Search

Processing-in-Memory (PIM) enhances memory with computational capabilities, potentially solving energy and latency issues associated with data transfer between memory and processors. However, managing concurrent computation and data flow within the PIM architecture incurs significant latency and energy penalty for applications. This paper introduces Shared-PIM, an architecture for in-DRAM PIM that strategically allocates rows in memory banks, bolstered by memory peripherals, for concurrent processing and data movement. Shared-PIM enables simultaneous computation and data transfer within a memory bank. When compared to LISA, a state-of-the-art architecture that facilitates data transfers for in-DRAM PIM, Shared-PIM reduces data movement latency and energy by 5x and 1.2x respectively. Furthermore, when integrated to a state-of-the-art (SOTA) in-DRAM PIM architecture (pLUTo), Shared-PIM achieves 1.4x faster addition and multiplication, and thereby improves the performance of matrix multiplication (MM) tasks by 40%, polynomial multiplication (PMM) by 44%, and numeric number transfer (NTT) tasks by 31%. Moreover, for graph processing tasks like Breadth-First Search (BFS) and Depth-First Search (DFS), Shared-PIM achieves a 29% improvement in speed, all with an area overhead of just 7.16% compared to the baseline pLUTo.

As the reliance on secure memory environments permeates across applications, memory encryption is used to ensure memory security. However, most effective encryption schemes, such as the widely used AES-CTR, inherently introduce extra overheads, including those associated with counter storage and version number integrity checks. Moreover, encryption only protects data content, and it does not fully address the memory access pattern leakage. While Oblivious RAM (ORAM) aims to obscure these patterns, its high performance costs hinder practical applications. We introduce Secure Scattered Memory (SSM), an efficient scheme provides a comprehensive security solution that preserves the confidentiality of data content without traditional encryption, protects access patterns, and enables efficient integrity verification. Moving away from traditional encryption-centric methods, SSM offers a fresh approach to protecting data content while eliminating counter-induced overheads. Moreover, SSM is designed to inherently obscure memory access patterns, thereby significantly enhancing the confidentiality of memory data. In addition, SSM incorporates lightweight, thus integrated mechanisms for integrity assurance, protecting against data tampering. We also introduce SSM+, an extension that adapts Path ORAM to offer even greater security guarantees for both data content and memory access patterns, demonstrating its flexibility and efficiency. Experimental results show that SSM incurs only a 10% performance overhead compared to non-protected memory and offers a 15% improvement over AES-CTR mode memory protection. Notably, SSM+ provides an 20% improvement against Path ORAM integrated with Intel SGX under the highest security guarantees.

Earth Mover's Distance (EMD) is an important similarity measure between two distributions, used in computer vision and many other application domains. However, its exact calculation is computationally and memory intensive, which hinders its scalability and applicability for large-scale problems. Various approximate EMD algorithms have been proposed to reduce computational costs, but they suffer lower accuracy and may require additional memory usage or manual parameter tuning. In this paper, we present a novel approach, NNS-EMD, to approximate EMD using Nearest Neighbor Search (NNS), in order to achieve high accuracy, low time complexity, and high memory efficiency. The NNS operation reduces the number of data points compared in each NNS iteration and offers opportunities for parallel processing. We further accelerate NNS-EMD via vectorization on GPU, which is especially beneficial for large datasets. We compare NNS-EMD with both the exact EMD and state-of-the-art approximate EMD algorithms on image classification and retrieval tasks. We also apply NNS-EMD to calculate transport mapping and realize color transfer between images. NNS-EMD can be 44x to 135x faster than the exact EMD implementation, and achieves superior accuracy, speedup, and memory efficiency over existing approximate EMD methods.

Privacy has rapidly become a major concern/design consideration. Homomorphic Encryption (HE) and Garbled Circuits (GC) are privacy-preserving techniques that support computations on encrypted data. HE and GC can complement each other, as HE is more efficient for linear operations, while GC is more effective for non-linear operations. Together, they enable complex computing tasks, such as machine learning, to be performed exactly on ciphertexts. However, HE and GC introduce two major bottlenecks: an elevated computational overhead and high data transfer costs. This paper presents PPIMCE, an in-memory computing (IMC) fabric designed to mitigate both computational overhead and data transfer issues. Through the use of multiple IMC cores for high parallelism, and by leveraging in-SRAM IMC for data management, PPIMCE offers a compact, energy-efficient solution for accelerating HE and GC. PPIMCE achieves a 107X speedup against a CPU implementation of GC. Additionally, PPIMCE achieves a 1,500X and 800X speedup compared to CPU and GPU implementations of CKKS-based HE multiplications. For privacy-preserving machine learning inference, PPIMCE attains a 1,000X speedup compared to CPU and a 12X speedup against CraterLake, the state-of-art privacy preserving computation accelerator.

