Scientists have made a big step forward in data storage technology, they've managed to improve the manufacturing process for 3D NAND flash memory. This type of storage technology stacks memory cells on top of each other to obtain higher data density. A team of experts from Lam Research, the University of Colorado Boulder, and Princeton Plasma Physics Lab came up with a better way to etch (the process of carving holes into alternating layers of silicon oxide and silicon nitride) by using hydrogen fluoride plasma. This new method cuts vertical channels through silicon-based materials twice as fast as before achieving 640 nanometers in just one minute.

The team found out that mixing in certain chemicals like phosphorus trifluoride helps the etching process. They also learned that some byproducts can slow down etching, but adding water can help fix this problem. "The salt can decompose at a lower temperature when water is present, which can accelerate etching", said Yuri Barsukov, a former PPPL researcher now working at Lam Research. This breakthrough is important as the need for data storage received a huge boost with the rise of AI programs, that need tons of storage.

The gaming industry is about to get massively disrupted. Instead of using game engines to power games, we are now witnessing an entirely new and crazy concept. A startup specializing in designing ASICs specifically for Transformer architecture, the foundation behind generative AI models like GPT/Claude/Stable Diffusion, has showcased a demo in partnership with Decart of a Minecraft clone being entirely generated and operated by AI instead of the traditional game engine. While we use AI to create images and videos based on specific descriptions and output pretty realistic content, having an AI model spit out an entire playable game is something different. Oasis is the first playable, real-time, real-time, open-world AI model that takes users' input and generates real-time gameplay, including physics, game rules, and graphics.

An interesting thing to point out is the hardware that powers this setup. Using a single NVIDIA H100 GPU, this 500-million parameter Oasis model can run at 720p resolution at 20 generated frames per second. Due to limitations of accelerators like NVIDIA's H100/B200, gameplay at 4K is almost impossible. However, Etched has its own accelerator called Sohu, which is specialized in accelerating transformer architectures. Eight NVIDIA H100 GPUs can power five Oasis models to five users, while the eight Sohu cards are capable of serving 65 Oasis runs to 65 users. This is more than a 10x increase in inference capability compared to NVIDIA's hardware on a single-use case alone. The accelerator is designed to run much larger models like future 100 billion-parameter generative AI video game models that can output 4K 30 FPS, all thanks to 144 GB of HBM3E memory, yielding 1,152 GB in eight-accelerator server configuration.