• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2020.11a
• /
• pp.241-243
• /
• 2020

The high computational complexity of deep learning algorithms has led to the development of specialized hardware architectures. However, soft errors (bit flip) may occur in these hardware systems due to voltage variation and high energy particles. Many error correction methods have been proposed to counter this problem. In this work, we analyze an error correction mechanism based on repetition codes and an activation function. We test this method by injecting errors into weight filters and define an ideal error rate range in which the proposed method complements the accuracy of the model in the presence of error.

• ETRI Journal
• /
• v.17 no.2
• /
• pp.21-30
• /
• 1995

A new architecture of field programmable gate array for high-speed datapath applications is presented. Its implementation is facilitated by a configurable interconnect technology based on a one-time, two-terminal programmable, very low-impedance anti-fuse and by a configurable logic module optimized for datapath applications. The configurable logic module can effectively implement diverse logic functions including sequential elements such as latches and flip-flops, and arithmetic functions such as one-bit full adder and two-bit comparator. A novel programming architecture is designed for supplying large current through the anti-fuse element, which drops the on-resistance of anti-fuse below $20{\Omega}$. The chip has been fabricated using a $0.8-{\mu}m$ n-well complementary metal oxide semiconductor technology with two layers of metalization.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.22 no.5
• /
• pp.1019-1025
• /
• 2012

Fault attacks manipulate the computation of an algorithm and get information about the private key from the erroneous result. It is the most powerful attack for the cryptographic device. Currently, the research on error detection methods and fault attacks have been studied actively. S. Pontarelli et al. introduced an error detection method in 2009. It can detect an error that occurs during Elliptic Curve Scalar Multiplication (ECSM). In this paper, we present a new fault attack. Our attack can avoid the error detection method introduced by S. Pontarelli et al. We inject a bit flip error in the Euclidean Addition Chain (EAC) on the private key in ECSM and retrieve the private key.

• Journal of the Semiconductor & Display Technology
• /
• v.21 no.4
• /
• pp.27-32
• /
• 2022

RISC-V is an open instruction set architecture (ISA) developed in 2010 at UC Berkeley, and active research is being conducted as a processor to compete with ARM. In this paper, we propose an SoC system including an RV32I ISA-based 32-bit 5-stage pipeline processor and AHB bus master. The proposed RISC-V processor supports 37 instructions, excluding FENCE, ECALL, and EBREAK instructions, out of a total of 40 instructions based on RV32I ISA. In addition, the RISC-V processor can be connected to peripheral devices such as BRAM, UART, and TIMER using the AHB-lite bus protocol through the proposed AHB bus master. The proposed SoC system was implemented in Arty A7-35T FPGA with 1,959 LUTs and 1,982 flip-flops. Furthermore, the proposed hardware has a maximum operating frequency of 50 MHz. In the Dhrystone benchmark, the proposed processor performance was confirmed to be 0.48 DMIPS.

• Journal of the Microelectronics and Packaging Society
• /
• v.17 no.4
• /
• pp.73-76
• /
• 2010

A two-facet approach was used to investigate the parametric performance of functional high-speed DDR3 (Double Data Rate) DRAM (Dynamic Random Access Memory) die placed in different types of BGA (Ball Grid Array) packages: wire-bonded BGA (FBGA, Fine Ball Grid Array), flip-chip (FCBGA) and lead-bonded $microBGA^{(R)}$. In the first section, packaged live DDR3 die were tested using automatic test equipment using high-resolution shmoo plots. It was found that the best timing and voltage margin was obtained using the lead-bonded microBGA, followed by the wire-bonded FBGA with the FCBGA exhibiting the worst performance of the three types tested. In particular the flip-chip packaged devices exhibited reduced operating voltage margin. In the second part of this work a test system was designed and constructed to mimic the electrical environment of the data bus in a PC's CPU-Memory subsystem that used a single DIMM (Dual In Line Memory Module) socket in point-to-point and point-to-two-point configurations. The emulation system was used to examine signal integrity for system-level operation at speeds in excess of 6 Gb/pin/sec in order to assess the frequency extensibility of the signal-carrying path of the microBGA considered for future high-speed DRAM packaging. The analyzed signal path was driven from either end of the data bus by a GaAs laser driver capable of operation beyond 10 GHz. Eye diagrams were measured using a high speed sampling oscilloscope with a pulse generator providing a pseudo-random bit sequence stimulus for the laser drivers. The memory controller was emulated using a circuit implemented on a BGA interposer employing the laser driver while the active DRAM was modeled using the same type of laser driver mounted to the DIMM module. A custom silicon loading die was designed and fabricated and placed into the microBGA packages that were attached to an instrumented DIMM module. It was found that 6.6 Gb/sec/pin operation appears feasible in both point to point and point to two point configurations when the input capacitance is limited to 2pF.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.1
• /
• pp.1-8
• /
• 2022