Lattice-based cryptographic algorithms built on ring learning with error theory are gaining importance due to their potential for providing post-quantum security. However, these algorithms involve complex polynomial operations, such as polynomial modular multiplication (PMM), which is the most time-consuming part of these algorithms. Accelerating PMM is crucial to make lattice-based cryptographic algorithms widely adopted by more applications. This work introduces a novel high-throughput and compact PMM accelerator, X-Poly, based on the crossbar (XB)-type compute-in-memory (CIM). We identify the most appropriate PMM algorithm for XB-CIM. We then propose a novel bit-mapping technique to reduce the area and energy of the XB-CIM fabric, and conduct processing engine (PE)-level optimization to increase memory utilization and support different problem sizes with a fixed number of XB arrays. X-Poly design achieves 3.1X10^6 PMM operations/s throughput and offers 200X latency improvement compared to the CPU-based implementation. It also achieves 3.9X throughput per area improvement compared with the state-of-the-art CIM accelerators., Comment: Accepted by 42nd International Conference on Computer-Aided Design (ICCAD)

In a number of machine learning models, an input query is searched across the trained class vectors to find the closest feature class vector in cosine similarity metric. However, performing the cosine similarities between the vectors in Von-Neumann machines involves a large number of multiplications, Euclidean normalizations and division operations, thus incurring heavy hardware energy and latency overheads. Moreover, due to the memory wall problem that presents in the conventional architecture, frequent cosine similarity-based searches (CSSs) over the class vectors requires a lot of data movements, limiting the throughput and efficiency of the system. To overcome the aforementioned challenges, this paper introduces COSIME, an general in-memory associative memory (AM) engine based on the ferroelectric FET (FeFET) device for efficient CSS. By leveraging the one-transistor AND gate function of FeFET devices, current-based translinear analog circuit and winner-take-all (WTA) circuitry, COSIME can realize parallel in-memory CSS across all the entries in a memory block, and output the closest word to the input query in cosine similarity metric. Evaluation results at the array level suggest that the proposed COSIME design achieves 333X and 90.5X latency and energy improvements, respectively, and realizes better classification accuracy when compared with an AM design implementing approximated CSS. The proposed in-memory computing fabric is evaluated for an HDC problem, showcasing that COSIME can achieve on average 47.1X and 98.5X speedup and energy efficiency improvements compared with an GPU implementation., Comment: Accepted by the 41st International Conference on Computer Aided Design (ICCAD), San Diego, USA

Recommendation systems (RecSys) suggest items to users by predicting their preferences based on historical data. Typical RecSys handle large embedding tables and many embedding table related operations. The memory size and bandwidth of the conventional computer architecture restrict the performance of RecSys. This work proposes an in-memory-computing (IMC) architecture (iMARS) for accelerating the filtering and ranking stages of deep neural network-based RecSys. iMARS leverages IMC-friendly embedding tables implemented inside a ferroelectric FET based IMC fabric. Circuit-level and system-level evaluation show that \fw achieves 16.8x (713x) end-to-end latency (energy) improvement compared to the GPU counterpart for the MovieLens dataset., Comment: Accepted by 59th Design Automation Conference (DAC)

This paper proposes IMCRYPTO, an in-memory computing (IMC) fabric for accelerating AES encryption and decryption. IMCRYPTO employs a unified structure to implement encryption and decryption in a single hardware architecture, with combined (Inv)SubBytes and (Inv)MixColumns steps. Because of this step-combination, as well as the high parallelism achieved by multiple units of random-access memory (RAM) and random-access/content-addressable memory (RA/CAM) arrays, IMCRYPTO achieves high throughput encryption and decryption without sacrificing area and power consumption. Additionally, due to the integration of a RISC-V core, IMCRYPTO offers programmability and flexibility. IMCRYPTO improves the throughput per area by a minimum (maximum) of 3.3x (223.1x) when compared to previous ASICs/IMC architectures for AES-128 encryption. Projections show added benefit from emerging technologies of up to 5.3x to the area-delay-power product of IMCRYPTO.

