the performance gap, aiming to combine the speed of DRAM with the persistence of Flash for next-generation computing systems.

- Official technology site: https://en.wikipedia.org/wiki/Random-access_memory
- Public AI Tinkerers demos and talks: 3
- Result page: 1 of 1

## Recent Public Talks and Demos

### [From Laptop to Supercluster: The New Era of Personal AI Supercomputing](https://paris.aitinkerers.org/talks/rsvp_tpKMaxLqXkQ)

AI development is shifting rapidly from centralized cloud infrastructure to powerful local AI systems that enable developers to build, test, and run large models directly on their desks. Lenovo’s new AI workstation portfolio powered by NVIDIA Blackwell and Grace-Blackwell architectures introduces a new class of systems—from compact personal AI appliances to multi-GPU developer workstations capable of running hundreds-billion-parameter models locally. This talk explores how AI development is moving closer to the developer, enabling faster iteration, lower cloud costs, and new experimentation workflows. We will walk through the new Lenovo AI workstation stack, including the ThinkStation PGX (Grace-Blackwell GB10 superchip) and Blackwell GPU developer workstations, and explain how these systems scale from personal AI experimentation to enterprise-grade model development. Questions the Talk Will Answer What does “personal AI supercomputing” actually mean for developers? How large of an AI model can realistically run locally on a workstation today? When should developers use local AI vs cloud GPU clusters? How do systems like PGX, P3, P5, P7, and PX map to different AI workloads? What does the Grace-Blackwell architecture change in AI workstation design? How can developers prototype and iterate faster using local AI hardware?

- Event context: High-Performance Local AI Development: Kick-off ThinkStation PGX — 2026-03-17 — Paris
- Public talk page: https://paris.aitinkerers.org/talks/rsvp_tpKMaxLqXkQ

### [From Autonomy to Observability: Running AI Agents Safely on Your Own Machine](https://seattle.aitinkerers.org/talks/rsvp_NbyMxlacro8)

AI agents are getting better at taking action, but most local setups still behave like black boxes. Once an agent can run commands, open ports, or spin up background processes, the real challenge isn’t intelligence, it’s knowing what’s actually happening on your machine. In this demo-driven talk, I’ll show how I’ve been experimenting with local-first AI agents that execute real system actions, and why I had to build observability alongside execution. I’ll demo two open-source tools: Deskmate, a local execution agent, and Riva, a local monitoring layer that makes agent behavior visible. We’ll look at real examples of agent workflows, failure modes, resource usage, and how visibility changes how much you can trust autonomy. The goal is not to pitch a framework, but to share lessons learned from running agents locally and open up a discussion on what responsible agent tooling should look like.

- Event context: AI Tinkerers Seattle Meetup: Dev Tools Track — March 9th, 2026 — 2026-03-10 — Seattle
- Public talk page: https://seattle.aitinkerers.org/talks/rsvp_NbyMxlacro8

### [Why do agents keep making the same mistakes? 🤯 Let's fix it!](https://munich.aitinkerers.org/talks/rsvp_8L-sUSIzyVA)

My side project: An agent setup that uses your personal MCP servers to complete tasks, and continuously improves itself based on the tasks it solves. Consider the task to prepare the meeting notes for today's upcoming meetings - here's the prompt you need to get it done: ** Out-of-the-box agents without self-improving ** "Fetch the meetings from my calendar for today. For each meeting, check whether it's an internal meeting or with a customer. For customer meetings, add an entry in the "Meeting Notes" notion database. Select the template for discovery meetings, and fill it out with the following information about the customer: First, find what their company does. Next, summarize ..." and so on. ** Self-improving agents ** "Prepare my meeting notes for today"

- Event context: AI Tinkerers Munich - July 25 — 2025-07-25 — Munich
- Public talk page: https://munich.aitinkerers.org/talks/rsvp_8L-sUSIzyVA

## Related Technologies

- [Agent SDK](https://aitinkerers.org/technologies/agent-sdk) ([Markdown](https://aitinkerers.org/technologies/agent-sdk.md)) — 3 public demos
- [AI infrastructure](https://aitinkerers.org/technologies/ai-infrastructure) ([Markdown](https://aitinkerers.org/technologies/ai-infrastructure.md)) — 2 public demos
- [ChatGPT](https://aitinkerers.org/technologies/chatgpt) ([Markdown](https://aitinkerers.org/technologies/chatgpt.md)) — 83 public demos
- [Claude Code](https://aitinkerers.org/technologies/claude-code) ([Markdown](https://aitinkerers.org/technologies/claude-code.md)) — 215 public demos
- [CPU](https://aitinkerers.org/technologies/cpu) ([Markdown](https://aitinkerers.org/technologies/cpu.md)) — 3 public demos
- [Deskmate](https://aitinkerers.org/technologies/deskmate) ([Markdown](https://aitinkerers.org/technologies/deskmate.md)) — 1 public demo
- [Finetuning](https://aitinkerers.org/technologies/finetuning) ([Markdown](https://aitinkerers.org/technologies/finetuning.md)) — 1 public demo
- [GPU](https://aitinkerers.org/technologies/gpu) ([Markdown](https://aitinkerers.org/technologies/gpu.md)) — 12 public demos
- [GPUs](https://aitinkerers.org/technologies/gpus) ([Markdown](https://aitinkerers.org/technologies/gpus.md)) — 5 public demos
- [Grace-Blackwell](https://aitinkerers.org/technologies/grace-blackwell) ([Markdown](https://aitinkerers.org/technologies/grace-blackwell.md)) — 2 public demos
- [Grace-Blackwell GB10](https://aitinkerers.org/technologies/grace-blackwell-gb10) ([Markdown](https://aitinkerers.org/technologies/grace-blackwell-gb10.md)) — 1 public demo
- [Linux](https://aitinkerers.org/technologies/linux) ([Markdown](https://aitinkerers.org/technologies/linux.md)) — 7 public demos
- [macOS](https://aitinkerers.org/technologies/macos) ([Markdown](https://aitinkerers.org/technologies/macos.md)) — 6 public demos
- [MCP](https://aitinkerers.org/technologies/mcp) ([Markdown](https://aitinkerers.org/technologies/mcp.md)) — 129 public demos
- [Multi-PGX clustering](https://aitinkerers.org/technologies/multi-pgx-clustering) ([Markdown](https://aitinkerers.org/technologies/multi-pgx-clustering.md)) — 1 public demo
- [Node](https://aitinkerers.org/technologies/node) ([Markdown](https://aitinkerers.org/technologies/node.md)) — 99 public demos
- [NVIDIA](https://aitinkerers.org/technologies/nvidia) ([Markdown](https://aitinkerers.org/technologies/nvidia.md)) — 7 public demos
- [NVIDIA AI Software Stack](https://aitinkerers.org/technologies/nvidia-ai-software-stack) ([Markdown](https://aitinkerers.org/technologies/nvidia-ai-software-stack.md)) — 1 public demo