Jesmin Jahan Tithi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jesmin Jahan Tithi has a total of 133 co-authors. Here are the top co-authors by number of shared publications:
- Fabrizio Petrini
Intel Labs 95054 Santa Clara CA USA
4 shared publications Fabrizio Petrini
Intel Labs, Santa Clara, CA, USA
3 shared publications- Fabio Checconi
Intel Corporation, Hillsboro, OR, USA
2 shared publications - Fabrizio Petrini
Intel Labs, Santa Clara, CA, USA
2 shared publications - Giovanni Sartor
Alma Mater Studiorum, University of Bologna
2 shared publications - Luca Alessandro Cappellini
Department of Biomedical Science, Humanitas University, Milan, Italy
2 shared publications
Latest Publications
- Open Access
- 14 September 2026
- Published by Springer Nature in AI and Ethics
- Vol. 6 (5) , 512
Co-design of a trustworthy AI-based prognostic tool for predicting patient outcome in acute stroke
This paper reports a descriptive, process-oriented case study of the application of Z-Inspection®, an ethically aligned co-design methodology, in the early design phase of an artificial intelligence (AI) system for healthcare. The methodology was applied within the Horizon Europe project VALIDATE [1], which develops and validates a prognostic clinical decision support system (CDSS) to support treatment decisions in acute ischaemic stroke. An interdisciplinary team of Z-Inspection® experts, AI developers, and clinical stakeholders jointly identified ethical, legal, and technical issues relevant to the planned system. The assessment resulted in 22 ethical issues, 12 dilemmas, 18 risks, and 48 derived requirements, each mapped to the European Commission’s trustworthy AI principles. The identified issues, dilemmas, risks, and derived requirements constitute the primary findings of this descriptive case study and document how general trustworthy-AI principles were translated into context-specific technical, clinical, governance, and organisational requirements. By documenting the process and publishing the complete requirements, the study provides a detailed case for critical scrutiny and potential uptake in other healthcare AI projects, though its transferability and scalability remain to be established in further studies.
- 14 September 2026
- Published by Springer Nature in AI and Ethics
- Vol. 6 (5) , 512
Open AccessCo-design of a trustworthy AI-based prognostic tool for predicting patient outcome in acute stroke
This paper reports a descriptive, process-oriented case study of the application of Z-Inspection®, an ethically aligned co-design methodology, in the early design phase of an artificial intelligence (AI) system for healthcare. The methodology was applied within the Horizon Europe project VALIDATE [1], which develops and validates a prognostic clinical decision support system (CDSS) to support treatment decisions in acute ischaemic stroke. An interdisciplinary team of Z-Inspection® experts, AI developers, and clinical stakeholders jointly identified ethical, legal, and technical issues relevant to the planned system. The assessment resulted in 22 ethical issues, 12 dilemmas, 18 risks, and 48 derived requirements, each mapped to the European Commission’s trustworthy AI principles. The identified issues, dilemmas, risks, and derived requirements constitute the primary findings of this descriptive case study and document how general trustworthy-AI principles were translated into context-specific technical, clinical, governance, and organisational requirements. By documenting the process and publishing the complete requirements, the study provides a detailed case for critical scrutiny and potential uptake in other healthcare AI projects, though its transferability and scalability remain to be established in further studies.
- Open Access
- 5 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 314-327
Closing the Efficiency Gap: AI Datacenter Co-design Roadmap for Scalable Training of LLMs
- 5 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 314-327
Open AccessClosing the Efficiency Gap: AI Datacenter Co-design Roadmap for Scalable Training of LLMs
- 20 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 36 (5) , 1025-1041
Accelerating Sparse Tensor Decomposition Using Adaptive Linearized Representation
High-dimensional sparse data emerge in many critical application domains such as healthcare and cybersecurity. To extract meaningful insights from massive volumes of these multi-dimensional data, scientists employ unsupervised analysis tools based on tensor decomposition (TD) methods. However, real-world sparse tensors exhibit highly irregular shapes and data distributions, which pose significant challenges for making efficient use of modern parallel processors. This study breaks the prevailing assumption that compressing sparse tensors into coarse-grained structures (i.e., tensor slices or blocks) or along a particular dimension/mode (i.e., mode-specific) is more efficient than keeping them in a fine-grained, mode-agnostic form. Our novel sparse tensor representation, Adaptive Linearized Tensor Order (${\sf ALTO}$ ), encodes tensors in a compact format that can be easily streamed from memory and is amenable to both caching and parallel execution. In contrast to existing compressed tensor formats,${\sf ALTO}$ constructs one tensor copy that is agnostic to both the mode orientation and the irregular distribution of nonzero elements. To demonstrate the efficacy of${\sf ALTO}$ , we accelerate popular TD methods that compute the Canonical Polyadic Decomposition (CPD) model across different types of sparse tensors. We propose a set of parallel TD algorithms that exploit the inherent data reuse of tensor computations to substantially reduce synchronization