Silicon-based static random access memories (SRAM) and digital Boolean logic have been the workhorse of the state-of-the-art computing platforms. Despite tremendous strides in scaling the ubiquitous metal-oxide-semiconductor transistor, the underlying von-Neumann computing architecture has remained unchanged. The limited throughput and energy-efficiency of the state-of-the-art computing systems, to a large extent, result from the well-known von-Neumann bottleneck. The energy and throughput inefficiency of the von-Neumann machines have been accentuated in recent times due to the present emphasis on data-intensive applications such as artificial intelligence, machine learning, and cryptography. A possible approach towards mitigating the overhead associated with the von-Neumann bottleneck is to enable in-memory Boolean computations. In this paper, we present an augmented version of the conventional SRAM bit-cells, called the X-SRAM, with the ability to perform in-memory, vector Boolean computations, in addition to the usual memory storage operations. We propose at least six different schemes for enabling in-memory vector computations, including NAND, NOR, IMP (implication), XOR logic gates, with respect to different bit-cell topologies - the 8T cell and the 8⁺T Differential cell. In addition, we also present a novel `read-compute-store' scheme, wherein the computed Boolean function can be directly stored in the memory without the need of latching the data and carrying out a subsequent write operation. The feasibility of the proposed schemes has been verified using the predictive transistor models and detailed Monte-Carlo variation analysis. As an illustration, we also present the efficacy of the proposed in-memory computations by implementing advanced encryption standard algorithm on a non-standard von-Neumann machine wherein the conventional SRAM is replaced by X-SRAM. Our simulations indicated that up to 75% of memory accesses can be saved using the proposed techniques.
15 August 2019 in IEEE Transactions on Very Large Scale Integration (VLSI) Systems
Large-scale digital computing almost exclusively relies on the von Neumann architecture, which comprises separate units for storage and computations. The energy-expensive transfer of data from the memory units to the computing cores results in the well-known von Neumann bottleneck. Various approaches aimed toward bypassing the von Neumann bottleneck are being extensively explored in the literature. These include in-memory computing based on CMOS and beyond CMOS technologies, wherein by making modifications to the memory array, vector computations can be carried out as close to the memory units as possible. Interestingly, in-memory techniques based on CMOS technology are of special importance due to the ubiquitous presence of field-effect transistors and the resultant ease of large-scale manufacturing and commercialization. On the other hand, perhaps the most important computation required for applications such as machine learning, etc., comprises the dot-product operation. Emerging nonvolatile memristive technologies have been shown to be very efficient in computing analog dot products in an in situ fashion. The memristive analog computation of the dot product results in much faster operation as opposed to digital vector in-memory bitwise Boolean computations. However, challenges with respect to large-scale manufacturing coupled with the limited endurance of memristors have hindered rapid commercialization of memristive-based computing solutions. In this paper, we show that the standard 8 transistor (8T) digital SRAM array can be configured as an analoglike in-memory multibit dot-product engine (DPE). By applying appropriate analog voltages to the read ports of the 8T SRAM array and sensing the output current, an approximate analog-digital DPE can be implemented. We present two different configurations for enabling multibit dot-product computations in the 8T SRAM cell array, without modifying the standard bit-cell structure. We also demonstrate the robustness of the present proposal in presence of nonidealities such as the effect of line resistances and transistor threshold voltage variations. Since our proposal preserves the standard 8T-SRAM array structure, it can be used as a storage element with standard read-write instructions and also as an on-demand analoglike dot-product accelerator.
30 March 2020 in IEEE Transactions on Circuits and Systems I: Regular Papers
`In-memory computing' is being widely explored as a novel computing paradigm to mitigate the well known memory bottleneck. This emerging paradigm aims at embedding some aspects of computations inside the memory array, thereby avoiding frequent and expensive movement of data between the compute unit and the storage memory. In-memory computing with respect to Silicon memories has been widely explored on various memory bit-cells. Embedding computation inside the 6 transistor (6T) SRAM array is of special interest since it is the most widely used on-chip memory. In this paper, we present a novel in-memory multiplication followed by accumulation operation capable of performing parallel dot products within 6T SRAM without any changes to the standard bitcell. We, further, study the effect of circuit non-idealities and process variations on the accuracy of the LeNet-5 and VGG neural network architectures against the MNIST and CIFAR-10 datasets, respectively. The proposed in-memory dot-product mechanism achieves 88.8% and 99% accuracy for the CIFAR-10 and MNIST, respectively. Compared to the standard von Neumann system, the proposed system is 6.24x better in energy consumption and 9.42x better in delay.