Deep random forest (DRF), which incorporates the core features of deep learning and random forest (RF), exhibits comparable classification accuracy, interpretability, and low memory and computational overhead when compared with deep neural networks (DNNs) in various information processing tasks for edge intelligence. However, the development of efficient hardware to accelerate DRF is lagging behind its DNN counterparts. The key for hardware acceleration of DRF lies in efficiently realizing the branch-split operation at decision nodes when traversing a decision tree. In this work, we propose to implement DRF through simple associative searches realized with ferroelectric analog content addressable memory (ACAM). Utilizing only two ferroelectric field effect transistors (FeFETs), the ultra-compact ACAM cell can perform a branch-split operation with an energy-efficient associative search by storing the decision boundaries as the analog polarization states in an FeFET. The DRF accelerator architecture and the corresponding mapping of the DRF model to the ACAM arrays are presented. The functionality, characteristics, and scalability of the FeFET ACAM based DRF and its robustness against FeFET device non-idealities are validated both in experiments and simulations. Evaluation results show that the FeFET ACAM DRF accelerator exhibits 10^6x/16x and 10^6x/2.5x improvements in terms of energy and latency when compared with other deep random forest hardware implementations on the state-of-the-art CPU/ReRAM, respectively., Comment: 44 pages, 16 figures

Hyperdimensional Computing (HDC) is an emerging computational framework that mimics important brain functions by operating over high-dimensional vectors, called hypervectors (HVs). In-memory computing implementations of HDC are desirable since they can significantly reduce data transfer overheads. All existing in-memory HDC platforms consider binary HVs where each dimension is represented with a single bit. However, utilizing multi-bit HVs allows HDC to achieve acceptable accuracies in lower dimensions which in turn leads to higher energy efficiencies. Thus, we propose a highly accurate and efficient multi-bit in-memory HDC inference platform called MIMHD. MIMHD supports multi-bit operations using ferroelectric field-effect transistor (FeFET) crossbar arrays for multiply-and-add and FeFET multi-bit content-addressable memories for associative search. We also introduce a novel hardware-aware retraining framework (HWART) that trains the HDC model to learn to work with MIMHD. For six popular datasets and 4000 dimension HVs, MIMHD using 3-bit (2-bit) precision HVs achieves (i) average accuracies of 92.6% (88.9%) which is 8.5% (4.8%) higher than binary implementations; (ii) 84.1x (78.6x) energy improvement over a GPU, and (iii) 38.4x (34.3x) speedup over a GPU, respectively. The 3-bit $\times$ is 4.3x and 13x faster and more energy-efficient than binary HDC accelerators while achieving similar accuracies., Comment: Accepted at ISLPED 2021

The memory wall bottleneck is a key challenge across many data-intensive applications. Multi-level FeFET-based embedded non-volatile memories are a promising solution for denser and more energy-efficient on-chip memory. However, reliable multi-level cell storage requires careful optimizations to minimize the design overhead costs. In this work, we investigate the interplay between FeFET device characteristics, programming schemes, and memory array architecture, and explore different design choices to optimize performance, energy, area, and accuracy metrics for critical data-intensive workloads. From our cross-stack design exploration, we find that we can store DNN weights and social network graphs at a density of over 8MB/mm^2 and sub-2ns read access latency without loss in application accuracy., Comment: Accepted at ISLPED 2021

Silicon ferroelectric field-effect transistors (FeFETs) with low-k interfacial layer (IL) between ferroelectric gate stack and silicon channel suffers from high write voltage, limited write endurance and large read-after-write latency due to early IL breakdown and charge trapping and detrapping at the interface. We demonstrate low voltage, high speed memory operation with high write endurance using an IL-free back-end-of-line (BEOL) compatible FeFET. We fabricate IL-free FeFETs with 28nm channel length and 126nm width under a thermal budget <400C by integrating 5nm thick Hf0.5Zr0.5O2 gate stack with amorphous Indium Tungsten Oxide (IWO) semiconductor channel. We report 1.2V memory window and read current window of 10^5 for program and erase, write latency of 20ns with +/-2V write pulses, read-after-write latency <200ns, write endurance cycles exceeding 5x10^10 and 2-bit/cell programming capability. Array-level analysis establishes IL-free BEOL FeFET as a promising candidate for logic-compatible high-performance on-chip buffer memory and multi-bit weight cell for compute-in-memory accelerators.