overhead, decrease memory footprint, and improve parallel performance. Additionally, we characterize the major execution bottlenecks of TD methods on multiple generations of the latest Intel Xeon Scalable processors, including Sapphire Rapids CPUs, and introduce dynamic adaptation heuristics to automatically select the best algorithm based on the sparse tensor characteristics. Across a diverse set of real-world data sets,${\sf ALTO}$ outperforms the state-of-the-art approaches, achieving more than an order-of-magnitude speedup over the best mode-agnostic formats. Compared to the best mode-specific formats, which require multiple tensor copies,${\sf ALTO}$ achieves$5.1\times$ geometric mean speedup at a fraction (25% ) of their storage costs. Moreover,${\sf ALTO}$ obtains$8.4\times$ geometric mean speedup over the state-of-the-art memoization approach, which reduces computations by using extra memory, while requiring 14% of its memory consumption.- 20 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 36 (5) , 1025-1041
Accelerating Sparse Tensor Decomposition Using Adaptive Linearized Representation
High-dimensional sparse data emerge in many critical application domains such as healthcare and cybersecurity. To extract meaningful insights from massive volumes of these multi-dimensional data, scientists employ unsupervised analysis tools based on tensor decomposition (TD) methods. However, real-world sparse tensors exhibit highly irregular shapes and data distributions, which pose significant challenges for making efficient use of modern parallel processors. This study breaks the prevailing assumption that compressing sparse tensors into coarse-grained structures (i.e., tensor slices or blocks) or along a particular dimension/mode (i.e., mode-specific) is more efficient than keeping them in a fine-grained, mode-agnostic form. Our novel sparse tensor representation, Adaptive Linearized Tensor Order (${\sf ALTO}$ ), encodes tensors in a compact format that can be easily streamed from memory and is amenable to both caching and parallel execution. In contrast to existing compressed tensor formats,${\sf ALTO}$ constructs one tensor copy that is agnostic to both the mode orientation and the irregular distribution of nonzero elements. To demonstrate the efficacy of${\sf ALTO}$ , we accelerate popular TD methods that compute the Canonical Polyadic Decomposition (CPD) model across different types of sparse tensors. We propose a set of parallel TD algorithms that exploit the inherent data reuse of tensor computations to substantially reduce synchronization overhead, decrease memory footprint, and improve parallel performance. Additionally, we characterize the major execution bottlenecks of TD methods on multiple generations of the latest Intel Xeon Scalable processors, including Sapphire Rapids CPUs, and introduce dynamic adaptation heuristics to automatically select the best algorithm based on the sparse tensor characteristics. Across a diverse set of real-world data sets,${\sf ALTO}$ outperforms the state-of-the-art approaches, achieving more than an order-of-magnitude speedup over the best mode-agnostic formats. Compared to the best mode-specific formats, which require multiple tensor copies,${\sf ALTO}$ achieves$5.1\times$ geometric mean speedup at a fraction (25% ) of their storage costs. Moreover,${\sf ALTO}$ obtains$8.4\times$ geometric mean speedup over the state-of-the-art memoization approach, which reduces computations by using extra memory, while requiring 14% of its memory consumption.- 20 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Micro
- Vol. 43 (5) , 78-87
The Intel Programmable and Integrated Unified Memory Architecture Graph Analytics Processor
High performance large scale graph analytics are essential to timely analyze relationships in big data sets. Conventional processor architectures suffer from inefficient resource usage and bad scaling on those workloads. To enable efficient and scalable graph analysis, Intel ® developed the Programmable Integrated Unified Memory Architecture (PIUMA) as a part of the DARPA Hierarchical Identify Verify Exploit (HIVE) program. PIUMA consists of many multi-threaded cores, fine-grained memory and network accesses, a globally shared address space, powerful offload engines and a tightly integrated optical interconnection network. This paper presents the PIUMA architecture, and documents our experience in designing and building a prototype chip and its bring-up process. PIUMA silicon has successfully powered on demonstrating key aspects of the architecture, some of which will be incorporated into future Intel products.
- 20 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Micro
- Vol. 43 (5) , 78-87
The Intel Programmable and Integrated Unified Memory Architecture Graph Analytics Processor
High performance large scale graph analytics are essential to timely analyze relationships in big data sets. Conventional processor architectures suffer from inefficient resource usage and bad scaling on those workloads. To enable efficient and scalable graph analysis, Intel ® developed the Programmable Integrated Unified Memory Architecture (PIUMA) as a part of the DARPA Hierarchical Identify Verify Exploit (HIVE) program. PIUMA consists of many multi-threaded cores, fine-grained memory and network accesses, a globally shared address space, powerful offload engines and a tightly integrated optical interconnection network. This paper presents the PIUMA architecture, and documents our experience in designing and building a prototype chip and its bring-up process. PIUMA silicon has successfully powered on demonstrating key aspects of the architecture, some of which will be incorporated into future Intel products.
