On Monday, Broadcom (AVGO) stock surged by over 11%, a performance primarily driven by the market’s continued optimism about its AI-driven growth prospects. This stock price increase moved in tandem with a spike in Alphabet (GOOGL) stock, which continues to gain market favor following the release of its new-generation AI model, Gemini 3.
Gemini 3, the latest AI model from internet search giant Alphabet, demonstrates several exceptional characteristics, including more powerful contextual understanding and advanced reasoning capabilities, enabling it to complete complex tasks without excessive prompting. This technological advancement has garnered widespread industry attention. Salesforce CEO Marc Benioff specifically highlighted Gemini 3’s significant advantages over OpenAI’s ChatGPT on social media, stating it represents an astonishing leap in reasoning speed, image and video processing, among other areas.
This breakthrough progress by Alphabet in the field of artificial intelligence has brought direct benefits to Broadcom’s semiconductor business. As a partner of Alphabet, Broadcom is responsible for producing custom Tensor Processing Units (TPUs). These specialized chips can effectively accelerate artificial intelligence workloads, particularly excelling in model training. Alphabet not only uses these TPUs for its own AI needs but also provides this core hardware to cloud computing customers through the Google Cloud Platform.
Regarding its business layout, Broadcom recently launched a 128G optical fiber channel platform for AI workloads, including the Brocade X-series controllers and the Brocade G820 switch. These new products are positioned as an industry first in the tech sector, with their eighth-generation optical fiber channel technology designed to provide storage security for the quantum era.
Although Broadcom is involved in multiple business segments within the technology sector, its AI semiconductor division is undoubtedly the most prominent growth engine. The core competitiveness of this division lies in its custom AI accelerators. These computing units achieve precise workload optimization through deep collaboration with end-users. Compared to the performance loss issues associated with NVIDIA’s (NVDA) graphics processors, Broadcom’s custom solutions can provide superior performance at a more competitive price, thereby significantly reducing the total cost of large-scale AI builds.
Financial data confirms this growth trend: In the third quarter of fiscal year 2025, Broadcom’s AI revenue surged 63% year-over-year to $5.2 billion. Management further anticipates fourth-quarter revenue will climb to $6.2 billion, fully reflecting the strong market demand for its custom AI hardware. As the AI competition continues to heat up, these semiconductors specifically optimized for AI are expected to gain broader market recognition, providing sustained momentum for Broadcom’s future business development.
