Horizon Ventures Backs Kneron as AI Shifts from Cloud to Edge
SingTao · 1 SOURCESabout 5 hours ago5 MIN
Summary
Horizon Ventures founder Chow Ka-yuen made a rare public appearance at a UBS technology forum held at the Rosewood Hotel in Hong Kong on Sunday . The presence of the reclusive investor signaled strong backing for Taiwanese chip designer Kneron (耐能), whose local representative Jonathan chaired a session at the event . Kneron founder Liu Jun-cheng (劉峻誠) presented the company's vision for edge AI, arguing that most simple tasks should be processed locally rather than sent to power-hungry cloud data centers . The company announced expanded partnership with Indonesia's Sinar Mas Group to deploy sovereign AI infrastructure across Southeast Asia .
Key Points
- Horizon Ventures founder Chow Ka-yuen attended the UBS tech forum at Rosewood Hotel, where Kneron held a presentation session
- Kneron has developed two core technologies: reconfigurable architecture allowing chips to scale from smartwatches to servers, and dynamic memory allocation eliminating the need for expensive HBM memory
- The company claims to have shipped over 100 million NPUs over the past decade, with shareholders including Qualcomm, Sequoia Capital, Foxconn, and Delta Electronics
- Kneron announced deepened cooperation with Indonesian conglomerate Sinar Mas Group to promote "sovereign AI" in Southeast Asia
- Liu predicts personal computers will evolve into "AI agent computers" where voice commands replace typing, potentially weakening the Wintel dominance
Why It Matters
Kneron's pitch targets a critical vulnerability in current AI infrastructure: the massive electricity consumption and privacy risks of cloud-based processing . For Hong Kong readers, the involvement of locally-rooted investor Horizon Ventures in thisTaiwanese company suggests the city could become a testing ground for edge AI applications in financial services, where data sovereignty and low-latency processing are paramount .
Kneron's pitch targets a critical vulnerability in current AI infrastructure: the massive electricity consumption and privacy risks of cloud-based processing . For Hong Kong readers, the involvement of locally-rooted investor Horizon Ventures in thisTaiwanese company suggests the city could become a testing ground for edge AI applications in financial services, where data sovereignty and low-latency processing are paramount .