S. Karen Khatamifard
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
S. Karen Khatamifard has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Ulya R. Karpuzcu
University of Minnesota
9 shared publications - Zamshed Iqbal Chowdhury
Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, USA
8 shared publications - Jian-Ping Wang
University of Minnesota
6 shared publications - Sachin S. Sapatnekar
Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, USA
6 shared publications - Masoud Zabihi
Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, USA
5 shared publications - Zhengyang Zhao
Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, USA
5 shared publications
Latest Publications
- 1 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computing
- Vol. 10 (4) , 2055-2071
CRAM-Seq: Accelerating RNA-Seq Abundance Quantification Using Computational RAM
RNA Sequence (RNA-Seq) abundance quantification is an important application in different fields of genomic studies, e.g., analysis of functionally similar genes in a biological sample. This application depends on the availability of high volume of sequence data for high accuracy abundance estimation, which is made possible by next generation sequencing platforms. Large scale data processing requirements of this quantification application push conventional computing systems to their limits due to excessive data movement required between processing and memory elements. Processing-In-memory presents a viable solution to this drawback, through in-situ processing of the genomic data. In this paper, we present CRAM-Seq, an accelerator for RNA-Seq abundance quantification based on Computational RAM (CRAM) an in-memory processing substrate capable of high degree of parallel processing with very low energy consumption. Through hardware/software co-design, we demonstrate that CRAM-Seq outperforms a commonly used state-of-the-art software abundance quantification algorithm, Kallisto in terms of throughput and energy efficiency, while being highly scalable.
- 1 March 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computing
- Vol. 10 (4) , 2055-2071
CRAM-Seq: Accelerating RNA-Seq Abundance Quantification Using Computational RAM
RNA Sequence (RNA-Seq) abundance quantification is an important application in different fields of genomic studies, e.g., analysis of functionally similar genes in a biological sample. This application depends on the availability of high volume of sequence data for high accuracy abundance estimation, which is made possible by next generation sequencing platforms. Large scale data processing requirements of this quantification application push conventional computing systems to their limits due to excessive data movement required between processing and memory elements. Processing-In-memory presents a viable solution to this drawback, through in-situ processing of the genomic data. In this paper, we present CRAM-Seq, an accelerator for RNA-Seq abundance quantification based on Computational RAM (CRAM) an in-memory processing substrate capable of high degree of parallel processing with very low energy consumption. Through hardware/software co-design, we demonstrate that CRAM-Seq outperforms a commonly used state-of-the-art software abundance quantification algorithm, Kallisto in terms of throughput and energy efficiency, while being highly scalable.
- 6 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 19 (6) , 3482-3496
GeNVoM: Read Mapping Near Non-Volatile Memory
DNA sequencing is the physical/biochemical process of identifying the location of the four bases (Adenine, Guanine, Cytosine, Thymine) in a DNA strand. As semiconductor technology revolutionized computing, modern DNA sequencing technology (termed Next Generation Sequenc-ing, NGS) revolutionized genomic research. As a result, modern NGS platforms can sequence hundreds of millions of short DNA fragments in parallel. The sequenced DNA fragments, representing the output of NGS platforms, are termed reads. Besides genomic variations, NGS imperfections induce noise in reads. Mapping each read to (the most similar portion of) a reference genome of the same species, i.e., read mapping, is a common critical first step in a diverse set of emerging bioinformatics applications. Mapping represents a search-heavy memory-intensive similarity matching problem, therefore, can greatly benefit from near-memory processing. Intuition suggests using fast associative search enabled by Ternary Content Addressable Memory (TCAM) by construction. However, the excessive energy consumption and lack of support for similarity matching (under NGS and genomic variation induced noise) renders direct application of TCAM infeasible, irrespective of volatility, where only non-volatile TCAM can accommodate the large memory footprint in an area-efficient way. This paper introduces GeNVoM, a scalable, energy-efficient and high-throughput solution. Instead of optimizing an algorithm developed for general-purpose computers or GPUs, GeNVoM rethinks the algorithm and non-volatile TCAM-based accelerator design together from the ground up. Thereby GeNVoM can improve the throughput by up to 3.67x; the energy consumption, by up to 1.36x, when compared to an ASIC baseline, which represents one of the highest-throughput implementations known.
