Abhishek Das
| Shri Abhishek Das | |
| |
| Citizenship | Indian |
| Technical disciplines |
|
| Current corporate office | Chief Executive Officer & Managing Director, Accelerate Compute India Pvt. Ltd. |
| Key past appointments | |
| National programs |
|
| Major architectures |
|
| Key international honor | Atos Joseph Fourier Prize (2021) |
Shri Abhishek Das is an Indian computational scientist, high-performance systems architect, and enterprise technology director specializing in supercomputer engineering, artificial intelligence compute fabrics, petascale and exascale infrastructure, and heterogeneous parallel processing platforms. He currently serves as the Chief Executive Officer (CEO) and Managing Director (MD) of Accelerate Compute India Pvt. Ltd.
Over a scientific career encompassing two decades of technical leadership within the Centre for Development of Advanced Computing (C-DAC) and the Digital India Corporation (DIC) under the Ministry of Electronics and Information Technology (MeitY), Das functioned as an architectural lead and project director for India's major national mission-mode computing programs—including the National Supercomputing Mission (NSM), the National Monsoon Mission (NMM), and the IndiaAI Mission.
He is widely recognized for architecting India's highest-ranking artificial intelligence supercomputing platforms on the global TOP500 index, including the 210-petaflops PARAM Siddhi-AI facility (ranked 62nd globally in 2020) and the 410-petaflops unified AIRAWAT platform (ranked 75th globally in 2023). Das also directed the deployment of PARAM Shivay at IIT (BHU) Varanasi, which marked the inaugural supercomputing deployment commissioned under the NSM "Build Approach."
Institutional career and scientific administration
Centre for Development of Advanced Computing (C-DAC)
Das entered C-DAC during the mid-2000s, progressing through successive scientific, architectural, and program directorate appointments:
- Technical Officer & Senior Technical Officer
- Principal Technical Officer
- Joint Director
- Program Director, HPC-AI Infrastructure Development
- Program Director, Exascale Computing Mission Mode Programme
- Officer on Special Duty (OSD) to the Director General
Within C-DAC, his technical remits included micro-architecture profiling, heterogeneous accelerator acceleration, high-speed InfiniBand and Intel Omni-Path interconnect fabric design, high-throughput parallel file storage, and data-center facilities engineering. He coordinated the turnkey staging, benchmarking, and operational handover of supercomputing nodes at central universities, scientific research institutions, and defense laboratories across India.
IndiaAI Mission and Digital India Corporation
Under the IndiaAI Mission executed by MeitY through the Digital India Corporation, Das served as General Manager of the IndiaAI Compute Capacity vertical. He coordinated the technical specification, commercial empanelment, and operational deployment of India's sovereign AI computing backbone, establishing subsidized access to accelerated compute instances for researchers, academic bodies, and artificial intelligence startups.
Enterprise leadership
Following his government scientific agency tenures, Das transitioned into enterprise technology leadership as CEO and Managing Director of Accelerate Compute India Pvt. Ltd., guiding industrial deployments of accelerated GPU infrastructure, custom HPC data center engineering, and enterprise computational fabrics.
National mission mode programs
IndiaAI Mission
- IndiaAI Expert Group Working Report (2023): Co-authored the foundational technical framework for IndiaAI Future Labs Compute, establishing a three-tier national infrastructure hierarchy comprising centralized tier-one supercomputing installations (targeting 10,000 GPUs delivering ~40 exaflops of AI compute), mid-tier regional academic clusters, and low-latency edge computing nodes.
- National Compute Aggregation Fabric: Led operations pooling more than 38,000 accelerated GPUs across 14 empanelled private and public cloud infrastructure partners, integrating processor architectures from NVIDIA, AMD, Intel, AWS, and Google Cloud Platform.
- IndiaAI Compute Portal: Directed the software engineering and deployment of the centralized digital portal allocating containerized GPU compute, multi-terabyte memory instances, and distributed object storage to researchers training Small Language Models (SLMs), Large Language Models (LLMs), and multimodal neural networks.
- 1.1-Exaflops AI Supercomputer (2025): Authored the technical specifications and Request for Proposal (RFP) for the 1.1-exaflops sovereign AI supercomputing installation sited at the National Data Centre in Shastri Park, New Delhi.
- Liquid Immersion Cooling Research: Co-authored research in December 2024 detailing single-phase dielectric coolant immersion systems for cloud AI workloads to reduce Power Usage Effectiveness (PUE) in tropical environments.
- Indigenous Model Evaluation: Served as technical evaluator for native Indian foundational AI initiatives, including BharatGen, Zenteiq, and Fractal.
