V.B. Varma
Dr Vijaykumar Varma

Materials science · Microfluidics · Thermal engineering

Dr Vijaykumar Varma

Senior R&D scientist and team lead in thermal engineering, simulation and advanced materials, hands-on from model to hardware.

Cooling the next decade of AI chips, from coolant physics to package-level thermal architecture.

R&D leader who builds teams, labs and programmes where thermal engineering, microfluidics and AI-driven materials discovery meet.

Technical co-founder and CTO candidate for deep-tech: I have built labs, devices and data pipelines from nothing, and taken them to industry partners.

16 years of research since 2010, across India, Singapore, Luxembourg and Belgium. I led a five-person team for five years at NTU Singapore, set up two labs from scratch (programmes of ~$1.01M and S$970k), and wrote €3M+ in funding proposals as PI at LIST Luxembourg. Today I deliver high-throughput liquid-cooling simulation for imec’s AI-chip roadmap.

Thermal and microfluidics R&D leader with 16 years of research: five years leading pump-free magnetofluidic cooling up to kilowatt scale, and now high-throughput liquid-cooling simulation of 2.5D/3D GPU and HBM packages for imec’s 2028–2035 AI-chip roadmap.

I still run the simulations and build the hardware myself, and I have led the team, budget and partnerships around it. Five years leading a five-person team of scientists, 29 devices delivered, 5+ industry partners, and now 3,000+ AI-chip cooling simulations for imec in under a month.

I build autonomous, AI-guided experimentation platforms and the groups that run them. Five years leading a research team at NTU Singapore, PI at LIST Luxembourg, two labs set up from scratch, 35 publications, and mentees who are now an assistant professor, a principal engineer and a staff engineer.

Thermal and microfluidics expert for AI hardware. I run automated CFD/CHT programmes for next-generation liquid-cooled GPU and HBM stacks at imec, and I have designed, built and validated 18+ cooling devices that beat conventional designs by up to 300%.

16 years turning physics into hardware: two labs set up from zero (~$1.01M and S$970k programmes), a provisional patent, 5+ industry partners, €3M+ in non-dilutive proposals written as PI, a LIST commercialisation project, and an advisory role at M3TRL Consulting.

  • AI-chip & electronics cooling
  • Liquid cooling & thermal packaging
  • CFD, CHT & porous-media models
  • Microfluidics & lab-on-chip
  • Multiphysics digital twins
  • AI-driven materials discovery
  • EU & FNR proposals

Open to senior R&D, expert and leadership roles. Based between Luxembourg and Leuven; Luxembourg preferred, BeNeLux, Germany and the EU considered.

Open to senior R&D, engineering, expert and leadership roles, and to early-stage deep-tech teams. Luxembourg preferred; Belgium, the Netherlands, Germany and the EU considered.

Open to thermal architect, cooling technology lead and R&D manager or director roles in semiconductors, AI hardware and data-center cooling. Luxembourg, BeNeLux, Germany and the EU first; international roles considered.

Available for US roles with O-1 sponsorship and ready to relocate. CV and evidence portfolio on request.

Open to CTO, technical co-founder and head-of-R&D roles with deep-tech startups in the EU and US.

Illustrative self-driving lab: a Bayesian optimiser picks each next experiment across an alloy-composition space and homes in on the best material in a few dozen runs instead of hundreds.
Illustrative package cross-section: GPU and HBM stacks on an interposer, cooled by a micropillar cold plate as chip power climbs. The problem my simulation and cooling work addresses.
Illustrative temperature field: coolant crossing a micropillar cold plate above two chip hotspots, the kind of design space my simulation programmes sweep thousands of times.

Track record at a glance

    From pump-free cooling to AI chips

    Few people have done the coolant physics, built the hardware and simulated the package. I have done all three, in that order.

      Where I work in the cooling stack

      Four layers, from the coolant to the facility, each with evidence from my own work.

      Where chip cooling is heading

      Six directions for the next generation of AI hardware, marked by where I have hands-on results and where I would lead next.

      Leadership track record

      What I have built, led, won and grown, with the numbers behind it.

      For recruiters and search committees

      For US recruiters and hiring managers

      For semiconductor and AI-hardware recruiters

      For founders and investors

      The short version, so you can decide in a minute whether to call.

