arXiv subjects

Aman Arora

Publications and source records attributed to Aman Arora.

35 records · Page 2Linked to original sources

arXiv2025

GAMA: High-Performance GEMM Acceleration on AMD Versal ML-Optimized AI Engines

General matrix-matrix multiplication (GEMM) is a fundamental operation in machine learning (ML) applications. We present the first comprehensive performance acceleration of GEMM workloads on AMD's second-generation AIE-ML (AIE2) architecture, which is specifically optimized for ML applications. Compared to AI-Engine (AIE1), AIE offers increased compute throughput and larger on-chip memory capacity. We propose a novel design that maximizes AIE memory utilization, incorporates custom buffer placement within the AIE2 and staggered kernel placement across the AIE2 array, significantly reducing performance bottlenecks such as memory stalls and routing congestion, resulting in improved performance and efficiency compared to the default compiler provided by AMD. We evaluate the performance benefits of our design at three levels: single AIE, pack of AIEs and the complete AIE array. GAMA achieves state-of-the-art performance, delivering up to 165 TOPS (85% of peak) for int8 precision and 83 TBFLOPS (86% of peak) for bfloat16 precision GEMM workloads. Our solution achieves 8.7%, 9%, 39% and 53.6% higher peak throughput efficiency compared to the state-of-the-art AIE1 frameworks AMA, MAXEVA, ARIES and CHARM, respectively.

cs.AR↗

arXiv2025

CarbonSet: A Dataset to Analyze Trends and Benchmark the Sustainability of CPUs and GPUs

Over the years, the chip industry has consistently developed high-performance processors to address the increasing demands across diverse applications. However, the rapid expansion of chip production has significantly increased carbon emissions, raising critical concerns about environmental sustainability. While researchers have previously modeled the carbon footprint (CFP) at both system and processor levels, a holistic analysis of sustainability trends encompassing the entire chip lifecycle remains lacking. This paper presents CarbonSet, a comprehensive dataset integrating sustainability and performance metrics for CPUs and GPUs over the past decade. CarbonSet aims to benchmark and assess the design of next-generation processors. Leveraging this dataset, we conducted detailed analysis of flagship processors' sustainability trends over the last decade. This paper further highlights that modern processors are not yet sustainably designed, with total carbon emissions increasing more than 50$\times$ in the past three years due to the surging demand driven by the AI boom. Power efficiency remains a significant concern, while advanced process nodes pose new challenges requiring to effectively amortize the dramatically increased manufacturing carbon emissions.

cs.AR↗

arXiv2025

SAF: Scalable Acceleration Framework for dynamic and flexible scaling of FPGAs

FPGAs are increasingly gaining traction in cloud and edge computing environments due to their hardware flexibility, low latency, and low energy consumption. However, the existing hardware stack of FPGA and the host-FPGA connectivity does not allow flexible scaling and simultaneous reconfiguration of multiple devices, which limits the adoption of FPGA at scale. In this paper, we present SAF -- an Ethernet-based scalable acceleration framework that allows FPGA to be hot-plugged into a network in a stand-alone fashion without connecting to a local host CPU, which enables flexible scalability. SAF provides a custom FPGA shell and a set of Ethernet protocols that allow FPGAs to connect with a remote host to accelerate application kernels. SAF can configure multiple FPGAs simultaneously, which significantly reduces the reconfiguration time in scaling effort. We implemented the SAF framework using Intel FPGA SDK for OpenCL and 20 Bittware 385A cards with Arria-10 FPGAs. We analyze a case study and conduct experiments to compare SAF with state-of-the-art multi-FPGA clusters. Results show that SAF provides 13X faster reconfiguration than sequential PCIe programming, reduces the hardware setup costs by 38%, application runtime by 25%, and energy consumption by 27%. We evaluated the performance scalability of SAF using the PTRANS benchmark of the HPCC FPGA benchmark suite and showed an almost linear speedup for strong and weak scaling scenarios.

cs.DC↗

arXiv2025

Systolic Sparse Tensor Slices: FPGA Building Blocks for Sparse and Dense AI Acceleration