- 6 October 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 19 (6) , 3482-3496
GeNVoM: Read Mapping Near Non-Volatile Memory
DNA sequencing is the physical/biochemical process of identifying the location of the four bases (Adenine, Guanine, Cytosine, Thymine) in a DNA strand. As semiconductor technology revolutionized computing, modern DNA sequencing technology (termed Next Generation Sequenc-ing, NGS) revolutionized genomic research. As a result, modern NGS platforms can sequence hundreds of millions of short DNA fragments in parallel. The sequenced DNA fragments, representing the output of NGS platforms, are termed reads. Besides genomic variations, NGS imperfections induce noise in reads. Mapping each read to (the most similar portion of) a reference genome of the same species, i.e., read mapping, is a common critical first step in a diverse set of emerging bioinformatics applications. Mapping represents a search-heavy memory-intensive similarity matching problem, therefore, can greatly benefit from near-memory processing. Intuition suggests using fast associative search enabled by Ternary Content Addressable Memory (TCAM) by construction. However, the excessive energy consumption and lack of support for similarity matching (under NGS and genomic variation induced noise) renders direct application of TCAM infeasible, irrespective of volatility, where only non-volatile TCAM can accommodate the large memory footprint in an area-efficient way. This paper introduces GeNVoM, a scalable, energy-efficient and high-throughput solution. Instead of optimizing an algorithm developed for general-purpose computers or GPUs, GeNVoM rethinks the algorithm and non-volatile TCAM-based accelerator design together from the ground up. Thereby GeNVoM can improve the throughput by up to 3.67x; the energy consumption, by up to 1.36x, when compared to an ASIC baseline, which represents one of the highest-throughput implementations known.
- 24 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 19 (2) , 155-158
Voltage Noise Mitigation With Barrier Approximation
Barrier synchronization constructs are placed between phases of parallel programs to ensure correctness in the execution- by preventing threads from proceeding to the subsequent phases of the program before all threads have completed the preceding stage(s). Upon release, threads leaving the barrier at the same time cause sudden change in activity that could potentially lead to voltage emergencies, i.e., timing error, due to electrical properties of power delivery network of the chip. In this paper, we demonstrate barrier approximation through relaxation techniques’ leveraging conservative barrier placement during code development and its impact on voltage noise.
- 24 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 19 (2) , 155-158
Voltage Noise Mitigation With Barrier Approximation
Barrier synchronization constructs are placed between phases of parallel programs to ensure correctness in the execution- by preventing threads from proceeding to the subsequent phases of the program before all threads have completed the preceding stage(s). Upon release, threads leaving the barrier at the same time cause sudden change in activity that could potentially lead to voltage emergencies, i.e., timing error, due to electrical properties of power delivery network of the chip. In this paper, we demonstrate barrier approximation through relaxation techniques’ leveraging conservative barrier placement during code development and its impact on voltage noise.
- Open Access
- 13 April 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Exploratory Solid-State Computational Devices and Circuits
- Vol. 6 (1) , 80-88
A DNA Read Alignment Accelerator Based on Computational RAM
Recent years have witnessed an increasing interest in processing-in-memory paradigm in computing, owing to it’s promise to improve the performance through reduction of energy hungry and long latency memory accesses. Joined with explosion of data to be processed, produced in genomics – particularly genome sequencing, processing-in-memory has become a promising candidate for accelerating genomics applications since they do not scale up well in conventional Von-Neumann systems. In this paper, we present an in-memory accelerator architecture for DNA read alignment. This architecture outperforms corresponding software implementation by >49X and >18000X, in terms of throughput and energy efficiency respectively, even under conservative assumptions.
- 13 April 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Exploratory Solid-State Computational Devices and Circuits
- Vol. 6 (1) , 80-88
Open AccessA DNA Read Alignment Accelerator Based on Computational RAM
Recent years have witnessed an increasing interest in processing-in-memory paradigm in computing, owing to it’s promise to improve the performance through reduction of energy hungry and long latency memory accesses. Joined with explosion of data to be processed, produced in genomics – particularly genome sequencing, processing-in-memory has become a promising candidate for accelerating genomics applications since they do not scale up well in conventional Von-Neumann systems. In this paper, we present an in-memory accelerator architecture for DNA read alignment. This architecture outperforms corresponding software implementation by >49X and >18000X, in terms of throughput and energy efficiency respectively, even under conservative assumptions.
