Tsuyoshi Hamada

• Comment: The title of this draft either has been disambiguated or will require disambiguation if accepted.
  If this draft has been disambiguated (renamed), submitters and reviewers are asked to consider whether the current title is the best possible disambiguation, and, if necessary, move (rename) this draft.
  If this draft is accepted, a hatnote will need to be added to the primary page to refer to this page. If there is already a hatnote on the primary page, please review whether a disambiguation page is in order instead. Please do not edit the hatnote on the primary page or insert a hatnote onto the primary page unless you are accepting this draft.
  The primary page that the hatnote should be added to is Tsuyoshi Hamada.
  You may ask for advice about hatnotes at the Teahouse. Robert McClenon (talk) 16:35, 3 August 2026 (UTC)

---

This article is about the computer scientist. For the retired boxer, see Tsuyoshi Hamada.

Tsuyoshi Hamada
Native name	濱田 剛
Born
💼 Occupation	Engineer

Tsuyoshi Hamada (濱田 剛, Hamada Tsuyoshi) is a Japanese computer scientist and entrepreneur known for GPGPU research and the DEGIMA supercomputer. He is a former associate professor at Nagasaki University, co-founder of Tier IV and ArkEdge Space, and CEO of DEGIMA AI.

Hamada received the Gordon Bell Prize in the price/performance category from the Association for Computing Machinery in 2009 and again in 2010 (honorable mention in the same category), for work on low-cost GPU clusters. Japanese newspapers reported that the DEGIMA system reached performance comparable to Japan's then-leading supercomputers at a fraction of typical construction cost.

Career

Hamada received a Doctor of Engineering from the University of Tokyo in 1999. He joined RIKEN in 2006 as a researcher on massively parallel systems using graphics processing units (GPUs). From 2008 until about 2017 he was associate professor and head of the Massively Parallel Computing Division at the Advanced Computing Research Center of Nagasaki University.

In December 2015 he co-founded Tier IV, a company working on autonomous driving software. In January 2018 he co-founded ArkEdge Space, a company developing small satellites and related systems. In April 2025 he became CEO of DEGIMA AI, Inc., a U.S.-based firm working on high-performance computing and AI services.

Research

Hamada's research focuses on computer science, especially GPGPU and large-scale N-body simulations. In the IPA Exploratory IT Human Resources Project (Mitou) in 2005, he developed PROGRAPE-4, an FPGA-based system that reached 243.2 GFLOPS on collisionless gravitational N-body problems while drawing about 5 W of power; IPA named him a "Genius Programmer / Super Creator" in 2006.

In 2007 Hamada and Toshiaki Iitaka posted "The Chamomile Scheme", an optimized N-body algorithm for programmable GPUs, to arXiv; it used NVIDIA's newly released CUDA platform on a GeForce 8800 GTX. The preprint is widely cited as one of the earliest scientific computing papers using CUDA. At public talks in 2008 he compared CPU and GPU performance on galaxy-collision demos, reporting on the order of 0.2 GFLOPS on CPU versus 270–470 GFLOPS on GPU.

With Rio Yokota (then at the University of Bristol) and Keigo Nitadori of RIKEN, Hamada built DEGIMA, connecting 760 GPUs in parallel for about ¥38 million. The team used a "multiple-walk" method to coordinate the GPUs and reported about 158 TFLOPS peak performance, which Japanese media described as exceeding the Earth Simulator for the workloads tested. An upgraded configuration with InfiniBand and additional GPUs reached about 190 TFLOPS with a cost increase of less than 20%, and received a second Gordon Bell price/performance award in 2010.

On the June 2011 TOP500 list, DEGIMA ranked 429th (42.83 TFLOPS measured). On the Green500 list for the same period it placed third in energy efficiency (1375.88 MFLOPS/W), behind two Blue Gene/Q prototype entries.

Hamada studied under astrophysicist Junichiro Makino, who noted Hamada's Gordon Bell awards in 2009.

In July 2026, at a Japanese government event on physical AI policy, NVIDIA CEO Jensen Huang cited Hamada and Nagasaki University's GPU cluster work—including early systems cooled with household fans for quantum-chemistry calculations—as an example of Japan's prior leadership in GPU supercomputing.

Views

In a 2008 interview, Hamada said the design ideas behind GPGPU and the University of Tokyo's GRAPE special-purpose computers were similar, and that experience from physics simulation in Japan could transfer to GPU computing. After the DEGIMA awards, he told Nishinippon Shimbun that he was skeptical of supercomputer projects that required very large budgets without pursuing cost reduction.

Awards

• May 2006 – IPA Genius Programmer / Super Creator
• November 2009 – Gordon Bell Prize, Low Price / Performance Category
• November 2010 – Gordon Bell Prize, Honorable Mention (Price / Performance)
• June 2011 – PRACE Award

See also

• DEGIMA
• Gordon Bell Prize
• GPGPU

External links

• ACM award profile
• Nagasaki University Advanced Computing Research Center (archived department page)

Category:Japanese computer scientists Category:Living people Category:GPGPU

This article "Tsuyoshi Hamada (computer scientist)" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Tsuyoshi Hamada (computer scientist). Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.

---