CONNECT WITH US
Sep 23
Gigabyte's Giga Computing sets up AI factory in Taiwan to accelerate rack validation

Gigabyte's AI server shipments continue to accelerate, lifting revenue and profit while prompting capacity expansion. For the first half of 2026, revenue reached NT$249.272 billion (US$7.9 billion), up 48.4% year-over-year; net profit after tax was NT$11.875 billion (US$373.3 million), up 92.6%; and earnings per share (EPS) hit NT$17.73, all record highs for the same period.

Chinese President Xi Jinping is set to make his first state visit to the US since 2015, with the Trump administration rolling out high-level welcoming ceremonies, a state dinner and other diplomatic pomp. But industry observers and think tanks say the real goal is not to settle structural disputes over trade, AI, rare earths and Taiwan in one shot, but to keep US-China competition from spiraling further out of control.

Anthropic will open its Singapore office in October, making it the company's fifth office in the Asia-Pacific region after Tokyo, Bengaluru, Seoul, and Sydney. The move underscores the AI company's push into Singapore as demand for Claude surges across startups, large enterprises, and government agencies.

Anthropic has announced a third-party on-site evaluation program, bringing in staff from Faculty, the AI company owned by Accenture, to test model safety measures with near-employee access. The two sides have each committed at least US$1 billion over the next five years to build what Anthropic calls a third-party embedded evaluation mechanism.

Qualcomm unveiled two flagship mobile platforms at the Snapdragon Summit 2026, including the Snapdragon 8 Elite Extreme Gen 6, which changes its packaging approach to improve heat dissipation. During a media Q&A, Qualcomm Senior Vice President and General Manager of Mobile Handsets Chris Patrick said the new package still falls under existing Package on Package (PoP) technology, but the memory is no longer stacked directly above the die and is shifted to the side to create a thermal path.
Supermicro begins shipping Nvidia Vera Rubin NVL72 AI racks
Sep 24, 10:11
Supermicro has begun shipping Nvidia's next-generation Vera Rubin NVL72 systems, moving the rack-scale platform from product demonstrations into customer deployments as AI infrastructure increasingly requires integrated computing, power, and liquid-cooling systems.
Qualcomm's PickNik deal brings robotics software to its edge AI push
Sep 24, 10:02
MoveIt's integration with Qualcomm's Dragonwing and Arduino platforms could give robot builders a more direct route from on-device AI to motion planning and control, while also testing whether the open-source project can retain its cross-platform standing under new ownership.
Intel is positioning closer coordination between CPUs, GPUs and other computing resources as a key requirement for the next phase of AI, as agentic AI shifts workloads from model inference toward complex task execution.
While generative AI has surged into mainstream consumer use thanks to ChatGPT, the adoption of AI across industrial sectors has progressed at a much slower pace. Industry analysts point out that the primary culprit is the tendency of current AI architectures to "hallucinate," a fatal vulnerability in manufacturing environments where there is no tolerance for error. As industrial AI development cannot rely on large language models (LLMs) alone, attention is rapidly shifting toward industry world models that anchor AI systems in the laws of physics. Recognizing this pivot, global industrial software giants are already aggressively positioning themselves in this space.

GMI Technology, a Taiwan-based electronics component distributor and application solutions provider, said its AI computing rental business is booming, with strong demand from US technology companies for computing power. The company said large users tend to secure computing capacity through long-term contracts, leaving the market short of supply.

When Donald Trump and Xi Jinping discuss artificial intelligence, the deeper divide may not be over whether AI needs rules, but over what those rules are designed to protect.

Microsoft is changing how companies evaluate their generative AI investments, as the technology evolves from its beginnings as a productivity tool into deeper uses in core business workflows. In addition to token consumption and model costs, Microsoft Taiwan general manager Sean Pien points out that corporations are also examining how AI actually benefits workflows and innovation, making return-on-investment (ROI) evaluations harder to judge solely on the basis of time or labor savings.