      Roles
      Senior or expert R&D scientist or engineer, head of R&D, R&D group or team lead, and CTO for deep-tech startups. In academia and research institutes: group leader, associate or assistant professor.
      Thermal architect, packaging thermal lead, cooling technology lead, R&D manager or director (thermal) in semiconductors, AI hardware and data-center cooling.
      R&D engineer or scientist, CFD and thermal simulation engineer, thermal architect or packaging thermal lead, materials scientist, process or development engineer, innovation or R&D project manager, technical expert, team or group lead and head of R&D.
      Group leader, associate or assistant professor, or head of an autonomous-experimentation, thermal-fluids or microfluidics lab.
      Staff, principal or lead thermal engineer and R&D manager in AI hardware, semiconductors, data-center cooling and advanced materials.
      CTO, technical co-founder or VP/head of R&D at deep-tech startups in thermal management, microfluidics, advanced materials or lab automation.
      Where
      Luxembourg first; Belgium, the Netherlands, Germany and the wider EU. On-site or hybrid.
      United States, relocating with O-1 sponsorship.
      EU or US; on-site, hybrid or remote with regular travel.
      Leadership scope
      Five-person research team led for five years; spend managed on two lab programmes (~$1.01M and S$970k); interns selected and 7 researchers trained; PI on €3M+ in proposals.
      Luxembourg ties
      Two years at LIST and the University of Luxembourg (2023–2025), Luxembourg phone and network, and advisor to M3TRL Consulting, Luxembourg.
      Availability
      Experience
      16 years of research since 2010, 9 of them after the PhD, in India, Singapore, Luxembourg and Belgium.
      Work status
      EU Blue Card holder, based in Leuven, Belgium.
      EU Blue Card holder in Belgium; requires US sponsorship (O-1).
      Languages
      English (professional), Hindi (native).

      Where my skills transfer

      Sectors where my thermal, simulation, materials and lab-automation experience applies from day one.

      What I’d build next

      My plan for the first two years in a leadership role, in industry or in academia.

      My plan for the first two years leading a chip-cooling or thermal team.

      My plan for the first two years leading an R&D team or group.

      My plan for the first two years as a group leader or faculty member.

      Problems I would build a company around, and what I bring on day one.

      Programmes I have led and delivered

      Each with the mandate, the team and budget, the approach and the impact.

      Details of partner work at imec are confidential; only public information is shown here.

      Perspectives

      Short pieces on where chip cooling is going, published on LinkedIn.

        How I lead

          What supervisors and team members say

          Full reference letters available on request.

          Recognition and impact

          Independent evidence of standing in the field. Documentation available on request.

          Technical depth

          Six areas where I can lead a programme end to end, from first model to validated hardware.

          Tools and methods

          Ansys Fluent, COMSOL, Python (PyAnsys, PyFluent), HPC/SLURM, Bayesian optimisation and active learning, LabVIEW, MATLAB, microfabrication (PDMS, soft lithography, photolithography), 3D printing, SEM/TEM/XRD, DSC/TGA, rheometry, DoE, FMEA, SPC, DMAIC.

          Experience

            Education

            • PhD, Materials Science & Engineering, Nanyang Technological University, Singapore, 2017. CGPA 4.3/5; S$184k research scholarship. Thesis: magnetic structures by micro-magnetofluidic techniques.
            • MSc, Physics (materials science), University of Pune, India, 2010. Thesis: noise analysis of field-emission current in graphene.
            • BSc, Physics, Chemistry, Mathematics, B. P. College, SRTM University, India, 2008. GPA 9/10.

            Training and certifications

            • MIT Applied AI & Data Science Program
            • AI Foundations: Machine Learning (2021)
            • Six Sigma Yellow Belt (2021)
            • GMP awareness; decision-making and collaboration principles
            • NTU University Teaching for Teaching Assistants, 8-module course with assessment (2013)

            Teaching

            University teaching at NTU: tutorial and laboratory sessions for undergraduate courses in metallic and ceramic materials and semiconductors, and PhD-level lab sessions I designed in microfluidics, microfabrication and thermal management. Scores are from official NTU student surveys (5 = strongly agree).

              In students’ words

              From anonymous NTU feedback forms. Full survey reports available on request. Lectures and tutorials are also on my YouTube channel.

              Publications

              Full profile on Google Scholar.

                In preparation

                  Talks and media

                  36 conference presentations: 2 keynotes, 14 invited talks, session chairs and contributed talks and posters.

                  Videos: feature, talks, demos and training

                  YearTypeTalk

                  Advisory

                  Technical advisor to , in a personal capacity. For project work, contact the consultancy; for roles, research or speaking, contact me directly.

                  • Thermal design review for cold plates, immersion and microchannel cooling, with CFD/CHT validation plans.
                  • Simulation automation: turning one-off Fluent or COMSOL models into scripted parameter sweeps on HPC.
                  • Self-driving lab feasibility: what to automate first, which hardware, and how to close the loop with Bayesian optimisation.

                  Let’s talk.

                  Recruiters, search firms, founders and hiring committees: email is fastest. I usually reply within one working day.