AlphaWire

ima_daily5min

Kimi K3 open source model released, good for Nvidia, Micron valuation discount needs to be repaired (UBS)

2026-07-21·ima-daily5min-0721-13-a6a6964b3e
Street Signal | Kimi K3 open source model released, good for Nvidia, Micron valuation discount needs to be repaired (UBS)

A UBS report pointed out that Kimi K3 released by Moonshot AI, as the largest open source model, will drive the growth in demand for open source models and benefit computing power manufacturers such as NVIDIA (NVDA).

At the same time, the valuation discount of Micron (MU) compared to SK Hynix is not in line with historical rules, and it will generate huge free cash flow for many years in the future, with repurchase potential.

The report also analyzes the recovery cycle pricing of analog chips and updates industry congestion data, pointing out that storage and some computing stocks are highly crowded by longs, and smartphone-related stocks are highly crowded by shorts. The market is paying attention to the demand for AI computing power.

The release of Kimi K3 has strengthened this logic, and Micron's valuation depression relative to SK Hynix has become a new focus.

One-sentence conclusion: The release of Kimi K3 once again confirmed the strong demand for computing power in the AI open source model, which is good for Nvidia; at the same time, Micron's significant valuation discount to SK Hynix, combined with its strong cash flow prospects, makes it a very attractive value target in the storage sector.

Positive/negative: Positive for Nvidia (NVDA) and Micron (MU). Micron's valuation discount has not yet been fully repaired by the market, and the positive impact of the Kimi K3 incident on Nvidia may have been partially priced in. Catalysts:

1) Actual number of users and inferred demand data for open source models such as Kimi K3;

2) Micron (MU) next quarter financial report, pay attention to updates on its free cash flow and repurchase plan;

3) Simulate the monthly revenue data and order visibility of the chip industry to judge the recovery process.

Full text

Kimi K3 open source model released, good for Nvidia, Micron valuation discount needs to be repaired (UBS)

A UBS report pointed out that Kimi K3 released by Moonshot AI, as the largest open source model, will drive the growth in demand for open source models and benefit computing power manufacturers such as NVIDIA (NVDA).

A UBS report pointed out that Kimi K3 released by Moonshot AI, as the largest open source model, will drive the growth in demand for open source models and benefit computing power manufacturers such as NVIDIA (NVDA). At the same time, the valuation discount of Micron (MU) compared to SK Hynix is not in line with historical rules, and it will generate huge free cash flow for many years in the future, with repurchase potential. The report also analyzes the recovery cycle pricing of analog chips and updates industry congestion data, pointing out that storage and some computing stocks are highly crowded by longs, and smartphone-related stocks are highly crowded by shorts. The market is paying attention to the demand for AI computing power. The release of Kimi K3 has strengthened this logic, and Micron's valuation depression relative to SK Hynix has become a new focus. One-sentence conclusion: The release of Kimi K3 once again confirmed the strong demand for computing power in the AI open source model, which is good for Nvidia; at the same time, Micron's significant valuation discount to SK Hynix, combined with its strong cash flow prospects, makes it a very attractive value target in the storage sector. Positive/negative: Positive for Nvidia (NVDA) and Micron (MU). Micron's valuation discount has not yet been fully repaired by the market, and the positive impact of the Kimi K3 incident on Nvidia may have been partially priced in. Catalysts: 1) Actual number of users and inferred demand data for open source models such as Kimi K3; 2) Micron (MU) next quarter financial report, pay attention to updates on its free cash flow and repurchase plan; 3) Simulate the monthly revenue data and order visibility of the chip industry to judge the recovery process.

← Back to archive