RISC-V is an open-source instruction set architecture which has a simple base structure and can be extensible depending on the purpose. In this paper, we designed a small and low-power 32-bit RISC-V processor to establish the base for research on RISC-V embedded systems. We designed a 2-stage pipelined processor which supports RISC-V base integer instruction set except for FENCE and EBREAK instructions. The processor also supports privileged ISA for trap handling. It used 1895 LUTs and 1195 flip-flops, and consumed 0.001W on Xilinx Zynq-7000 FPGA when synthesized using Vivado Design Suite. GPIO, UART, and timer peripherals are additionally used to compose the system. We verified the operation of the processor on FPGA with FreeRTOS at 16MHz. We used Dhrystone and Coremark benchmarks to measure the performance of the processor. This study aims to provide a low-power, high-efficiency microprocessor for future extension.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.33A no.11
• /
• pp.140-151
• /
• 1996

In this study, we designed the routing structure and logic block of a SRAM cell-based FPGA with symmetrical-array architecture. The designed routing structure is composed of switch matrices, routing channels and I/O blocks, and the routing channels can be subdivided into single length channels, double length channels and global length channels. The interconnection between wires is made through SRAM cell-controlled pass transistors. To reduce the signal delay in pass transistors, we proposed a scheme raising the gate-control voltage to 7V. The designed SRAM cells have built-in shift register capability, so there is no need for separate shift registers. We designed SRAM cells in the LUTs(look-up tables) to enable the wirte operations to be performed synchronously with the clock for ease of system application. Each logic block (LFU) has four 4-input LUTs, flip-flops and other gates, and the LUTs can be used a sSRAM memory. The LFU also has a dedicated carry logic, so a 4-bit adder can be implemented in one LFU. We designed our FPGA using 0.6.mu.m CMOS technology, and simulation shows proper operation of a 4 bit counter at 100MHz.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.33 no.1C
• /
• pp.140-148
• /
• 2008

Using the self duality of an optimal normal basis(ONB) of type II, we present a bit parallel and bit serial systolic arrays over GF($2^m$) which has a low hardware complexity and a low latency. We show that our multiplier has a latency m+1 and the basic cell of our circuit design needs 5 latches(flip-flops). Comparing with other arrays of the same kinds, we find that our array has significantly reduced latency and hardware complexity.

• Journal of IKEEE
• /
• v.27 no.3
• /
• pp.220-224
• /
• 2023

Row hammering is a phenomenon in which bit flips occur in adjacent rows when accessing a particular row continuously, causing data damage, security problems, and poor computing performance. This paper analyzes the cause and response method of row hammering through TCAD simulation in 2ynm DRAM. In the experiment, the row hammering is reproduced while changing the parameters of the trap and the device structure, and the trap density, temperature. It analyzes the relationship with Active Wisdom, etc. As a result, it was confirmed that changes in trap parameters and device structures directly affect ΔV_cap/pulse. This enables a fundamental understanding of low hammering and finding countermeasures, and can contribute to improving the stability and security of DRAM.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.39 no.12
• /
• pp.1174-1180
• /
• 2011

Field Programmable Gate Array(FPGA)s are replacing traditional integrated circuits for space applications due to their lower development cost as well as reconfigurability. However, they are very sensitive to single event upset (SEU) caused by space radiation environment. In order to mitigate the SEU, on-board computer of STSAT-3 employed a triple modular redundancy(TMR) and scrubbing scheme. Experimental results showed that upset threshold energy was improved from 10.6 MeV to 20.3 MeV when the TMR and the scrubbing were applied to the on-board computer. Combining the experimental results with the orbit simulation results, calculated bit-flip rate of on-board computer is 1.23 bit-flips/day assuming in the worst case of STSAT-3 orbit.