This is a random message on random Fridays about random topics. Typing fast, so pardon my random typos. Today’s Focus: AI, CUDA, and the Hardware Connection
- Updating an old project: Indeed, today’s focus is exactly the title of what is now a 4 years old experiment in my public GitHub. I'm updating to new libraries, showing interesting results, soon to be published. Yet, reading the old content is still worthwhile, even for me, showing the value of documentation.
- Apple MLX for Nvidia hardware. If only every AI announcement matched reality. Apple MLX is alive and kicking for Apple hardware. For the Nvidia support, maybe it is worth asking AI to “define soon”. Just saying pip install “mlx[cuda]” will work is a beautiful story… Maybe I’ll just try a DGX Spark.
- Quantum computing: yet another market for GPUs? It is becoming fascinating how many markets GPUs and CUDA can serve. Seeing videos about the NVQLink shows an interesting prospect. It is important however to hear from the skeptics.
Happy November!