Very recent development shows that SK Hynix pulled ahead their HBM4 schedule and is ready for Nvidia’s Rubin GPU, continues its leadership in the market with more than 50%. Samsung is following up very quickly on HBM4 as well leveraging their memory/logic foundry integration (HBM4 bottom die uses logic processing), expected to increase its market share to over 30%. This squeeze MU to 3rd place with market share expected to shrink to less than 10%. CEO tried to clarify their HBM4 delay rumors and use of their internal fab as a temporary solution to the base logic die. Reading between the lines, it confirms their lag behind the competition. More importantly, they dont have enough volume. Given AVGO got crashed by the Wall Street even as an absolute leader in AI ASIC, interconnect switches, a
The Fed confirmed they are injecting liquidity by purchasing $40 billion in short-term Treasuries over the coming month. Operations officially started today, Dec 12. While the market is focusing on Powell's comments, the plumbing is getting fixed. The effects of liquidity ops usually lag by a few weeks.