FPGA architectures have recently been enhanced to meet the substantial computational demands of modern deep neural networks (DNNs). To this end, both FPGA vendors and academic researchers have proposed in-fabric blocks that perform efficient tensor computations. However, these blocks are primarily optimized for dense computation, while most DNNs exhibit sparsity. To address this limitation, we propose incorporating structured sparsity support into FPGA architectures. We architect 2D systolic in-fabric blocks, named systolic sparse tensor (SST) slices, that support multiple degrees of sparsity to efficiently accelerate a wide variety of DNNs. SSTs support dense operation, 2:4 (50%) and 1:4 (75%) sparsity, as well as a new 1:3 (66.7%) sparsity level to further increase flexibility. When demonstrating on general matrix multiplication (GEMM) accelerators, which are the heart of most current DNN accelerators, our sparse SST-based designs attain up to 5x higher FPGA frequency and 10.9x lower area, compared to traditional FPGAs. Moreover, evaluation of the proposed SSTs on state-of-the-art sparse ViT and CNN models exhibits up to 3.52x speedup with minimal area increase of up to 13.3%, compared to dense in-fabric acceleration.

cs.AR↗

arXiv2024

Iterative NLP Query Refinement for Enhancing Domain-Specific Information Retrieval: A Case Study in Career Services

Retrieving semantically relevant documents in niche domains poses significant challenges for traditional TF-IDF-based systems, often resulting in low similarity scores and suboptimal retrieval performance. This paper addresses these challenges by introducing an iterative and semi-automated query refinement methodology tailored to Humber College's career services webpages. Initially, generic queries related to interview preparation yield low top-document similarities (approximately 0.2--0.3). To enhance retrieval effectiveness, we implement a two-fold approach: first, domain-aware query refinement by incorporating specialized terms such as resources-online-learning, student-online-services, and career-advising; second, the integration of structured educational descriptors like "online resume and interview improvement tools." Additionally, we automate the extraction of domain-specific keywords from top-ranked documents to suggest relevant terms for query expansion. Through experiments conducted on five baseline queries, our semi-automated iterative refinement process elevates the average top similarity score from approximately 0.18 to 0.42, marking a substantial improvement in retrieval performance. The implementation details, including reproducible code and experimental setups, are made available in our GitHub repositories \url{https://github.com/Elipei88/HumberChatbotBackend} and \url{https://github.com/Nisarg851/HumberChatbot}. We also discuss the limitations of our approach and propose future directions, including the integration of advanced neural retrieval models.

cs.IR↗

arXiv2024

HLSFactory: A Framework Empowering High-Level Synthesis Datasets for Machine Learning and Beyond

Machine learning (ML) techniques have been applied to high-level synthesis (HLS) flows for quality-of-result (QoR) prediction and design space exploration (DSE). Nevertheless, the scarcity of accessible high-quality HLS datasets and the complexity of building such datasets present challenges. Existing datasets have limitations in terms of benchmark coverage, design space enumeration, vendor extensibility, or lack of reproducible and extensible software for dataset construction. Many works also lack user-friendly ways to add more designs, limiting wider adoption of such datasets. In response to these challenges, we introduce HLSFactory, a comprehensive framework designed to facilitate the curation and generation of high-quality HLS design datasets. HLSFactory has three main stages: 1) a design space expansion stage to elaborate single HLS designs into large design spaces using various optimization directives across multiple vendor tools, 2) a design synthesis stage to execute HLS and FPGA tool flows concurrently across designs, and 3) a data aggregation stage for extracting standardized data into packaged datasets for ML usage. This tripartite architecture ensures broad design space coverage via design space expansion and supports multiple vendor tools. Users can contribute to each stage with their own HLS designs and synthesis results and extend the framework itself with custom frontends and tool flows. We also include an initial set of built-in designs from common HLS benchmarks curated open-source HLS designs. We showcase the versatility and multi-functionality of our framework through seven case studies: I) ML model for QoR prediction; II) Design space sampling; III) Fine-grained parallelism backend speedup; IV) Targeting Intel's HLS flow; V) Adding new auxiliary designs; VI) Integrating published HLS data; VII) HLS tool version regression benchmarking.

