Demand for AI chip testing continues to heat up. ChipMOS (6510) today (3rd) announced its July consolidated revenue, surpassing the NT$600 million mark to reach NT$615 million, up 7.1% month-on-month and 49.9% year-on-year, setting a new single-month record. Cumulative consolidated revenue for the first seven months of the year reached NT$3.612 billion, up 30% year-on-year. As data centers and cloud service providers continue to expand investments in AI infrastructure and customers' next-generation products begin shipping, ChipMOS’s growth momentum is extending from GPU to ASIC, with testing demand for both chip types becoming a key driver for operations.
AI and HPC Shipments Heat Up – July Revenue Breaks NT$600 Million
ChipMOS stated that the record-breaking July revenue was primarily driven by strong shipment momentum in artificial intelligence (AI) and high-performance computing (HPC) related products. Compared to June, July revenue increased by 7.1%; compared to the same period last year, it surged by 49.9%, indicating continued intensification in AI chip testing demand.
Cumulatively, ChipMOS’s consolidated revenue for the first seven months of the year reached NT$3.612 billion, up 30% year-on-year. The company noted that it currently has “good order visibility,” and with AI applications continuing to drive demand in end devices and cloud markets, it expects full-year revenue to maintain growth momentum.
Agentic AI Fuels Compute Investment – GPU Demand Remains Strong
From a demand structure perspective, the wave of Agentic AI is prompting data centers and cloud service providers to further expand investments, thereby boosting demand for GPU-related chips, modules, and systems. Unlike traditional generative AI, which primarily responds to single commands, Agentic AI executes multi-step tasks, requiring greater inference computing and data processing capabilities.
As cloud providers accelerate the deployment of AI computing power, demand for GPUs and related systems continues, driving shipments of high-end chip test interface products. This has become one of the main drivers behind ChipMOS’s record-breaking July revenue.
ASIC Products Gradually Scaling Up – Forming Two Growth Engines
In addition to sustained strong GPU demand, as customers’ next-generation products begin shipping and edge computing and inference applications become increasingly diverse, ASIC chip-related applications are also gradually scaling up.
This signifies that ChipMOS’s AI growth momentum is expanding from the GPU, which has been the market’s primary focus, to customized ASIC chips. The simultaneous growth in GPU and ASIC testing demand not only broadens the application scope of AI-related products but also forms two key engines supporting future operations.
In response to the continued penetration of AI applications across industries, ChipMOS stated it will strengthen competitiveness through capacity expansion and technological upgrades, and flexibly adjust product mix and capacity allocation based on end-market demand changes to capture high-end chip testing demand.
AI Chip Power and Size Increase – Test Interfaces Evolve Accordingly
As AI chip computing performance and power consumption continue to rise, chip size, packaging complexity, and I/O count are also increasing, driving test interfaces toward higher power, higher pin count, and larger sizes.
ChipMOS will showcase AI test interface boards, high-power burn-in test boards, high-pin-count probe cards, and large-size ball grid array (BGA) packaging boards at SEMICON Taiwan, an international semiconductor exhibition taking place from September 2 to 4, demonstrating its strategic positioning in the AI chip and high-end semiconductor testing markets.
The company will also continue to leverage AI technology to optimize design processes, improve the design accuracy and efficiency of test interface products, and meet customers’ testing requirements for next-generation GPUs, ASICs, and advanced packaging products through key manufacturing capabilities.
FACT BOX
- Source: PR Times
- Category: 財務業績