- Open Access
- 4 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Exploratory Solid-State Computational Devices and Circuits
- Vol. 5 (2) , 206-214
Spintronic In-Memory Pattern Matching
Traditional Von Neumann computing is falling apart in the era of exploding data volumes as the overhead of data transfer becomes forbidding. Instead, it is more energyefficient to fuse compute capability with memory where the data reside. This is particularly critical for pattern matching, a key computational step in large-scale data analytics, which involves repetitive search over very large databases residing in memory. Emerging spintronic technologies show remarkable versatility for the tight integration of logic and memory. In this paper, we introduce SpinPM, a novel high-density, reconfigurable spintronic in-memory pattern matching Spin-Orbit-Torque (SOT) – specifically Spin-Hall-Effect (SHE) – substrate based on Computational RAM (CRAM); and demonstrate the performance benefit SpinPM can achieve over conventional and near memory processing systems.
- 4 November 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Exploratory Solid-State Computational Devices and Circuits
- Vol. 5 (2) , 206-214
Open AccessSpintronic In-Memory Pattern Matching
Traditional Von Neumann computing is falling apart in the era of exploding data volumes as the overhead of data transfer becomes forbidding. Instead, it is more energyefficient to fuse compute capability with memory where the data reside. This is particularly critical for pattern matching, a key computational step in large-scale data analytics, which involves repetitive search over very large databases residing in memory. Emerging spintronic technologies show remarkable versatility for the tight integration of logic and memory. In this paper, we introduce SpinPM, a novel high-density, reconfigurable spintronic in-memory pattern matching Spin-Orbit-Torque (SOT) – specifically Spin-Hall-Effect (SHE) – substrate based on Computational RAM (CRAM); and demonstrate the performance benefit SpinPM can achieve over conventional and near memory processing systems.
- Open Access
A Machine Learning Accelerator In-Memory for Energy Harvesting
There is increasing demand to bring machine learning capabilities to low power devices. By integrating the computational power of machine learning with the deployment capabilities of low power devices, a number of new applications become possible. In some applications, such devices will not even have a battery, and must rely solely on energy harvesting techniques. This puts extreme constraints on the hardware, which must be energy efficient and capable of tolerating interruptions due to power outages. Here, as a representative example, we propose an in-memory support vector machine learning accelerator utilizing non-volatile spintronic memory. The combination of processing-in-memory and non-volatility provides a key advantage in that progress is effectively saved after every operation. This enables instant shut down and restart capabilities with minimal overhead. Additionally, the operations are highly energy efficient leading to low power consumption.
Open AccessA Machine Learning Accelerator In-Memory for Energy Harvesting
There is increasing demand to bring machine learning capabilities to low power devices. By integrating the computational power of machine learning with the deployment capabilities of low power devices, a number of new applications become possible. In some applications, such devices will not even have a battery, and must rely solely on energy harvesting techniques. This puts extreme constraints on the hardware, which must be energy efficient and capable of tolerating interruptions due to power outages. Here, as a representative example, we propose an in-memory support vector machine learning accelerator utilizing non-volatile spintronic memory. The combination of processing-in-memory and non-volatility provides a key advantage in that progress is effectively saved after every operation. This enables instant shut down and restart capabilities with minimal overhead. Additionally, the operations are highly energy efficient leading to low power consumption.
- Open Access
Computational RAM to Accelerate String Matching at Scale
Traditional Von Neumann computing is falling apart in the era of exploding data volumes as the overhead of data transfer becomes forbidding. Instead, it is more energy-efficient to fuse compute capability with memory where the data reside. This is particularly critical for pattern matching, a key computational step in large-scale data analytics, which involves repetitive search over very large databases residing in memory. Emerging spintronic technologies show remarkable versatility for the tight integration of logic and memory. In this paper, we introduce CRAM-PM, a novel high-density, reconfigurable spintronic in-memory compute substrate for pattern matching.
Open AccessComputational RAM to Accelerate String Matching at Scale
Traditional Von Neumann computing is falling apart in the era of exploding data volumes as the overhead of data transfer becomes forbidding. Instead, it is more energy-efficient to fuse compute capability with memory where the data reside. This is particularly critical for pattern matching, a key computational step in large-scale data analytics, which involves repetitive search over very large databases residing in memory. Emerging spintronic technologies show remarkable versatility for the tight integration of logic and memory. In this paper, we introduce CRAM-PM, a novel high-density, reconfigurable spintronic in-memory compute substrate for pattern matching.
