RIDM PTE. LTD.
Category: AI Chips / Semiconductors
Singapore-based AI computing startup developing a proprietary dataflow architecture (DODA) to solve the memory wall bottleneck in AI and HPC, offering a C++-programmable alternative to NVIDIA's CUDA ecosystem. RIDM PTE. LTD. was founded in 2023. The company is led by Jinho Lee. Based in Singapore, Singapore. Team size: 1-10. Latest round: Seed. Key investors include The Invention Lab, Qube Research & Technologies (QRT).
- Founded
- 2023
- Headquarters
- Singapore, Singapore
- Team size
- 1-10
Value proposition
Solves the memory wall bottleneck in AI computing with a novel dataflow architecture that reduces spatial compilation to linear time, supports standard C++ programming as an alternative to CUDA, and delivers ultra-low latency with high energy efficiency.
Products and solutions
DODA (Dynamically Orchestrated Dataflow Architecture) — a proprietary reconfigurable computing architecture for AI and HPC, FPGA acceleration solutions, near-sensor processor for edge devices, run-time programmable dataflow architecture
Unique value
DODA architecture reduces the decades-old spatial compilation problem to linear time, enabling faster development cycles for specialized hardware while supporting standard C++ — offering a competitive alternative to NVIDIA's CUDA ecosystem.
Target customer
Enterprise AI/HPC customers; quantitative investment firms (e.g., QRT); semiconductor/IC design industry; edge computing device manufacturers
Industries served
AI Computing, High-Performance Computing (HPC), Semiconductor/Chip Design, Quantitative Finance, Edge Computing
Technology advantage
DODA architecture; exclusive license to NUS-held patent on general-purpose dataflow architecture; 3-4 international patents filed; run-time programmable dataflow; support for dynamic re-programming and time multiplexing; FPGA acceleration expertise; near-sensor processing capability
How they differentiate
Proprietary DODA (Dynamically Orchestrated Dataflow Architecture) that solves the memory wall bottleneck; reduces spatial compilation to linear time (a decades-hard problem); supports standard C++ programming (not proprietary languages); backed by 3-4 international patents with exclusive license from NUS; run-time programmable and reconfigurable; focus on energy efficiency/green computing
Main competitors
SambaNova Systems (dataflow architecture AI chips), Groq (tensor-streaming AI inference), Cerebras Systems (wafer-scale AI chips)
Key partnerships
Qube Research & Technologies (QRT) — joint PoC on FPGA acceleration, The Invention Lab — strategic mentoring and seed investment, National University of Singapore (NUS) — patent licensing and research collaboration
Notable customers
Qube Research & Technologies (QRT) — running PoC for FPGA acceleration
Major milestones
Founded 2023 by NUS researchers, Developed DODA (Dynamically Orchestrated Dataflow Architecture), Filed 3-4 international patents, Secured exclusive license from NUS for dataflow architecture patent, Completed Seed round with The Invention Lab and QRT (Feb 27, 2026), Awarded TIPS grant (up to S$700K) from South Korea's startup R&D program (Jul 31, 2026), Running PoC with QRT on FPGA acceleration
Growth metrics
Seed stage startup; lean team structure; running PoC with QRT; TIPS grant recipient
Market positioning
Early-stage deep-tech startup positioning as an alternative to NVIDIA's CUDA ecosystem with a novel dataflow architecture targeting the memory wall bottleneck in AI/HPC. Competes with established AI chip companies like SambaNova (dataflow architecture), Groq (tensor-streaming), and Cerebras (wafer-scale), but at an earlier stage with focus on FPGA acceleration and edge devices.
Geographic focus
Singapore (HQ), South Korea (investor base), Global
Patents and IP
3-4 international patents filed; exclusive license to NUS-held patent on general-purpose dataflow architecture (CEO Jinho Lee's doctoral research)
About Jinho Lee
PhD in Computer Science from National University of Singapore (NUS) specializing in general-purpose dataflow architecture; NUS GRIP (Graduate Research Innovation Programme) alumni; Entrepreneur/computer architecture researcher
Latest news about RIDM PTE. LTD.
More AI Chips / Semiconductors companies
Official website: https://www.ridm.tech