Dominant mechanisms of transient-radiation upset in CMOS RAM VLSI circuits realized in SOS technology

The dominant mechanisms are analyzed of transient-radiation upset in CMOS RAM VLSI circuits realized in SOS technology. Data reliability under transient irradiation is discussed in relation to photocurrents, rail-span collapse, and the circuit and layout design of memory cells. The response is simulated of SOS integrated resistors to transient radiation. Optimal parameter values are thus determined for the resistor used in the RC network of a memory cell. It is found that the data reliability of the memory circuits considered is affected by the cross coupling of memory cells sharing a read/write line. The lifetime of radiation-induced charge carriers is estimated by experiment and computer simulation.     

Design considerations and strategies for high-reliable STT-MRAM

Benefiting from Spin Transfer Torque (STT) switching approach, second generation of Magnetic RAM (MRAM) promises low power, great miniaturization prospective (<22 nm) and easy integration with CMOS process. It becomes actually a strong non-volatile memory candidate for both embedded and standalone applications. However STT-MRAM suffers from important reliability issues compared with the conventional one based on magnetic field switching, for example, a read-current could write erroneously the stored data, the low Resistance Area (RA) value drives high sensing error rate. This paper presents the considerations and strategies from design point of view for the reliability enhancement. Mixed transient and statistical simulations have been performed by using a STT-MRAM compact model and CMOS 65 nm design kit.     

An efficient NBTI sensor and compensation circuit for stable and reliable SRAM cells

Aggressive technology scaling causes unavoidable reliability issues in modern high-performance integrated circuits. The major reliability factors in nanoscale VLSI design is the negative bias temperature instability (NBTI) degradation and soft-errors in the space and terrestrial environment. In this paper, an on-chip analog adaptive body bias (OA-ABB) circuit to compensate the degradation due to NBTI aging is presented. The OA-ABB is used to compensate the parameter variations and improves the SRAM circuit yield regarding read current, hold SNM, read SNM, write margin and word line write margin (WLWM). The OA-ABB consists of standby leakage current sensor circuit, decision circuit and body bias control circuit. Circuit level simulation for SRAM cell is performed for pre- and post-stress of 10 years NBTI aging. The proposed OA-ABB reduces the effect of NBTI on the stability of SRAM cell. The simulation results show the hold SNM, read SNM and WLWM decreases by 10.55%, 8.55%, and 3.25% respectively in the absence of OA-ABB whereas hold SNM, read SNM and WLWM decreases by only 0.61%, 1.48%, and 0.72% respectively by using OA-ABB to compensate the degradation. The figure of merit of 6T SRAM cell also improved by 17.24% with the use of OA-ABB.     

A third-generation SPARC V9 64-b microprocessor

This quad-issue processor achieves 1-GHz operation through improved dynamic circuit techniques in critical paths and a more extensive on-chip memory system which scales in both bandwidth and latency. Critical logic paths use domino, delayed clocked domino, and logic embedded in dynamic flip-flops for minimum delay. A 64-KB sum-addressed memory data cache combines the address offset add with the cache decode, allowing the average memory latency to scale by more than the clock ratio. Memory bandwidth is improved by using wave pipelined SRAM designs for on-chip caches and a write cache for store traffic. Memory power is controlled without increased latency by use of delayed-reset logic decoders. The chip operates at 1000 MHz and dissipates less than 80 W from a 1.6-V supply. It contains 23 million transistors (12 million in RAM cells) on a 244 mm² die     

基于FPGA的SDRAM控制器设计与实现

针对SDRAM(Synchronous Dynamic Random Access Memory)在缓存图像数据时时序的控制比较复杂的问题,在研究SDRAM的特点和原理的基础上,提出了一种基于现场可编程逻辑器件FPGA(Field Programmable Gate Array)为核心的SDRAM控制器的设计方案。采用分模块的思想,把SDRAM的控制分成不同的功能模块,各模块之间通过信号状态线相互关联,并且相关模块利用状态机来控制整个时序的过程。另外,为了提高SDRAM的缓存速度,选择了SDRAM工作在页突发操作模式下,使SDRAM的读写速度有了大幅的提升。整个控制系统经过仿真和在线逻辑分析仪验证表明:控制器能准确地对SDRAM进行读写控制,稳定可靠,可应用于不同的高速缓存系统。     

A CMOS/SOS 4K static RAM

Discusses high density CMOS/SOS technology used to develop a fully static 4096-bit RAM with a five-transistor storage cell. Selection of a five-transistor memory cell has reduced the access to the flip-flop storage element to a single word line transistor and bit line. The word line transistor must be able to prevent data altering currents from entering the memory cell at all times except for the write operation. The write operation is enhanced by reducing the bias voltage across the memory cell, thereby making the current needed to alter the cell smaller. Through the use of a 5 /spl mu/m design rule, the memory cell occupies 2913 /spl mu/m/SUP 2/. The 4096-bit static CMOS/SOS RAM contains 22553 transistors in 20 mm/SUP 2/. Organised as 1024 4-bit words, the RAM has a read cycle time of 350 ns and standby power dissipation of 50 /spl mu/W at V/SUB cc/=5 V and temperature of 27/spl deg/C.     

