SEOUL, South Korea and SANTA CLARA, Calif. — September 11, 2025 — Leads & Copy — Moreh unveiled its distributed inference system on AMD and showcased collaborations with Tenstorrent and SGLang at the AI Infra Summit 2025 in Santa Clara, held September 9–11.
The AI Infra Summit, a large AI conference dedicated to the infrastructure layer of AI & Machine Learning, attracted 3,500 attendees and over 100 partners.
Moreh CEO Gangwon Jo introduced the company’s distributed inference system and benchmark results, demonstrating optimization of deep learning models more efficiently than NVIDIA. A next-generation AI semiconductor system combining Moreh’s software with Tenstorrent’s hardware was also unveiled, offering cost-competitive alternatives to NVIDIA.
Moreh co-hosted a presentation with SGLang and organized a booth and networking sessions to strengthen collaboration with the global AI ecosystem. Moreh plans to jointly develop an AMD-based distributed inference system with SGLang.
Moreh CEO Gangwon Jo stated that Moreh is conducting PoC projects with leading LLM companies and aims to establish itself as a global company through collaboration with AMD, Tenstorrent, and SGLang.
Moreh is developing its own core AI infrastructure engine and securing technological capabilities through its foundation LLM subsidiary Motif Technologies.
Contact:
Gangwon Jo, CEO
Source: Moreh