Nearest neighbor (NN) search is an essential operation in many applications, such as one/few-shot learning and image classification. As such, fast and low-energy hardware support for accurate NN search is highly desirable. Ternary content-addressable memories (TCAMs) have been proposed to accelerate NN search for few-shot learning tasks by implementing $L_\infty$ and Hamming distance metrics, but they cannot achieve software-comparable accuracies. This paper proposes a novel distance function that can be natively evaluated with multi-bit content-addressable memories (MCAMs) based on ferroelectric FETs (FeFETs) to perform a single-step, in-memory NN search. Moreover, this approach achieves accuracies comparable to floating-point precision implementations in software for NN classification and one/few-shot learning tasks. As an example, the proposed method achieves a 98.34% accuracy for a 5-way, 5-shot classification task for the Omniglot dataset (only 0.8% lower than software-based implementations) with a 3-bit MCAM. This represents a 13% accuracy improvement over state-of-the-art TCAM-based implementations at iso-energy and iso-delay. The presented distance function is resilient to the effects of FeFET device-to-device variations. Furthermore, this work experimentally demonstrates a 2-bit implementation of FeFET MCAM using AND arrays from GLOBALFOUNDRIES to further validate proof of concept., Comment: To be published in DATE'21

Homomorphic encryption (HE) allows direct computations on encrypted data. Despite numerous research efforts, the practicality of HE schemes remains to be demonstrated. In this regard, the enormous size of ciphertexts involved in HE computations degrades computational efficiency. Near-memory Processing (NMP) and Computing-in-memory (CiM) - paradigms where computation is done within the memory boundaries - represent architectural solutions for reducing latency and energy associated with data transfers in data-intensive applications such as HE. This paper introduces CiM-HE, a Computing-in-memory (CiM) architecture that can support operations for the B/FV scheme, a somewhat homomorphic encryption scheme for general computation. CiM-HE hardware consists of customized peripherals such as sense amplifiers, adders, bit-shifters, and sequencing circuits. The peripherals are based on CMOS technology, and could support computations with memory cells of different technologies. Circuit-level simulations are used to evaluate our CiM-HE framework assuming a 6T-SRAM memory. We compare our CiM-HE implementation against (i) two optimized CPU HE implementations, and (ii) an FPGA-based HE accelerator implementation. When compared to a CPU solution, CiM-HE obtains speedups between 4.6x and 9.1x, and energy savings between 266.4x and 532.8x for homomorphic multiplications (the most expensive HE operation). Also, a set of four end-to-end tasks, i.e., mean, variance, linear regression, and inference are up to 1.1x, 7.7x, 7.1x, and 7.5x faster (and 301.1x, 404.6x, 532.3x, and 532.8x more energy efficient). Compared to CPU-based HE in a previous work, CiM-HE obtain 14.3x speed-up and >2600x energy savings. Finally, our design offers 2.2x speed-up with 88.1x energy savings compared to a state-of-the-art FPGA-based accelerator., Comment: 14 pages

Ferroelectric field effect transistors (FeFETs) are being actively investigated with the potential for in-memory computing (IMC) over other non-volatile memories (NVMs). Content Addressable Memories (CAMs) are a form of IMC that performs parallel searches for matched entries over a memory array for a given input query. CAMs are widely used for data-centric applications that involve pattern matching and search functionality. To accommodate the ever expanding data, it is attractive to resort to analog CAM for memory density improvement. However, the digital CAM design nowadays based on standard CMOS or emerging nonvolatile memories (e.g., resistive storage devices) is already challenging due to area, power, and cost penalties. Thus, it can be extremely expensive to achieve analog CAM with those technologies due to added cell components. As such, we propose, for the first time, a universal compact FeFET based CAM design, FeCAM, with search and storage functionality enabled in digital and analog domain simultaneously. By exploiting the multi-level-cell (MLC) states of FeFET, FeCAM can store and search inputs in either digital or analog domain. We perform a device-circuit co-design of the proposed FeCAM and validate its functionality and performance using an experimentally calibrated FeFET model. Circuit level simulation results demonstrate that FeCAM can either store continuous matching ranges or encode 3-bit data in a single CAM cell. When compared with the existing digital CMOS based CAM approaches, FeCAM is found to improve both memory density by 22.4X and energy saving by 8.6/3.2X for analog/digital modes, respectively. In the CAM-related application, our evaluations show that FeCAM can achieve 60.5X/23.1X saving in area/search energy compared with conventional CMOS based CAMs., Comment: 8 pages, 8 figures, accepted

