Arm and Qualcomm Eye Alphawave as Acquisition Target

On April 1, Reuters reported that Arm, the world's largest semiconductor IP provider and a subsidiary of SoftBank, is exploring the acquisition of Alphawave, the fourth-largest semiconductor IP provider globally. The move aims to strengthen Arm's AI processor capabilities by securing Alphawave's SerDes (Serializer/Deserializer) technology, a crucial component for high-speed data transmission in AI, high-performance computing (HPC), and 5G applications.

SerDes technology is essential for AI-driven data centers, as it enables high-speed, point-to-point serial communication while optimizing bandwidth and reducing costs. Alphawave, along with Broadcom, holds a strong position in this sector, with major clients like Google and OpenAI relying on high-speed SerDes for custom AI chips. The custom chip market is projected to reach $60 billion by 2028, making SerDes a key enabler of AI growth.

In March 2024, Alphawave expanded its optical-electrical product lineup, introducing 3nm DSPs for 800G and 1.6T links and coherent optical modulation—critical technologies for AI-driven data centers. Arm, which currently lacks advanced SerDes technology, would benefit from an acquisition to complement its existing CPU, GPU, and NPU IP portfolio.

According to insiders, developing high-end SerDes from scratch takes at least two years and requires deep technical expertise, making an acquisition a faster and more strategic option. Arm and Alphawave have already collaborated in the past, including joint development of AI/HPC-focused chips and participation in the Arm Total Design ecosystem, which integrates PCIe Gen 6/7, UCIe, and HBM.

However, sources indicate that Arm ultimately decided to withdraw from acquisition talks after preliminary discussions. Meanwhile, Qualcomm has emerged as a new contender, reportedly submitting a buyout offer for Alphawave. Under UK takeover rules, Qualcomm must submit a formal bid by April 29 or withdraw from the process.

Following reports of a potential acquisition, Alphawave's stock surged 21% in London, marking its biggest jump since September 2021, before closing up 14.84% at 107.38 pence, giving it a market capitalization of $1.04 billion.

AMD Completes $4.9 Billion Acquisition of ZT Systems

On March 31, AMD announced the completion of its $4.9 billion acquisition of ZT Systems, a leading provider of hyperscale computing and AI infrastructure. The deal, structured as a cash-and-stock transaction, includes up to $400 million in contingent payments based on performance milestones.

The acquisition will allow AMD to integrate its CPU, GPU, networking chips, and ROCm software into ZT Systems' server designs, accelerating the development of end-to-end AI infrastructure solutions.

ZT Systems, which ships over 100,000 servers annually, has previously relied on Intel and NVIDIA chips. However, with AMD's acquisition, the company is expected to shift its focus toward AMD-powered AI servers, reducing its dependence on competitors.

AMD projects that the ZT Systems acquisition will be accretive on a non-GAAP basis by the end of 2025. The ZT Systems design team will join AMD's Data Center Solutions Group, led by Executive Vice President Forrest Norrod.

Additionally, ZT Systems founder Frank Zhang will join AMD as Senior VP of ZT Manufacturing, overseeing the planned divestiture of ZT's U.S. data center infrastructure business in 2025. Former ZT Systems President Doug Huang will serve as Senior VP of Data Center Platform Engineering, spearheading server design and customer support for AMD's data center AI solutions.

“For AI innovation, reducing the time-to-deployment for large-scale AI systems is a critical advantage,” said Forrest Norrod, EVP of AMD's Data Center Solutions Group. “Acquiring ZT Systems is a major milestone in our AI strategy, allowing us to optimize end-to-end solutions tailored to customer environments while leveraging our open ecosystem approach.”

As AI workloads continue to expand, ZT Systems' expertise in high-volume server production is expected to strengthen AMD's position in the hyperscale AI data center market, intensifying competition with Intel and NVIDIA.