- Open Access
An Optimal Level-synchronous Shared-memory Parallel BFS Algorithm with Optimal parallel Prefix-sum Algorithm and its Implications for Energy Consumption
We present a work-efficient parallel level-synchronous Breadth First Search (BFS) algorithm for shared-memory architectures which achieves the theoretical lower bound on parallel running time. The optimality holds regardless of the shape of the graph. We also demonstrate the implication of this optimality for the energy consumption of the program empirically. The key idea is never to use more processing cores than necessary to complete the work in any computation step efficiently. We keep the rest of the cores idle to save energy and to reduce other resource contentions (e.g., bandwidth, shared caches, etc). Our BFS does not use locks and atomic instructions and is easily extendible to shared-memory coprocessors.
Open AccessAn Optimal Level-synchronous Shared-memory Parallel BFS Algorithm with Optimal parallel Prefix-sum Algorithm and its Implications for Energy Consumption
We present a work-efficient parallel level-synchronous Breadth First Search (BFS) algorithm for shared-memory architectures which achieves the theoretical lower bound on parallel running time. The optimality holds regardless of the shape of the graph. We also demonstrate the implication of this optimality for the energy consumption of the program empirically. The key idea is never to use more processing cores than necessary to complete the work in any computation step efficiently. We keep the rest of the cores idle to save energy and to reduce other resource contentions (e.g., bandwidth, shared caches, etc). Our BFS does not use locks and atomic instructions and is easily extendible to shared-memory coprocessors.
- Open Access
- 29 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Technology and Society
- Vol. 3 (4) , 272-289
Assessing Trustworthy AI in Times of COVID-19: Deep Learning for Predicting a Multiregional Score Conveying the Degree of Lung Compromise in COVID-19 Patients
The paper’s main contributions are twofold: to demonstrate how to apply the general European Union’s High-Level Expert Group’s (EU HLEG) guidelines for trustworthy AI in practice for the domain of healthcare; and to investigate the research question of what does “trustworthy AI” mean at the time of the COVID-19 pandemic. To this end, we present the results of a post-hoc self-assessment to evaluate the trustworthiness of an AI system for predicting a multi-regional score conveying the degree of lung compromise in COVID-19 patients, developed and verified by an interdisciplinary team with members from academia, public hospitals, and industry in time of pandemic. The AI system aims to help radiologists to estimate and communicate the severity of damage in a patient’s lung from Chest X-rays. It has been experimentally deployed in the radiology department of the ASST Spedali Civili clinic in Brescia (Italy) since December 2020 during pandemic time. The methodology we have applied for our post-hoc assessment, called Z-Inspection, uses socio-technical scenarios to identify ethical, technical and domain-specific issues in the use of the AI system in the context of the pandemic.
- 29 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Technology and Society
- Vol. 3 (4) , 272-289
Open AccessAssessing Trustworthy AI in Times of COVID-19: Deep Learning for Predicting a Multiregional Score Conveying the Degree of Lung Compromise in COVID-19 Patients
The paper’s main contributions are twofold: to demonstrate how to apply the general European Union’s High-Level Expert Group’s (EU HLEG) guidelines for trustworthy AI in practice for the domain of healthcare; and to investigate the research question of what does “trustworthy AI” mean at the time of the COVID-19 pandemic. To this end, we present the results of a post-hoc self-assessment to evaluate the trustworthiness of an AI system for predicting a multi-regional score conveying the degree of lung compromise in COVID-19 patients, developed and verified by an interdisciplinary team with members from academia, public hospitals, and industry in time of pandemic. The AI system aims to help radiologists to estimate and communicate the severity of damage in a patient’s lung from Chest X-rays. It has been experimentally deployed in the radiology department of the ASST Spedali Civili clinic in Brescia (Italy) since December 2020 during pandemic time. The methodology we have applied for our post-hoc assessment, called Z-Inspection, uses socio-technical scenarios to identify ethical, technical and domain-specific issues in the use of the AI system in the context of the pandemic.
