NVIDIA AI Architect

Mocha, whose full name is Mochammad Asri, has built a career at the center of global AI development while keeping Indonesia in view. At NVIDIA, he works as an AI Architect focused on chip systems and architectures for generative AI, a role that places him in one of the most strategic areas of modern computing.His path is notable not only because of the companies he has worked for, but also because of the technical depth behind his specialization. Mocha’s work sits at the intersection of efficiency, memory systems, and large-scale AI infrastructure, the kind of expertise that has become increasingly valuable as the industry races to build faster and smarter systems.A technical career built on strong academic foundationsMocha’s journey began far from Silicon Valley. He graduated from SMA Negeri 8 Jakarta before continuing his studies overseas and eventually earning a doctorate in the United States.He received a Ph.D. in Computer Engineering from The University of Texas at Austin, with research focused on data movement optimization in heterogeneous systems. Before that, he completed a Master of Engineering and a Bachelor of Science at Tokyo Institute of Technology in Computer Science.His academic background was further strengthened by edX certifications in Deploying TinyML, Fundamentals of TinyML, and Applications of TinyML. That mix of formal education and specialized training helped shape a career that now spans both research and industrial architecture work.Why his NVIDIA role mattersMocha’s specialization goes well beyond chip design alone. His expertise includes Generative AI, Efficient AI, Large Language Models, heterogeneous systems, HW/SW co-design, and System-on-Chip architecture.He also works in areas such as bandwidth versus latency shaping, QoS-aware memory controllers, caching techniques, Network-on-Chip design, parallelism, and advanced memory and computing optimization. These are core building blocks for generative AI systems, automotive platforms, and server chips.At NVIDIA, his responsibilities include end-to-end memory system development for generative AI, automotive applications, and server chip platforms. He also works on optimizing bandwidth, latency, and intelligent QoS systems for modern architectures.Focus AreaExamplesWhy It MattersAI ArchitectureGenerative AI, LLMs, SoCSupports modern AI compute platformsMemory and Data MovementCache, NoC, QoS-aware controllersImproves efficiency and system performanceSystem Co-DesignHW/SW co-design, heterogeneous systemsAligns hardware and software for advanced workloadsExperience that sharpened his global profileBefore joining NVIDIA, Mocha worked at Meta as a Chip Architect from July 2020 to September 2023 in Menlo Park, California. There, he contributed to computing architecture development for AR, VR, and AI.His work at Meta centered on data movement systems such as cache and Network-on-Chip, experience that helped strengthen his reputation as an Indonesian engineer capable of operating at a global level. He then moved to NVIDIA in September 2023 and now serves as Lead Architect.Why his return to Indonesia mattersDespite his international career, Mocha has repeatedly expressed a desire to return to Indonesia and contribute to education. He sees strong potential in the country’s young talent, but also recognizes the need for a stronger ecosystem and wider access to quality learning.He has pointed to technology fields such as AI, semiconductors, and computer systems as areas where Indonesia still needs more support and opportunity. That perspective gives his story a broader meaning: a globally successful engineer who still looks at home as the place where long-term impact could matter most.For many young Indonesians, Mocha stands as proof that a path from local schools to the heart of global AI is possible. His career combines academic discipline, high-level industry experience, and a consistent concern for how technology education can be strengthened back home. Source: id.mashable.com

From NVIDIA To Indonesia, Mocha’s Return Plan Puts Education First