National Monsoon Mission (NMM)
Operating as Principal Investigator (PI) for the Portable CFS Model with Performance Optimization and Enhanced Usability project at the Indian Institute of Tropical Meteorology (IITM), Pune, Das directed:
- The migration of the Climate Forecast System (CFS) from proprietary IBM Power architectures to scalable commodity x86-64 server environments.
- Algorithmic thread parallelization, cache optimization, and loop vectorization across coupled ocean-atmosphere Earth systems codebases.
- The design of a secure web-based execution interface for operational meteorologists, reducing turnaround times for monsoon prediction models.
National Supercomputing Mission (NSM)
Under the joint DST–MeitY mission, Das directed systems engineering and hardware deployment for multiple petascale and sub-petascale supercomputers:
- PARAM Siddhi-AI (2020): Led the systems architecture, hardware-software integration, and commissioning of India's flagship AI supercomputer at C-DAC, Pune. Generating a peak AI capacity of 210 petaflops and an FP64 Rpeak of 5.267 petaflops, the system ranked 62nd on the global TOP500 list at SC20. It incorporates a 10.5-petabyte high-throughput parallel storage fabric.
- PARAM Shivay (2019): Led the engineering deployment of the 833-teraflops PARAM Shivay system at IIT (BHU) Varanasi, the first supercomputer commissioned under the NSM "Build Approach," inaugurated by Prime Minister Narendra Modi on 19 February 2019.
- Indigenous NSM Deployments:
- PARAM Ganga: 1.3-petaflops hybrid supercomputer installed at IIT Roorkee (March 2022).
- PARAM Ananta: 833-teraflops system commissioned at IIT Gandhinagar (May 2022).
- PARAM Smriti: 833-teraflops compute facility deployed at the National Agri-Food Biotechnology Institute (NABI), Mohali (November 2021).
- AI Compute Deployments: Coordinated the staging and installation of dedicated AI supercomputing nodes at the School of Artificial Intelligence, IIT Delhi (~250 AI petaflops) and the National Informatics Centre (~50 AI petaflops).
- SAMHAR-COVID19 Task Force (2020): Coordinated supercomputing allocations during the pandemic for structural drug repurposing, viral genomic sequencing, and epidemiological spread modeling.
- ARM-based HPC Evaluation: Directed the deployment of a 100-teraflops testbed at C-DAC Pune featuring Fujitsu A64FX processors to benchmark Scalable Vector Extension (SVE) performance against x86 architectures.
AIRAWAT Supercomputing Platform
Das served as a principal architect of the AIRAWAT (AI Research, Analytics, and Knowledge Dissemination Platform) environment at C-DAC. Reaching an initial capacity of 200 AI petaflops and scaling to 410 combined AI petaflops alongside PARAM Siddhi-AI, the installation attained the 75th global position on the TOP500 list at the International Supercomputing Conference (ISC 2023) in Germany.
Academic cluster deployments and desktop supercomputing
Institutional supercomputing facilities
Between 2007 and 2023, Das led engineering teams delivering specialized computing clusters across Indian scientific research centers:
| Year | Facility / System | Host Institution | Architecture / Scale | Research Focus |
|---|---|---|---|---|
| 2007 | HPC Cluster | Jawaharlal Nehru University (JNU), New Delhi | 64-node AMD Opteron cluster | Computational biology, information systems |
| 2007 | CSC Cluster | IIT Delhi | 16-node AMD Opteron cluster | Academic high-performance computing |
| 2008 | Research Cluster | North-Eastern Hill University (NEHU), Shillong | 4-teraflops x86 cluster | Parallel processing research |
| 2008 | HPC System | Physical Research Laboratory (PRL), Ahmedabad | 3-teraflops cluster, 10 TB storage | Planetary science, astrophysics |
| 2008 | PARAM GEM | CSIR-NEIST, Jorhat | Specialized geoscientific cluster | Seismic processing, disaster modeling |
| 2009 | Operational Cluster | NCMRWF, Noida | 32-node compute facility | Operational weather prediction |
| 2010 | CAS Cluster | Centre for Atmospheric Sciences, IIT Delhi | 16-node Intel Xeon cluster | Monsoon dynamics, climate modeling |
| 2010 | Science Cluster | Bharathidasan University, Tamil Nadu | 32-node scientific cluster | Applied mathematics, physical sciences |
| 2013 | Serendipity | NABI, Mohali | Fujitsu enterprise cluster, 20 TB storage | Agricultural biotechnology, genomics |
| 2014 | HPC Facility | NIT Silchar, Assam | 15-teraflops cluster | Parallel computing education |
| 2016 | PARAM Kanchenjunga | NIT Sikkim | 15-teraflops cluster | Mountain sciences, regional engineering |
| 2016 | PARAM Ishan | IIT Guwahati | 250-teraflops hybrid cluster | Weather forecasting, computational chemistry |
| 2020 | PARAM Ambar | National Atmospheric Research Laboratory (NARL), Tirupati | 1.65-petaflops (Intel Xeon / NVIDIA V100) | Atmospheric research, storm-surge modeling |
| 2023 | Secure HPC Node | NTRO, New Delhi | 960 TF HPC / 30 PF AI (NVIDIA A100) | Strategic data analysis |
PARAM Shavak desktop supercomputing series
In 2016, Das formulated the systems architecture for the PARAM Shavak Deep Learning GPU and PARAM Shavak HPC series—compact, self-contained desktop supercomputers delivering up to 25 teraflops of single-precision performance. The platform was designed to decentralize advanced scientific computing across institutions lacking dedicated data center facilities.