- 25 July 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 17 (2) , 201-204
A New Class of Covert Channels Exploiting Power Management Vulnerabilities
The energy-crisis of computing leads to even more sophisticated power management. To facilitate dynamic yet accurate power management, hardware activity has to be tracked and processed at a very fine granularity in both space and time. Specialized activity monitors in the form of performance counters or sensors dispersed across chip serve the purpose. The abundance of such activity monitors, their exposure of control to software layers, and the need for tight global control, make power delivery and management systems particularly vulnerable to covert communication over a previously unexplored class of channels, which form the focus of this paper.
- 25 July 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 17 (2) , 201-204
A New Class of Covert Channels Exploiting Power Management Vulnerabilities
The energy-crisis of computing leads to even more sophisticated power management. To facilitate dynamic yet accurate power management, hardware activity has to be tracked and processed at a very fine granularity in both space and time. Specialized activity monitors in the form of performance counters or sensors dispersed across chip serve the purpose. The abundance of such activity monitors, their exposure of control to software layers, and the need for tight global control, make power delivery and management systems particularly vulnerable to covert communication over a previously unexplored class of channels, which form the focus of this paper.
- 12 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 17 (2) , 117-121
On Memory System Design for Stochastic Computing
Growing uncertainty in design parameters (and therefore, in design functionality) renders stochastic computing particularly promising, which represents and processes data as quantized probabilities. However, due to the difference in data representation, integrating conventional memory (designed and optimized for non-stochastic computing) in stochastic computing systems inevitably incurs a significant data conversion overhead. Barely any stochastic computing proposal to-date covers the memory impact. In this paper, as the first study of its kind to the best of our knowledge, we rethink the memory system design for stochastic computing. The result is a seamless stochastic system, StochMem, which features analog memory to trade the energy and area overhead of data conversion for computation accuracy. In this manner StochMem can reduce the energy (area) overhead by up-to 52.8% (93.7%) at the cost of at most 0.7% loss in computation accuracy.
- 12 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Computer Architecture Letters
- Vol. 17 (2) , 117-121
On Memory System Design for Stochastic Computing
Growing uncertainty in design parameters (and therefore, in design functionality) renders stochastic computing particularly promising, which represents and processes data as quantized probabilities. However, due to the difference in data representation, integrating conventional memory (designed and optimized for non-stochastic computing) in stochastic computing systems inevitably incurs a significant data conversion overhead. Barely any stochastic computing proposal to-date covers the memory impact. In this paper, as the first study of its kind to the best of our knowledge, we rethink the memory system design for stochastic computing. The result is a seamless stochastic system, StochMem, which features analog memory to trade the energy and area overhead of data conversion for computation accuracy. In this manner StochMem can reduce the energy (area) overhead by up-to 52.8% (93.7%) at the cost of at most 0.7% loss in computation accuracy.
- 17 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multi-Scale Computing Systems
- Vol. 4 (2) , 141-151
On Approximate Speculative Lock Elision
Each synchronization point represents a point of serialization, and thereby can easily hurt parallel scalability. As demonstrated by recent studies, approximating synchronization by eliminating a subset of synchronization points spatio-temporally can help improve parallel scalability, as long as approximation incurred violations of basic execution semantics remain predictable and controllable. Even if the divergence from fully-synchronized execution renders lower computation accuracy rather than catastrophic program termination, for approximation to be viable, the accuracy loss must be bounded. In this paper, we assess the viability of approximate synchronization using Speculative Lock Elision, which was adopted by hardware transactional memory implementations from industry, as a baseline for comparison.
- 17 November 2017
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multi-Scale Computing Systems
- Vol. 4 (2) , 141-151
On Approximate Speculative Lock Elision
Each synchronization point represents a point of serialization, and thereby can easily hurt parallel scalability. As demonstrated by recent studies, approximating synchronization by eliminating a subset of synchronization points spatio-temporally can help improve parallel scalability, as long as approximation incurred violations of basic execution semantics remain predictable and controllable. Even if the divergence from fully-synchronized execution renders lower computation accuracy rather than catastrophic program termination, for approximation to be viable, the accuracy loss must be bounded. In this paper, we assess the viability of approximate synchronization using Speculative Lock Elision, which was adopted by hardware transactional memory implementations from industry, as a baseline for comparison.