cs.AR↗

arXiv2024

GreenFPGA: Evaluating FPGAs as Environmentally Sustainable Computing Solutions

Growing global concerns about climate change highlight the need for environmentally sustainable computing. The ecological impact of computing, including operational and embodied, is a key consideration. Field Programmable Gate Arrays (FPGAs) stand out as promising sustainable computing platforms due to their reconfigurability across various applications. This paper introduces GreenFPGA, a tool estimating the total carbon footprint (CFP) of FPGAs over their lifespan, considering design, manufacturing, reconfigurability (reuse), operation, disposal, and recycling. Using GreenFPGA, the paper evaluates scenarios where the ecological benefits of FPGA reconfigurability outweigh operational and embodied carbon costs, positioning FPGAs as a environmentally sustainable choice for hardware acceleration compared to Application-Specific Integrated Circuits (ASICs). Experimental results show that FPGAs have lower CFP than ASICs, particularly for multiple distinct, low-volume applications, or short application lifespans.

cs.AR↗

arXiv2024

Efficient Approaches for GEMM Acceleration on Leading AI-Optimized FPGAs

FPGAs are a promising platform for accelerating Deep Learning (DL) applications, due to their high performance, low power consumption, and reconfigurability. Recently, the leading FPGA vendors have enhanced their architectures to more efficiently support the computational demands of DL workloads. However, the two most prominent AI-optimized FPGAs, i.e., AMD/Xilinx Versal ACAP and Intel Stratix 10 NX, employ significantly different architectural approaches. This paper presents novel systematic frameworks to optimize the performance of General Matrix Multiplication (GEMM), a fundamental operation in DL workloads, by exploiting the unique and distinct architectural characteristics of each FPGA. Our evaluation on GEMM workloads for int8 precision shows up to 77 and 68 TOPs (int8) throughput, with up to 0.94 and 1.35 TOPs/W energy efficiency for Versal VC1902 and Stratix 10 NX, respectively. This work provides insights and guidelines for optimizing GEMM-based applications on both platforms, while also delving into their programmability trade-offs and associated challenges.

cs.AR↗

arXiv2024

Field-Programmable Gate Array Architecture for Deep Learning: Survey & Future Directions

Deep learning (DL) is becoming the cornerstone of numerous applications both in datacenters and at the edge. Specialized hardware is often necessary to meet the performance requirements of state-of-the-art DL models, but the rapid pace of change in DL models and the wide variety of systems integrating DL make it impossible to create custom computer chips for all but the largest markets. Field-programmable gate arrays (FPGAs) present a unique blend of reprogrammability and direct hardware execution that make them suitable for accelerating DL inference. They offer the ability to customize processing pipelines and memory hierarchies to achieve lower latency and higher energy efficiency compared to general-purpose CPUs and GPUs, at a fraction of the development time and cost of custom chips. Their diverse high-speed IOs also enable directly interfacing the FPGA to the network and/or a variety of external sensors, making them suitable for both datacenter and edge use cases. As DL has become an ever more important workload, FPGA architectures are evolving to enable higher DL performance. In this article, we survey both academic and industrial FPGA architecture enhancements for DL. First, we give a brief introduction on the basics of FPGA architecture and how its components lead to strengths and weaknesses for DL applications. Next, we discuss different styles of DL inference accelerators on FPGA, ranging from model-specific dataflow styles to software-programmable overlay styles. We survey DL-specific enhancements to traditional FPGA building blocks such as logic blocks, arithmetic circuitry, and on-chip memories, as well as new in-fabric DL-specialized blocks for accelerating tensor computations. Finally, we discuss hybrid devices that combine processors and coarse-grained accelerator blocks with FPGA-like interconnect and networks-on-chip, and highlight promising future research directions.

cs.AR↗

arXiv2023

PIMSAB: A Processing-In-Memory System with Spatially-Aware Communication and Bit-Serial-Aware Computation

Bit-serial Processing-In-Memory (PIM) is an attractive paradigm for accelerator architectures, for parallel workloads such as Deep Learning (DL), because of its capability to achieve massive data parallelism at a low area overhead and provide orders-of-magnitude data movement savings by moving computational resources closer to the data. While many PIM architectures have been proposed, improvements are needed in communicating intermediate results to consumer kernels, for communication between tiles at scale, for reduction operations, and for efficiently performing bit-serial operations with constants. We present PIMSAB, a scalable architecture that provides spatially aware communication network for efficient intra-tile and inter-tile data movement and provides efficient computation support for generally inefficient bit-serial compute patterns. Our architecture consists of a massive hierarchical array of compute-enabled SRAMs (CRAMs) and is codesigned with a compiler to achieve high utilization. The key novelties of our architecture are: (1) providing efficient support for spatially-aware communication by providing local H-tree network for reductions, by adding explicit hardware for shuffling operands, and by deploying systolic broadcasting, and (2) taking advantage of the divisible nature of bit-serial computations through adaptive precision, bit-slicing and efficient handling of constant operations. When compared against a similarly provisioned modern Tensor Core GPU (NVIDIA A100), across common DL kernels and an end-to-end DL network (Resnet18), PIMSAB outperforms the GPU by 3x, and reduces energy by 4.2x. We compare PIMSAB with similarly provisioned state-of-the-art SRAM PIM (Duality Cache) and DRAM PIM (SIMDRAM) and observe a speedup of 3.7x and 3.88x respectively.

cs.AR↗

arXiv2023

MaxEVA: Maximizing the Efficiency of Matrix Multiplication on Versal AI Engine

The increasing computational and memory requirements of Deep Learning (DL) workloads has led to outstanding innovations in hardware architectures. An archetype of such architectures is the novel Versal AI Engine (AIE) by AMD/Xilinx. The AIE comprises multiple programmable processors optimized for vector-based algorithms. An AIE array consisting of 400 processor cores, operating at 1.25 GHz is able to deliver a peak throughput of 8 TFLOPs for 32-bit floating-point (fp32), and 128 TOPs for 8-bit integer (int8) precision. In this work, we propose MaxEVA: a novel framework to efficiently map Matrix Multiplication (MatMul) workloads on Versal AIE devices. Our framework maximizes the performance and energy efficiency of MatMul applications by efficiently exploiting features of the AIE architecture and resolving performance bottlenecks from multiple angles. When demonstrating on the VC1902 device of the VCK190 board, MaxEVA accomplishes up to 5.44 TFLOPs and 77.01 TOPs throughput for fp32 and int8 precisions, respectively. In terms of energy efficiency, MaxEVA attains up to 124.16 GFLOPs/W for fp32, and 1.16 TOPs/W for int8. Our proposed method substantially outperforms the state-of-the-art approach by exhibiting up to 2.19x throughput gain and 20.4% higher energy efficiency. The MaxEVA framework provides notable insights to fill the knowledge gap in effectively designing MatMul-based DL workloads on the new Versal AIE devices.

cs.AR↗

arXiv2023

HLSDataset: Open-Source Dataset for ML-Assisted FPGA Design using High Level Synthesis

Machine Learning (ML) has been widely adopted in design exploration using high level synthesis (HLS) to give a better and faster performance, and resource and power estimation at very early stages for FPGA-based design. To perform prediction accurately, high-quality and large-volume datasets are required for training ML models.This paper presents a dataset for ML-assisted FPGA design using HLS, called HLSDataset. The dataset is generated from widely used HLS C benchmarks including Polybench, Machsuite, CHStone and Rossetta. The Verilog samples are generated with a variety of directives including loop unroll, loop pipeline and array partition to make sure optimized and realistic designs are covered. The total number of generated Verilog samples is nearly 9,000 per FPGA type. To demonstrate the effectiveness of our dataset, we undertake case studies to perform power estimation and resource usage estimation with ML models trained with our dataset. All the codes and dataset are public at the github repo.We believe that HLSDataset can save valuable time for researchers by avoiding the tedious process of running tools, scripting and parsing files to generate the dataset, and enable them to spend more time where it counts, that is, in training ML models.