15 October 2019 in IEEE Transactions on Circuits and Systems I: Regular Papers
In-memory computing architectures present a promising solution to address the memory- and the power-wall challenges by mitigating the bottleneck between processing units and storage. Such architectures incorporate computing functionalities inside memory arrays to make better use of the large internal memory bandwidth, thereby, avoiding frequent data movements. In-DRAM computing architectures offer high throughput and energy improvements in accelerating modern data-intensive applications like machine learning etc. In this manuscript, we propose a vector addition methodology inside DRAM arrays through functional read enabled on local word-lines. The proposed primitive performs majority-based addition operations by storing data in transposed manner. Majority functions are achieved in DRAM cells by activating odd number of rows simultaneously. The proposed majority based bit-serial addition enables huge parallelism and high throughput. We validate the robustness of the proposed in-DRAM computing methodology under process variations to ascertain its reliability. Energy evaluation of the proposed scheme shows 21.7X improvement compared to normal data read operations in standard DDR3-1333 interface. Moreover, compared to state-of-the-art in-DRAM compute proposals, the proposed scheme provides one of the fastest addition mechanisms with low area overhead (<1% of DRAM chip area). Our system evaluation running the k-Nearest Neighbor (kNN) algorithm on the MNIST handwritten digit classification dataset shows 11.5X performance improvement compared to a conventional von-Neumann machine.
24 April 2019 in IEEE Transactions on Circuits and Systems I: Regular Papers
Deep neural networks are biologically inspired class of algorithms that have recently demonstrated the state-of-the-art accuracy in large-scale classification and recognition tasks. Hardware acceleration of deep networks is of paramount importance to ensure their ubiquitous presence in future computing platforms. Indeed, a major landmark that enables efficient hardware accelerators for deep networks is the recent advances from the machine learning community that have demonstrated the viability of aggressively scaled deep binary networks. In this paper, we demonstrate how deep binary networks can be accelerated in modified von Neumann machines by enabling binary convolutions within the static random access memory (SRAM) arrays. In general, binary convolutions consist of bit-wise exclusive-nor (xnor) operations followed by a population count (popcount). We present two proposals: one based on charge sharing approach to perform vector xnor and approximate popcount and another based on bit-wise xnor followed by a digital bit-tree adder for accurate popcount. We highlight the various tradeoffs in terms of circuit complexity, speed-up, and classification accuracy for both the approaches. Few key techniques presented as a part of the manuscript are the use of low-precision, low-overhead analog-to-digital converter (ADC), to achieve a fairly accurate popcount for the charge-sharing scheme and proposal for sectioning of the SRAM array by adding switches onto the read-bitlines, thereby achieving improved parallelism. Our results on benchmark image classification datasets for CIFAR-10 and SVHN on a binarized neural network architecture show energy improvements of up to 6.1x and 2.3x for the two proposals, compared to conventional SRAM banks. In terms of latency, improvements of up to 15.8x and 8.1x were achieved for the two respective proposals.
1 March 2017 in IEEE Transactions on Electron Devices
The efficiency of the human brain in performing classification tasks has attracted considerable research interest in brain-inspired neuromorphic computing. The spiking neuromorphic architectures attempt to mimic the computations performed in the brain through a dense interconnection of the neurons and synaptic weights. A leaky-integrate-fire (LIF) spiking model is widely used to emulate the dynamics of the biological neurons. In this paper, we propose a spin-based LIF spiking neuron using the magnetoelectric (ME) switching of ferromagnets. The voltage across the ME oxide exhibits a typical leaky-integrate behavior, which in turn switches an underlying ferromagnet. Due to the effect of thermal noise, the ferromagnet exhibits probabilistic switching dynamics, which is reminiscent of the stochasticity exhibited by biological neurons. The energy efficiency of the ME switching mechanism coupled with the intrinsic nonvolatility of ferromagnets results in lower energy consumption, when compared with a CMOS LIF neuron. A device to system-level simulation framework has been developed to investigate the feasibility of the proposed LIF neuron for a hand-written digit recognition application.