Tuesday☕️

14 Nov 2023 — 4 min read

Trending:

Markets:

APEC Summit:

The Asia-Pacific Economic Cooperation (APEC) summit, scheduled from Wednesday to Friday in San Francisco, is a high-profile international conference. This event garners significant global attention, with key world leaders such as President Biden and Chinese President Xi Jinping in attendance, discussing economic and policy issues relevant to the Asia-Pacific region.
In preparation for this important summit, San Francisco authorities conducted extensive clearances of homeless camps across the city's downtown areas. This action was revealed through internal emails among city officials who had discussed the operation in advance. The move, coming right before the summit, has raised questions, especially regarding the relocation of the homeless individuals who were affected by the sweep. The exact details of where these individuals were moved to remain unclear.

Nvidia’s New AI Chip:

Nvidia recently announced the H200, a new graphics processing unit (GPU) aimed at enhancing artificial intelligence model development, particularly for generative AI. This GPU is an evolution of the H100, which was notably used by OpenAI for training GPT-4.
The H200 stands out for its significant memory upgrade to 141GB of advanced "HBM3" memory, enhancing its capability for "inference" tasks like generating text, images, or predictions. Demand for such GPUs is high among major corporations, startups, and government entities, although they come at a steep cost, with H100 models priced between $25,000 and $40,000.
Nvidia's advancements in AI GPUs have fueled a surge in its stock value, rising over 230% in 2023, and the company anticipates a 170% increase in revenue for the fiscal third quarter compared to the previous year, projecting about $16 billion.
The H200 is set to be twice as fast as its predecessor in generating outputs, as demonstrated in tests with Meta's Llama 2 LLM. Slated for release in the second quarter of 2024, it will be competing with AMD's MI300X GPU, which also features enhanced memory for handling large AI models.

Statistic: