This blog post is written by ABI Research Industry Analyst Paul Shell, who was invited as a guest to the Snapdragon Summit. All experiences were hosted, but no additional compensation was received.

Artificial Intelligence (AI) server manufacturers have experienced surging demand as data center operators require significantly more computing power than before the advent of ChatGPT and other Generative Artificial Intelligence (Gen AI) tools. Enterprises are investing billions of dollars in cloud infrastructure and Gen AI/Agentic AI development, fueling strong demand for high-performance AI servers.

The Artificial Intelligence (AI) server market is entering a pivotal growth phase. ABI Research forecasts that the global market size will reach US$245 billion in 2025, up 25% from 2024. By 2030, AI server sales will grow even further, pushing the market to US$524 billion, representing an 18% Compound Annual Growth Rate (CAGR). Dell, Hewlett-Packard Enterprise (HPE), Inspur, and Lenovo are market leaders.

Qualcomm is strategically leveraging its heritage in connectivity to carve out a broad position in the Artificial Intelligence (AI) landscape. Its approach is about embedding intelligence seamlessly from the network edge into a vast array of devices, from battery-powered embedded devices and edge AI “boxes” to AI Personal Computers (PCs) and now desktops, and finally the data center. COMPUTEX did not see any new silicon from Qualcomm, but it did offer a convention of partners and customers, and messaging about Qualcomm’s strategy as robust across most locations of compute with the exception of the data center.

NVIDIA was one of the few semiconductor vendors with significant announcements, and a larger presence over last year’s COMPUTEX. The company has forged closer ties with Taiwan, in no small part linked to Jensen Huang’s personal ties to the island, and this was also symbolically demonstrated by the decision to build another “Headquarters (HQ)” near Taipei named NVIDIA Constellation. This is also a manifestation of the divergence from the United States and the rest of the world brought on by the risk of trade restrictions, and a sign of Taiwan’s importance in the high-end technology supply chain.

A new frontline is the low-end Artificial Intelligence (AI) workstation. As with enterprise and consumer AI Personal Computers (PCs), this is not just about packing more Tera Operations per Second (TOPS); it is a strategic battle for AI developers, ecosystem dominance, and resulting market share in a segment poised for growth. At COMPUTEX 2025, both Intel and AMD unveiled their competitive strategies for monetizing more demanding enterprise AI workloads, such as fine-tuning and even small training workloads, that will reside outside of the cloud for cost, data privacy, and latency reasons.

• The Emergence of a New Battleground: Intel Versus AMD in AI Workstations
• NVIDIA's Strategy: Dominating AI Through Ecosystem, Access, and Interconnect
• Qualcomm's AI Strategy: Leveraging Connectivity Legacy Across Edge AI, Capturing AI PC Market Share, and a Foray into Data Center via NVLink Fusion

As we gear up for COMPUTEX 2025, the Artificial Intelligence (AI) Personal Computer (PC) narrative, which saturated the 2024 show with its narrow focus on Terra-Operations per Second (TOPS) and vendor benchmarking battles between Intel, AMD, and, increasingly, Qualcomm, remains under pressure to mature. The 2024 drive for raw performance was the first step, providing Independent Software Vendors (ISVs) with platforms for developing AI applications. The 2025 spotlight must shift toward genuine user experience and productivity enhancement. We'll be scrutinizing how the next wave of silicon, beyond simply meeting Microsoft's Copilot+ thresholds, enables applications that move beyond impressive but limited Proofs of Concept (PoCs) like stable diffusion demos. The key will be whether developers, armed with more robust tools and access to runtimes like Copilot+, can finally deliver the innovative, everyday creativity and productivity boosters that actually justify widespread device upgrades.

Key Takeaways:

• Cloud AI chipset revenue is surging. The market will grow 65% in 2025 to reach $63.6 billion and is expected to hit $137 billion by 2031, nearly quadrupling in size from 2024.
• Generative AI is the main growth driver. The rise of Gen AI, cloud computing, and AI-optimized silicon is fueling record demand for powerful chipsets that can train and run large AI models.
• GPUs lead the market. Graphics Processing Units remain the top choice for cloud AI workloads, while ASICs are gaining traction and CPUs are growing more slowly.
• Energy use and scalability are key challenges. Power constraints and data center bottlenecks are pushing the industry toward greener designs, hybrid setups, and edge computing.
• Hybrid and multi-cloud strategies are essential. Businesses should invest in flexible cloud infrastructures and monitor fast-moving AI chipset innovations to stay competitive.

According to the latest forecast from ABI Research, the cloud AI chipset market size is set to grow 65% in 2025, reaching US$63.6 billion. Growing at a 21% Compound Annual Growth Rate (CAGR) between 2024 and 2031,  the market is projected to reach US$137 billion by 2031. This translates to a nearly quadrupling of the market size compared to 2024,

The market is expanding rapidly, driven by advancements in Generative Artificial Intelligence (Gen AI), cloud computing, and AI-optimized silicon chipsets. Businesses and organizations are increasingly leveraging cloud AI to train complex Large Language Models (LLMs), run new applications, and integrate Machine Learning (ML) capabilities into their operations.

U.K.-based semiconductor Intellectual Property (IP) house Imagination Technologies recently made the headlines for securing a US$100 million investment to bolster the development of its edge Artificial Intelligence (AI) portfolio. The company has an illustrious legacy, once claiming the top position in Graphics Processing Unit (GPU) innovation, ahead of NVIDIA. In the early days of mobile platforms, Imagination’s footprint was immense: PowerVR IP was in a range of Apple products, including the first iPhone in 2007, MediaTek Systems-on-Chip (SoCs), gaming consoles, automotive systems, and more. By 2008, PowerVR graphics had been shipped in more than 100 million consumer devices, reaching 1 billion by 2013.