The demand for artificial intelligence (AI) servers is surging globally, driven by substantial investments from cloud service providers keen to train AI models and deploy these innovations in real-world settings. According to market research by TrendForce, the AI server market is anticipated to reach a staggering $187 billion in revenue by 2024, marking a 69% increase compared to the previous year. This growth offers significant prospects for multiple industries involved in AI server components and technology.

Among the key players capitalising on this boom are Micron Technology and Marvell Technology. These companies provide critical components that enhance the performance and efficiency of AI servers, underpinning the expansion of this burgeoning market.

Micron Technology's Advances in High-Bandwidth Memory

Micron Technology, a prominent semiconductor company, has positioned itself strategically within the AI server market through its production of high-bandwidth memory (HBM) chips. These chips are crucial for AI operations, particularly in AI server chips like graphics processing units (GPUs), facilitating quicker data transfer, reduced processing times, and lower power consumption.

The company reports that demand for its HBM is so robust that its capacity is fully booked for both the current and forthcoming year. Micron's latest iteration, the HBM3E chip, promises to outpace competitor offerings by providing a 20% reduction in power consumption and a 50% enhancement in capacity. The company forecasts launching mass production of the HBM3E by early 2025, intending to scale up output throughout that year.

Reports suggest that by next year, Micron aims to secure 20% to 25% of the HBM market. With the HBM market's revenue expected to rocket to $25 billion by 2025 from a mere $4 billion in 2023, this could considerably bolster Micron's growth, projecting a 52% revenue increase to $38 billion for their fiscal year starting August 30.

Marvell Technology's Rise through Custom AI Chips

Marvell Technology is gaining traction thanks to its application-specific integrated circuits (ASICs), custom chips designed specifically for AI server needs. Major cloud service providers, including Meta Platforms, Google, and Amazon, are increasingly opting for ASICs to cut costs by developing proprietary processing solutions.

ASICs are projected to constitute 26% of the AI server chip market by 2024, creating a revenue opportunity that could reach $150 billion. Despite a 5% year-on-year revenue decline to $1.27 billion for the fiscal second quarter ending August 3, Marvell recorded a 92% increase in data centre revenue, up to $881 million, highlighting the potential of AI-driven growth areas.

CEO Matt Murphy has noted that Marvell's AI silicon programmes are advancing positively, with significant projects moving into production phases. The company anticipates its data centre business to experience substantial growth in the upcoming quarter, forecasting revenue improvement to $1.45 billion compared to the previous year's $1.42 billion in the same period.

Overall, both Micron and Marvell are set to benefit immensely from the AI server market's anticipated rise over the next few years. Their innovative components and growing capacities highlight their crucial roles in realizing the potential of AI in cloud computing and related technologies. As such, they represent critical stakeholders in the ongoing expansion and capabilities of AI-driven infrastructures worldwide.

Source: Noah Wire Services