April 25, 2024 - BESIY
Amidst the technical jargon of semiconductor assembly and discussions of cyclical downturns, Besi's Q1 2024 earnings call revealed a significant development: a potential shift in the high-bandwidth memory (HBM) landscape that could position Besi as a key player in a DRAM revolution. While the call focused on the company's strength in advanced packaging, particularly in hybrid bonding, it was the subtle discussion of TCB Next, Besi's next-generation thermal compression bonding (TCB) system, and its implications for HBM that caught attention. Besi revealed that leading memory customers are pursuing a "parallel path" for the adoption of both hybrid bonding AND next-generation TCB assembly processes for HBM. This suggests that the memory industry is not solely betting on hybrid bonding, the supposed successor to TCB for HBM5 and beyond. This "parallel path" approach, with memory giants like Micron, Samsung, and SK Hynix evaluating both technologies side-by-side, indicates a crucial battle brewing in the high-stakes world of HBM. While hybrid bonding boasts superior performance and lower energy consumption, critical for data centers, TCB retains a cost advantage. Besi, uniquely positioned with leading solutions in both technologies, stands to benefit regardless of the victor. But here's where it gets intriguing: Besi explicitly stated that their TCB Next system boasts "tighter specs than other systems currently available in the market." This implies that Besi's TCB Next may be closing the performance gap with hybrid bonding, potentially disrupting the expected narrative of hybrid bonding's inevitable dominance in HBM5.
Could Besi's superior TCB Next solution prolong the relevance of TCB in HBM, even for HBM5? The recent JEDEC announcement allowing for a higher stack height in DRAM packages strengthens this possibility. This change allows for extended TCB usage (potentially up to 16 layers) in HBM4, directly competing with hybrid bonding in a segment previously thought to be exclusively transitioning to hybrid bonding. If Besi's TCB Next delivers performance close to hybrid bonding at a lower cost, it could become the preferred choice for some HBM4 applications and potentially even for initial HBM5 implementations.
Besi's "parallel path" observation implies that the memory giants are hedging their bets, unwilling to prematurely commit to a single technology. This hesitation creates a unique opportunity for Besi. Their expertise in both TCB and hybrid bonding offers a powerful competitive advantage, allowing them to cater to the evolving needs of all three major memory players, regardless of their final technology choices. This dual expertise becomes a formidable moat, solidifying Besi's position as the go-to supplier for HBM assembly solutions and positioning them for substantial market share gains.
The earnings call indicated that Besi expects to ship 25 to 35 hybrid bonding systems in Q2, a substantial number, but predominantly for logic applications. Notably, Besi expects to receive a second evaluation order for their TCB Next system, indicating growing interest from memory customers. While it's too early to quantify the potential impact on Besi's revenue, the "parallel path" observation suggests a more complex and potentially larger opportunity in the HBM market than previously anticipated. If Besi's TCB Next gains traction alongside hybrid bonding, it could significantly amplify their already strong growth trajectory in advanced packaging.
The following chart showcases Besi's projected shipments for Q2 2024, illustrating the current focus on hybrid bonding for logic applications and the potential emergence of TCB Next for memory.
Besi's Q1 earnings call revealed more than just solid financial performance. It hinted at a potential shift in the HBM landscape, where the expected dominance of hybrid bonding is facing a challenge from Besi's advanced TCB Next solution. This development could reshape the DRAM industry and propel Besi to the forefront of a memory revolution. Besi, a company known for its prudence and conservative guidance, rarely makes bold pronouncements. Their "parallel path" observation, therefore, should be taken seriously. It's a hidden signal, overlooked by most, that points to a significant opportunity with potentially explosive implications for Besi's future.
"Fun Fact: Thermal compression bonding (TCB) has been a workhorse of the semiconductor industry for decades, dating back to the 1960s. It's a testament to the technology's robustness and adaptability that it's still relevant in the cutting-edge world of high-bandwidth memory."