IP licensing revenue increased & automotive/mobile phone chip design projects went smoothly, 24Q3 revenue increased 17% month-on-month.

Revenue for the first three quarters of 24 was 1.65 billion yuan, down 6.50% year on year. Net profit to mother was -0.396 billion yuan, net profit not to mother was -0.422 billion yuan. 24Q3 single quarter revenue was 0.718 billion yuan, up 16.96% month-on-month, net profit to mother was -0.111 billion yuan, net profit not to mother was -0.118 billion yuan; gross profit margin was 40.07%, down 3.67 pct from month to month. The month-on-month increase in revenue is due to a month-on-month increase in IP licensing fees and chip design business revenue.

Q3 chip design revenue was +23% month-on-month, and customer inventory improvements benefiting from mass production orders remained at a high level.

By business, 24Q3 IP licensing revenue was 0.243 billion yuan, an increase of 32.35% over the previous month. Among them, IP license fee revenue was 0.218 billion yuan, an increase of 36.62% over the previous month. Specifically, 24Q3 semiconductor IP licenses were 53 times, and the average IP licensing fee was about 4.12 million yuan/time, an increase of 67.55% over the previous month.

24Q3 chip customization business revenue was 0.474 billion yuan, up 10.84% month-on-month. (1) The mass production business revenue was 0.236 billion yuan, a slight increase of 0.79% over the previous month. The main benefit was that downstream customer inventory conditions had improved markedly, and mass production orders remained at a high level. (2) The chip design business revenue was 0.238 billion yuan, an increase of 23.02% over the previous month, mainly benefiting from the smooth progress of advanced process design service projects at 14nm and below.

The amount of on-hand orders has exceeded 2 billion yuan for six consecutive quarters, laying a solid foundation for future performance transformation.

The 24Q3 company signed a new order of 0.648 billion yuan. Current orders have remained high for four consecutive quarters. As of the end of 24Q3, the company's on-hand orders were 2.138 billion yuan, of which the proportion converted into revenue within one year was 78.26%.

The company's order situation is good, laying a solid foundation for future performance transformation.

Actively embrace AIGC and automotive intelligence, expand the product matrix, and deepen cooperation with leading customers in the industry.

The company's processor IP can meet diverse AI computing needs. In the first three quarters of 24, IP licensing revenue related to AI computing power was 0.232 billion yuan, accounting for 48.59% of revenue. (1) After NPU IP is expanded through multiple convolutional computing cores, the computing power can reach 400 TOPs, which can be better used in the field of AI. As of 24H1, NPU IP has been used by 72 customers in 128 AI chips, and more than 0.1 billion AI chips with NPU IP have been shipped. (2) GPU IP can support large-scale general-purpose computing and AIGC related applications, and has been used and deployed by customers in various high-performance AI chips for data centers, high-performance computing, automobiles, etc.

In the first three quarters of 24, automotive electronics revenue was 0.172 billion yuan, up 50.19% year on year. As of 24H1, IP has been authorized to more than 20 automotive electronics customers around the world. Many world-renowned automotive OEMs use the company's GPUs for in-vehicle infotainment systems or dashboards, and NPU IP has also been used by many customers for ADAS products.

Domestic IP & ASIC unicorns embrace Chiplet's 100 billion market and maintain a “buy” rating.

The advent of AIGC such as ChatGPT marks an epoch-making moment for AI, and VeriSilicon is expected to fully benefit as the leading domestic IP &ASIC king. Net profit for 2024-2026 is expected to be -0.416/0.053/0.158 billion yuan, respectively, corresponding to 25/26 PE of 274.5/92.6 times, maintaining a “buy” rating.

Risk warning: geopolitical risk; technology licensing risk; intellectual property risk; technology iteration risk.