cs.AR↗

arXiv2023

ULEEN: A Novel Architecture for Ultra Low-Energy Edge Neural Networks

The deployment of AI models on low-power, real-time edge devices requires accelerators for which energy, latency, and area are all first-order concerns. There are many approaches to enabling deep neural networks (DNNs) in this domain, including pruning, quantization, compression, and binary neural networks (BNNs), but with the emergence of the "extreme edge", there is now a demand for even more efficient models. In order to meet the constraints of ultra-low-energy devices, we propose ULEEN, a model architecture based on weightless neural networks. Weightless neural networks (WNNs) are a class of neural model which use table lookups, not arithmetic, to perform computation. The elimination of energy-intensive arithmetic operations makes WNNs theoretically well suited for edge inference; however, they have historically suffered from poor accuracy and excessive memory usage. ULEEN incorporates algorithmic improvements and a novel training strategy inspired by BNNs to make significant strides in improving accuracy and reducing model size. We compare FPGA and ASIC implementations of an inference accelerator for ULEEN against edge-optimized DNN and BNN devices. On a Xilinx Zynq Z-7045 FPGA, we demonstrate classification on the MNIST dataset at 14.3 million inferences per second (13 million inferences/Joule) with 0.21 $μ$s latency and 96.2% accuracy, while Xilinx FINN achieves 12.3 million inferences per second (1.69 million inferences/Joule) with 0.31 $μ$s latency and 95.83% accuracy. In a 45nm ASIC, we achieve 5.1 million inferences/Joule and 38.5 million inferences/second at 98.46% accuracy, while a quantized Bit Fusion model achieves 9230 inferences/Joule and 19,100 inferences/second at 99.35% accuracy. In our search for ever more efficient edge devices, ULEEN shows that WNNs are deserving of consideration.

cs.AR↗

arXiv2022

CoMeFa: Compute-in-Memory Blocks for FPGAs

Block RAMs (BRAMs) are the storage houses of FPGAs, providing extensive on-chip memory bandwidth to the compute units implemented using Logic Blocks (LBs) and Digital Signal Processing (DSP) slices. We propose modifying BRAMs to convert them to CoMeFa (Compute-In-Memory Blocks for FPGAs) RAMs. These RAMs provide highly-parallel compute-in-memory by combining computation and storage capabilities in one block. CoMeFa RAMs utilize the true dual port nature of FPGA BRAMs and contain multiple programmable single-bit bit-serial processing elements. CoMeFa RAMs can be used to compute in any precision, which is extremely important for evolving applications like Deep Learning. Adding CoMeFa RAMs to FPGAs significantly increases their compute density. We explore and propose two architectures of these RAMs: CoMeFa-D (optimized for delay) and CoMeFa-A (optimized for area). Compared to existing proposals, CoMeFa RAMs do not require changing the underlying SRAM technology like simultaneously activating multiple rows on the same port, and are practical to implement. CoMeFa RAMs are versatile blocks that find applications in numerous diverse parallel applications like Deep Learning, signal processing, databases, etc. By augmenting an Intel Arria-10-like FPGA with CoMeFa-D (CoMeFa-A) RAMs at the cost of 3.8% (1.2%) area, and with algorithmic improvements and efficient mapping, we observe a geomean speedup of 2.5x (1.8x), across several representative benchmarks. Replacing all or some BRAMs with CoMeFa RAMs in FPGAs can make them better accelerators of modern compute-intensive workloads.