Turnkey installations were deployed across national institutions, including NIT Kurukshetra, SVNIT Surat, NIT Meghalaya, NIT Delhi, NIT Agartala, Jadavpur University, Assam University, Tezpur University, VSSUT Burla, L.D. College of Engineering Ahmedabad, Saurashtra University, CET Bhubaneswar, IGIT Sarang, PMEC Berhampur, and the CSIR-National Botanical Research Institute (NBRI) in Lucknow.
Selected publications and conference papers
Das has authored research papers published across proceedings of the IEEE, the ACM, and Springer:
- Das, A. et al. (2023). "Multi-GPU Enabled Quantum Computing on HPC-AI System." Proceedings of the 7th International Conference on Data Management, Analytics and Innovation (ICDMAI 2023), Springer.
- Das, A. et al. (2022). "Performance Analysis of Quantum Algorithms on PARAM Siddhi-AI System." Trends in Quantum Computing and Emerging Business Technologies, IEEE.
- Das, A. et al. (2022). "Strategies to Get Better Performance of TensorFlow on Modern GPU Architecture." 4th International Conference on Advances in Electronics, Computers and Communications (ICAECC), IEEE.
- Das, A. et al. (2021). "Running a Single Instruction Execution Stream to Massively Parallelized Computational Operations." 2nd International Conference on Technology, Engineering, Management for Societal Impact (TEMSMET), IEEE. (Conferred Best Paper Award).
- Das, A. et al. (2021). "Performance Analysis of Python-Based Finite Volume Solver ANUGA on Modern Architectures." 13th International Conference on Contemporary Computing (IC3), ACM.
- Das, A. et al. (2021). "Molecular Dynamics Simulations Accelerate on Elastic Multi-GPU Architecture." Lecture Notes in Networks and Systems (LNNS), Springer.
- Das, A. et al. (2021). "Performance Analysis of Different Distributions of Python and TensorFlow to Efficiently Utilize CPU on HPC Cluster." International Conference on Electrical, Computer and Energy Technologies (ICECET), IEEE.
- Das, A. et al. (2013). "Evaluation of Rodinia Codes on Intel Xeon Phi." International Conference on Intelligent Systems Modelling and Simulation (ISMS), ACM.
- Das, A. et al. (2013). "e-Onama: Mobile High-Performance Computing for Engineering Research." 3rd International Conference on Innovative Computing Technology (INTECH), IEEE.
- Das, A. et al. (2012). "C-DAC Pan India Strategy in High-Performance Computing for Human Advancement." 6th International Conference on Theory and Practice of Electronic Governance (ICEGOV), ACM.
Honors and recognitions
- Atos Joseph Fourier Prize (2021): Conferred for contributions to advanced supercomputing architecture and applied artificial intelligence research.
- Dr. A.P.J. Abdul Kalam Award: Awarded as part of C-DAC's core high-performance computing division for contributions to Indian computational self-reliance.
- Director General's Commendation (Multiple): Conferred by C-DAC for systems engineering leadership and the commissioning of PARAM Shivay under the National Supercomputing Mission.
- Manthan Award (2013): Conferred for the conceptualization and deployment of the Onama / e-Onama mobile computing system.
See also
- Supercomputing in India
- PARAM series
- National Supercomputing Mission
- Centre for Development of Advanced Computing
- TOP500
External links
- Official Corporate Portal: Accelerate Compute India Pvt. Ltd.
- Mission Repository: National Supercomputing Mission (NSM India)