AI People Wiki

Builders & open source

Hani Jamjoom

Systems security researcher listed with IBM Research; co-author of confidential-computing studies for AI

Hani Jamjoom is listed with IBM Research as a co-author of papers on computing systems and their security. These include "Blueprint, Bootstrap, and Bridge" (2025), a security study of NVIDIA GPU Confidential Computing, which aims to provide secure execution for AI workloads, and "Confidential Inference via Ternary Model Partitioning" (first submitted in 2018), which uses trusted execution environments to limit information leakage from deep learning inference inputs. Jamjoom is also listed with the IBM T.J. Watson Research Center as a co-author of a 2026 reference architecture for quantum-centric supercomputers that combine quantum processors with GPUs and CPUs.123

At a glance

Current role
None confirmed in the sources
Work listed
2
Sources cited
7
Profile reviewed

Research

Several of Jamjoom's co-authored papers concern confidential computing for machine learning. With Zhongshu Gu and colleagues, Jamjoom co-wrote a 2018 paper on confidential inference through ternary model partitioning and a 2018 design called CalTrain for keeping training data confidential while preserving model accountability in collaborative learning. A 2021 paper on separation of powers in federated learning addressed information leakage from an untrusted aggregation server.245

Jamjoom has also co-authored analyses of hardware trusted execution: a 2023 top-down study of Intel Trust Domain Extensions (TDX) and the 2025 security study of NVIDIA GPU Confidential Computing. Other recent co-authored systems work includes Rex, a 2025 design for safe and usable kernel extensions, and a 2026 reference architecture for quantum-centric supercomputers.6173

Work and contributions

  • Security study of NVIDIA GPU Confidential Computing Published research

    Co-author of "Blueprint, Bootstrap, and Bridge: A Security Look at NVIDIA GPU Confidential Computing" (2025), which reconstructs the architecture and bootstrap process of GPU confidential computing and tests whether data transfers between trusted CPU and GPU domains stay protected.1

  • Confidential Inference via Ternary Model Partitioning Published research

    Co-author of a 2018 paper that studies how inference inputs can leak in deep learning image classification pipelines and proposes a ternary model partitioning mechanism that uses trusted execution environments.2

Timeline

  1. Co-authored "Confidential Inference via Ternary Model Partitioning."2
  2. Co-authored CalTrain, a design for data confidentiality and model accountability in collaborative learning.4
  3. Co-authored "Separation of Powers in Federated Learning."5
  4. Co-authored "Intel TDX Demystified," a top-down study of Intel Trust Domain Extensions.6
  5. Co-authored Rex, a system for safe and usable kernel extensions.7
  6. Co-authored a security study of NVIDIA GPU Confidential Computing.1
  7. Co-authored a reference architecture for quantum-centric supercomputers.3

Documented roles

The sources do not confirm a current role; these are the roles they document.

  • Listed with IBM Research on the 2025 GPU confidential computing paper and with the IBM T.J. Watson Research Center on a 2026 arXiv paper, IBM Research13

Companies in their history

  • IBM 29 people in the wiki

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

7 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. Blueprint, Bootstrap, and Bridge: A Security Look at NVIDIA GPU Confidential Computing (arXiv 2507.02770)arXiv · Paper · published Jul 3, 2025 · accessed Oct 21, 2026
  2. Confidential Inference via Ternary Model Partitioning (arXiv 1807.00969)arXiv · Paper · published Jul 3, 2018 · accessed Oct 21, 2026
  3. Reference Architecture of a Quantum-Centric Supercomputer (arXiv 2603.10970)arXiv · Paper · published Mar 11, 2026 · accessed Oct 21, 2026
  4. Reaching Data Confidentiality and Model Accountability on the CalTrainarXiv · Paper · published Dec 7, 2018 · accessed Oct 21, 2026
  5. Separation of Powers in Federated LearningarXiv · Paper · published May 19, 2021 · accessed Oct 21, 2026
  6. Intel TDX Demystified: A Top-Down ApproacharXiv · Paper · published Mar 27, 2023 · accessed Oct 21, 2026
  7. Safe and usable kernel extensions with RexarXiv · Paper · published Feb 26, 2025 · accessed Oct 21, 2026