HyperX: A Hybrid RRAM-SRAM partitioned system for error recovery in memristive Xbars
Adarsh Kostaa, Efstathia Souflerib, Indranil Chakrabortyc, Amogh Agrawald, Aayush Ankite and Kaushik Royf
Purdue University, West Lafayette, USA
aakosta@purdue.edu
besoufler@purdue.edu
cichakra@purdue.edu
dagrawa64@purdue.edu
eaankit@purdue.edu
fkaushik@purdue.edu
ABSTRACT
Memristive crossbars based on Non-volatile Memory (NVM) technologies such as RRAM, have recently shown great promise for accelerating Deep Neural Networks (DNNs). They achieve this by performing efficient Matrix-Vector-Multiplications (MVMs) while offering dense on-chip storage and minimal offchip data movement. However, their analog nature of computing introduces functional errors due to non-ideal RRAM devices, significantly degrading the application accuracy. Further, RRAMs suffer from low endurance and high write costs, hindering on-chip trainability. To alleviate these limitations, we propose HyperX, a hybrid RRAM-SRAM system that leverages the complementary benefits of NVM and CMOS technologies. Our proposed system consists of a fixed RRAM block offering area and energy-efficient MVMs and an SRAM block enabling on-chip training to recover the accuracy drop due to the RRAM non-idealities. The improvements are reported in terms of energy and product of latency and area (ms × mm2), termed as area-normalized latency. Our experiments on CIFAR datasets using ResNet-20 show up to 2.88× and 10.1× improvements in inference energy and area-normalized latency, respectively. In addition, for a transfer learning task from ImageNet to CIFAR datasets using ResNet-18, we observe up to 1.58× and 4.48× improvements in energy and area-normalized latency, respectively. These improvements are with respect to an all-SRAM baseline.
Keywords: Hybrid, RRAM, SRAM, NVM, Crossbar, Energyefficiency, Transfer-Learning.