- Open Access
A New Parallel Algorithm for Sinkhorn Word-Movers Distance and Its Performance on PIUMA and Xeon CPU
The Word Movers Distance (WMD) measures the semantic dissimilarity between two text documents by computing the cost of optimally moving all words of a source/query document to the most similar words of a target document. Computing WMD between two documents is costly because it requires solving an $O(V^3log(V))$ optimization problem where $V$ is the number of unique words in the document. Fortunately, WMD can be framed as an Earth Mover's Distance (EMD) for which the algorithmic complexity can be reduced to $O(V^2)$ by adding an entropy penalty to the optimization problem and solving it using the Sinkhorn-Knopp algorithm. Additionally, the computation can be made highly parallel by adopting a batching approach, i.e., computing the WMD of a single query document against multiple target documents at once. Sinkhorn WMD is a key kernel used in many ML/NLP applications. and usually gets implemented in Python. However, a straightforward Python implementation may leave significant performance on the table even though it may internally call optimized C++ BLAS routines. We present a new sparse {P}arallel {A}lgorithm for {S}inkhorn-Knopp {W}ord-movers {D}istance to compute the semantic distance of one document to many other documents by adopting the $O(V^2)$ EMD algorithm. We algorithmically transform $O(V^2)$ dense compute-heavy EMD version into an equivalent sparse one using new fused SDDMM-SpMM (sparse selection of dense-dense matrix-, sparse-dense matrix-multiplication) kernels. We implemented and optimized this algorithm for two very different architectures -- the new Intel Programmable Integrated Unified Memory Architecture (PIUMA) and Intel Xeon CPUs. We show that we were able to reach close to peak performance on both platforms.Open AccessA New Parallel Algorithm for Sinkhorn Word-Movers Distance and Its Performance on PIUMA and Xeon CPU
The Word Movers Distance (WMD) measures the semantic dissimilarity between two text documents by computing the cost of optimally moving all words of a source/query document to the most similar words of a target document. Computing WMD between two documents is costly because it requires solving an $O(V^3log(V))$ optimization problem where $V$ is the number of unique words in the document. Fortunately, WMD can be framed as an Earth Mover's Distance (EMD) for which the algorithmic complexity can be reduced to $O(V^2)$ by adding an entropy penalty to the optimization problem and solving it using the Sinkhorn-Knopp algorithm. Additionally, the computation can be made highly parallel by adopting a batching approach, i.e., computing the WMD of a single query document against multiple target documents at once. Sinkhorn WMD is a key kernel used in many ML/NLP applications. and usually gets implemented in Python. However, a straightforward Python implementation may leave significant performance on the table even though it may internally call optimized C++ BLAS routines. We present a new sparse {P}arallel {A}lgorithm for {S}inkhorn-Knopp {W}ord-movers {D}istance to compute the semantic distance of one document to many other documents by adopting the $O(V^2)$ EMD algorithm. We algorithmically transform $O(V^2)$ dense compute-heavy EMD version into an equivalent sparse one using new fused SDDMM-SpMM (sparse selection of dense-dense matrix-, sparse-dense matrix-multiplication) kernels. We implemented and optimized this algorithm for two very different architectures -- the new Intel Programmable Integrated Unified Memory Architecture (PIUMA) and Intel Xeon CPUs. We show that we were able to reach close to peak performance on both platforms. - Open Access
MISIM: A Neural Code Semantics Similarity System Using the Context-Aware Semantics Structure
Code semantics similarity can be used for many tasks such as code recommendation, automated software defect correction, and clone detection. Yet, the accuracy of such systems has not yet reached a level of general purpose reliability. To help address this, we present Machine Inferred Code Similarity (MISIM), a neural code semantics similarity system consisting of two core components: (i)MISIM uses a novel context-aware semantics structure, which was purpose-built to lift semantics from code syntax; (ii)MISIM uses an extensible neural code similarity scoring algorithm, which can be used for various neural network architectures with learned parameters. We compare MISIM to four state-of-the-art systems, including two additional hand-customized models, over 328K programs consisting of over 18 million lines of code. Our experiments show that MISIM has 8.08% better accuracy (using MAP@R) compared to the next best performing system.
Open AccessMISIM: A Neural Code Semantics Similarity System Using the Context-Aware Semantics Structure
Code semantics similarity can be used for many tasks such as code recommendation, automated software defect correction, and clone detection. Yet, the accuracy of such systems has not yet reached a level of general purpose reliability. To help address this, we present Machine Inferred Code Similarity (MISIM), a neural code semantics similarity system consisting of two core components: (i)MISIM uses a novel context-aware semantics structure, which was purpose-built to lift semantics from code syntax; (ii)MISIM uses an extensible neural code similarity scoring algorithm, which can be used for various neural network architectures with learned parameters. We compare MISIM to four state-of-the-art systems, including two additional hand-customized models, over 328K programs consisting of over 18 million lines of code. Our experiments show that MISIM has 8.08% better accuracy (using MAP@R) compared to the next best performing system.