An analog synapse circuit based on ferroelectric-metal field-effect transistors is proposed, that offers 6-bit weight precision. The circuit is comprised of volatile least significant bits (LSBs) used solely during training, and non-volatile most significant bits (MSBs) used for both training and inference. The design works at a 1.8V logic-compatible voltage, provides 10^10 endurance cycles, and requires only 250ps update pulses. A variant of LeNet trained with the proposed synapse achieves 98.2% accuracy on MNIST, which is only 0.4% lower than an ideal implementation of the same network with the same bit precision. Furthermore, the proposed synapse offers improvements of up to 26% in area, 44.8% in leakage power, 16.7% in LSB update pulse duration, and two orders of magnitude in endurance cycles, when compared to state-of-the-art hybrid synaptic circuits. Our proposed synapse can be extended to an 8-bit design, enabling a VGG-like network to achieve 88.8% accuracy on CIFAR-10 (only 0.8% lower than an ideal implementation of the same network)., Comment: Accepted at ISCAS'20 for oral presentation

We propose a technology-independent method, referred to as adjacent connection matrix (ACM), to efficiently map signed weight matrices to non-negative crossbar arrays. When compared to same-hardware-overhead mapping methods, using ACM leads to improvements of up to 20% in training accuracy for ResNet-20 with the CIFAR-10 dataset when training with 5-bit precision crossbar arrays or lower. When compared with strategies that use two elements to represent a weight, ACM achieves comparable training accuracies, while also offering area and read energy reductions of 2.3x and 7x, respectively. ACM also has a mild regularization effect that improves inference accuracy in crossbar arrays without any retraining or costly device/variation-aware training., Comment: Accepted at DAC'20

A new spintronic nonvolatile memory cell analogous to 1T DRAM with non-destructive read is proposed. The cells can be used as neural computing units. A dual-circuit neural network architecture is proposed to leverage these devices against the complex operations involved in convolutional networks. Simulations based on HSPICE and Matlab were performed to study the performance of this architecture when classifying images as well as the effect of varying the size and stability of the nanomagnets. The spintronic cells outperform a purely charge-based implementation of the same network, consuming about 100 pJ total per image processed.

As cost and performance benefits associated with Moore's Law scaling slow, researchers are studying alternative architectures (e.g., based on analog and/or spiking circuits) and/or computational models (e.g., convolutional and recurrent neural networks) to perform application-level tasks faster, more energy efficiently, and/or more accurately. We investigate cellular neural network (CeNN)-based co-processors at the application-level for these metrics. While it is well-known that CeNNs can be well-suited for spatio-temporal information processing, few (if any) studies have quantified the energy/delay/accuracy of a CeNN-friendly algorithm and compared the CeNN-based approach to the best von Neumann algorithm at the application level. We present an evaluation framework for such studies. As a case study, a CeNN-friendly target-tracking algorithm was developed and mapped to an array architecture developed in conjunction with the algorithm. We compare the energy, delay, and accuracy of our architecture/algorithm (assuming all overheads) to the most accurate von Neumann algorithm (Struck). Von Neumann CPU data is measured on an Intel i5 chip. The CeNN approach is capable of matching the accuracy of Struck, and can offer approximately 1000x improvements in energy-delay product.

Deep neural network (DNN) accelerators with improved energy and delay are desirable for meeting the requirements of hardware targeted for IoT and edge computing systems. Convolutional neural networks (CoNNs) belong to one of the most popular types of DNN architectures. This paper presents the design and evaluation of an accelerator for CoNNs. The system-level architecture is based on mixed-signal, cellular neural networks (CeNNs). Specifically, we present (i) the implementation of different layers, including convolution, ReLU, and pooling, in a CoNN using CeNN, (ii) modified CoNN structures with CeNN-friendly layers to reduce computational overheads typically associated with a CoNN, (iii) a mixed-signal CeNN architecture that performs CoNN computations in the analog and mixed signal domain, and (iv) design space exploration that identifies what CeNN-based algorithm and architectural features fare best compared to existing algorithms and architectures when evaluated over common datasets -- MNIST and CIFAR-10. Notably, the proposed approach can lead to 8.7$\times$ improvements in energy-delay product (EDP) per digit classification for the MNIST dataset at iso-accuracy when compared with the state-of-the-art DNN engine, while our approach could offer 4.3$\times$ improvements in EDP when compared to other network implementations for the CIFAR-10 dataset., Comment: 25 pages