AI Infrastructure Spending Boosting Companies Across the Tech Sector
AI infrastructure spending is yielding substantial returns for firms involved in the supply of processors, networking equipment, servers, and electrical systems. Recent earnings reports underscore the varying business conditions facing technology companies, particularly in light of shifting market demands and a strong focus on artificial intelligence advancement.
For instance, Qualcomm is actively seeking to penetrate the data center market, addressing the contraction in its handset revenues. Meanwhile, Supermicro is swiftly ramping up server production, and manufacturers of electrical equipment are experiencing an increase in backlogs. This dynamic is particularly relevant for data center operators and IT procurement personnel, who are witnessing firsthand how the escalating demand for AI solutions impacts suppliers across computation, networking, servers, and electrical infrastructure. These shifts also bring to light the accompanying cost and capacity pressures that are being faced by industry players.
Chipmakers Respond to AI Compute Demands
AMD, a major player in the semiconductor space, recently reported an impressive third-quarter revenue of $13.4 billion, marking a 45% increase from the previous year. The Data Center segment alone generated an unprecedented $8.5 billion, representing a staggering 97% year-over-year growth rate. This surge can largely be attributed to the rising demand for EPYC server processors and Instinct GPUs, although AMD did not break down its revenue between server CPUs and AI accelerators. CEO Lisa Su confirmed that both product lines significantly contributed to this growth, emphasizing the company’s optimistic outlook moving forward.
“We delivered a record quarter with revenue exceeding the high end of our guidance as Data Center segment sales nearly doubled year-over-year,” Su stated in the company’s announcement of third-quarter earnings. Looking ahead, AMD is forecasting fourth-quarter revenue to be approximately $17 billion, with a potential year-over-year growth rate of about 66%.
On a different trajectory, Qualcomm reported an 8% decline in fiscal fourth-quarter revenue, totaling $10.32 billion, primarily driven by challenges within its handset business, which saw a notable 26% drop in revenue. However, there were bright spots too, as automotive revenue surged by 89% to nearly $2 billion, and IoT revenue saw a 4% increase, reaching $1.89 billion. Qualcomm President and CEO Cristiano Amon indicated that fiscal 2027 would signal a critical turning point for the company’s data center business.
“Fiscal 2027 marks the start of our nonhandset revenue inflection, including the ramp of our Data Center business,” Amon said, projecting nonhandset revenue growth of over 60% year-over-year, reaffirming a target of $5 billion in revenue for the Data Center segment by then.
However, this ambitious target comes amidst a backdrop of challenges, including increasing semiconductor input costs. Qualcomm has noted the rise in expenses related to wafer fabrication, advanced packaging, and other materials, which prompted the company to adjust product prices consequently. Senior analyst Shivani Parashar from Counterpoint Research anticipates Qualcomm’s data center revenue will reach over $650 million by the December 2026 quarter, with robust year-over-year growth likely to extend into 2028.
Networking Demand Driven by AI Clusters
In the networking sector, Arista Networks posted third-quarter revenue of $3.66 billion, showing a 58.5% year-over-year increase. CEO Jayshree Ullal referred to this performance as "momentous," attributing growth to hyperscaler deployments of 800G networking and Ethernet fabrics. As Arista’s Etherlink SU-144 architecture illustrates, the company is making strides in creating robust Ethernet architectures to support the growing demand for AI workloads across data centers.
According to Ron Westfall, vice president at HyperFrame Research, high-performance networking is expected to capture an increasingly significant portion of total AI capital expenditures as GPU clustering encounters communication bottlenecks. He pointed out that thermal management, optical interconnect density, and power consumption pose significant challenges for larger AI clusters, as high-speed transceivers may account for a substantial percentage of total cluster power.
Supermicro and Eaton’s Robust Growth
Supermicro has also reported impressive growth, with preliminary results indicating that its server business generated net sales of $19.9 billion for the quarter ended September 30, substantially up from $5 billion a year earlier. Although gross margins have declined, the company remains optimistic about the future, forecasting second-quarter revenues between $19.5 billion and $21.5 billion and aiming for full-year fiscal 2027 revenues in the range of $72 billion to $82 billion.
Eaton, focusing on the electrical side of infrastructure, reported third-quarter sales of $8.9 billion, representing a 28% increase year-over-year. The company’s Electrical Americas sector alone recorded $4.2 billion in sales, showcasing strong organic growth. Eaton’s rising backlog indicates burgeoning demand across its broader electrical markets, which include utility and industrial clients.
As the technological landscape continues to evolve, driven by the increasing significance of AI, these companies exemplify the diverse strategies being employed to capitalize on the burgeoning infrastructure needs. While challenges remain, particularly regarding costs and supply chain constraints, the overall trend indicates a robust future for AI infrastructure spending, further solidifying its place at the heart of technological advancement.