cs.AR↗

arXiv2022

Weightless Neural Networks for Efficient Edge Inference

Weightless Neural Networks (WNNs) are a class of machine learning model which use table lookups to perform inference. This is in contrast with Deep Neural Networks (DNNs), which use multiply-accumulate operations. State-of-the-art WNN architectures have a fraction of the implementation cost of DNNs, but still lag behind them on accuracy for common image recognition tasks. Additionally, many existing WNN architectures suffer from high memory requirements. In this paper, we propose a novel WNN architecture, BTHOWeN, with key algorithmic and architectural improvements over prior work, namely counting Bloom filters, hardware-friendly hashing, and Gaussian-based nonlinear thermometer encodings to improve model accuracy and reduce area and energy consumption. BTHOWeN targets the large and growing edge computing sector by providing superior latency and energy efficiency to comparable quantized DNNs. Compared to state-of-the-art WNNs across nine classification datasets, BTHOWeN on average reduces error by more than than 40% and model size by more than 50%. We then demonstrate the viability of the BTHOWeN architecture by presenting an FPGA-based accelerator, and compare its latency and resource usage against similarly accurate quantized DNN accelerators, including Multi-Layer Perceptron (MLP) and convolutional models. The proposed BTHOWeN models consume almost 80% less energy than the MLP models, with nearly 85% reduction in latency. In our quest for efficient ML on the edge, WNNs are clearly deserving of additional attention.

cs.AR↗

arXiv2021

Compute RAMs: Adaptable Compute and Storage Blocks for DL-Optimized FPGAs

The configurable building blocks of current FPGAs -- Logic blocks (LBs), Digital Signal Processing (DSP) slices, and Block RAMs (BRAMs) -- make them efficient hardware accelerators for the rapid-changing world of Deep Learning (DL). Communication between these blocks happens through an interconnect fabric consisting of switching elements spread throughout the FPGA. In this paper, a new block, Compute RAM, is proposed. Compute RAMs provide highly-parallel processing-in-memory (PIM) by combining computation and storage capabilities in one block. Compute RAMs can be integrated in the FPGA fabric just like the existing FPGA blocks and provide two modes of operation (storage or compute) that can be dynamically chosen. They reduce power consumption by reducing data movement, provide adaptable precision support, and increase the effective on-chip memory bandwidth. Compute RAMs also help increase the compute density of FPGAs. In our evaluation of addition, multiplication and dot-product operations across multiple data precisions (int4, int8 and bfloat16), we observe an average savings of 80% in energy consumption, and an improvement in execution time ranging from 20% to 80%. Adding Compute RAMs can benefit non-DL applications as well, and make FPGAs more efficient, flexible, and performant accelerators.

cs.AR↗

arXiv2021

Koios: A Deep Learning Benchmark Suite for FPGA Architecture and CAD Research

With the prevalence of deep learning (DL) in many applications, researchers are investigating different ways of optimizing FPGA architecture and CAD to achieve better quality-of-results (QoR) on DL-based workloads. In this optimization process, benchmark circuits are an essential component; the QoR achieved on a set of benchmarks is the main driver for architecture and CAD design choices. However, current academic benchmark suites are inadequate, as they do not capture any designs from the DL domain. This work presents a new suite of DL acceleration benchmark circuits for FPGA architecture and CAD research, called Koios. This suite of 19 circuits covers a wide variety of accelerated neural networks, design sizes, implementation styles, abstraction levels, and numerical precisions. These designs are larger, more data parallel, more heterogeneous, more deeply pipelined, and utilize more FPGA architectural features compared to existing open-source benchmarks. This enables researchers to pin-point architectural inefficiencies for this class of workloads and optimize CAD tools on more realistic benchmarks that stress the CAD algorithms in different ways. In this paper, we describe the designs in our benchmark suite, present results of running them through the Verilog-to-Routing (VTR) flow using a recent FPGA architecture model, and identify key insights from the resulting metrics. On average, our benchmarks have 3.7x more netlist primitives, 1.8x and 4.7x higher DSP and BRAM densities, and 1.7x higher frequency with 1.9x more near-critical paths compared to the widely-used VTR suite. Finally, we present two example case studies showing how architectural exploration for DL-optimized FPGAs can be performed using our new benchmark suite.

cs.AR↗