The rise of XPUs: Exploring the hardware behind modern AI
Understand how CPUs, GPUs, NPUs, and other processors fit different AI workloads.
Back to all posts
Topic
AI
Explore 10 articles about AI and more.
My writing on AI covers the hardware behind models, where inference runs, and how to deploy and manage it. You’ll also find experiments with generative AI and perspectives on how AI agents affect open-source communities.
Start with AI infrastructure
Understand the hardware
Choose where inference runs
Where should your AI run?
Explore how latency, memory, connectivity, and privacy affect edge, cloud, and hybrid architectures.
Explore deployment options
All the ways you can self-host an LLM: 7 options from a laptop to a GPU cluster
Compare seven approaches, from running a model on your laptop to deploying it across GPU infrastructure.
Manage model traffic
What are AI gateways, and do you even need them?
Decide when routing, provider failover, access controls, and cost tracking justify a gateway.
All AI articles (10)
How to write an eval your AI app can actually fail
• 12 min read
Your AI app builds, the endpoint responds, and Claude has run another smoke test. Let's find out whether the app actually does its job.
Where should your AI run?
• 13 min read
An interactive exploration of edge, cloud, and everything in between, where you build an architecture, change the constraints, and see what breaks.
Turning Spotify history into generative art
• 10 min read
How a Spotify track ID, album art, playback state, and listening history become a dithered generative composition.
The slow death of open source in the age of AI agents
• 11 min read
AI Agents are here; but the bigger question is... Is Open-source ready? If so, why are maintainers starting to close the door?
The rise of XPUs: Exploring the hardware behind modern AI
• 10 min read
A practical guide to XPUs, comparing CPUs, GPUs, TPUs, NPUs, DPUs, VPUs, ASICs, and FPGAs across modern AI workloads.
All the ways you can self-host an LLM: 7 options from a laptop to a GPU cluster
• 12 min read
Compare options for self-hosting LLMs locally, on GPUs, in containers, at the edge, and across cloud or on-prem infrastructure.
Killing DeepSeek's bias
• 6 min read
Are models inherently biased? Read on to meet the middleware that decides what your AI is allowed to say; and how you can fight back!
What are AI gateways, and do you even need them?
• 5 min read
The AI gold rush has led to a very modern problem: too many shovels (models), too little gold (control). Are AI gateways the solution?
Leaked system prompts of popular LLMs and what they tell us
• 6 min read
Read on for actionable advice on crafting the perfect prompt for the LLM you love!
DeepSeek is here. Is it production-ready?
• 3 min read
Read on to learn more about this tech disruptor and how you can choose the right LLM for your project!