Assessment of STT-MRAMs based on double-barrier MTJs for cache applications by means of a device-to-system level simulation framework

This paper explores non-volatile cache memories implemented by spin-transfer torque magnetic random access memories (STT-MRAMs) based on state-of-the-art perpendicular magnetic tunnel junctions (MTJs) and FinFETs. The use of double-barrier MTJs with two reference layers (DMTJs) is benchmarked against solutions relying on single-barrier MTJs (SMTJs) at different technology nodes (from 28-nm down to 20-nm). Our study is carried out through a cross-layer simulation platform, starting from the device- up to the system-level. Our results point out that, thanks to the reduced switching currents, DMTJ-based STT-MRAMs allow decreasing write access time of about 63% as compared to their SMTJ-based counterparts. This is achieved while assuring reduced energy consumption under both write (−42%) and read (−28%) accesses, lower area occupancy (−40%) and smaller leakage power (−25%), at the only cost of worsened read access time. This makes DMTJ-based STT-MRAM a promising candidate to replace conventional semiconductor-based cache memory for the next-generation of low-power microprocessors with on-chip non-volatility.     

An experimental nanosecond Josephson 1K RAM using 5-µm Pb-alloy technology

A high-speed fully decoded Josephson 1K RAM has been designed and tested. Several bits of the 1K RAM were successfully operated with a typical read access time of 3.3 ns and associated power dissipation of 2.0 mW. The chip, containing about 10 000 Josephson junctions, was fabricated using 5-µm Pb-alloy technology, including a novel junction oxide barrier formation technique. A nondestructive readout (NDRO) Josephson ring cell operating with all current levels equal and an on-chip timing circuit for read/write operations were employed.     

Board-level built-in self-repair method of RAM

This paper describes the method of built-in self-repairing of RAM on board, designs hardware circuit, and logic for the RAM’s faults self-repairing system based on FPGA. The key technology is that it utilizes FPGA to test RAM according to some algorithm to find out failure memory units and replace the faulty units with FPGA. Then it can build a memory that has no fault concern to external controller, and realizes the logic binding between external controller and RAM. Micro Controller Unit (MCU) can operate external RAM correctly even if RAM has some fault address units. Conventional MCS-51 is used to simulate the operation of MCU operating external memory. Simulation shows FPGA can complete the faulty address units’ mapping and MCU can normally read and write external RAM. This design realizes the RAM’s built-in self-repairing on board.     

CMOS static RAM chip with high-speed optical read and write

We present the first demonstration of a dense VLSI RAM technology with high-speed optical read and optical write capability. The CMOS-based Static-RAM technology is capable of parallel optical access with read/write speeds limited by the native RAM access times. We fabricated a 2/spl times/2 mm optoelectronic-VLSI test chip incorporating 800-b storage and 200 optical I/O based on the hybrid integration of GaAs-AlGaAs MQW modulators on CMOS. Results from the photonic-SRAM test-chip confirm 6.2 ns read and 8-ns write capability.     

Effective march algorithms for testing single-order addressed memories

Today's commonly used macro generators provide for read/write memories of type SRAM, Register File, Multi-Port RAM, Single-Order Addressed Memory (e.g. FIFO), CAM (Content Addressable Memory), etc. In addition to automatically generating the required momory, the appropriate test, which may be applied externally or internally as a BIST, has to be determined.Current literature provides tests for most memory types; however, tests for single-order addressed (SOA) memories, whereby the address can only change in one direction (e.g. from address 0 ton-1) have not been published yet. SOA memories are used in FIFOs and in applications where the BIST area overhead and/or speed penalty for normal (dual) order addressing are not acceptable.This article illustrates the testing problems and presents a family of march algorithms optimized for testing SOA memories.     

Write Disturbance Modeling and Testing for MRAM

The magnetic random access memory (MRAM) is considered one of the potential candidates that will replace current on-chip memories (RAM, EEPROM, and flash memory) in the future. The MRAM is fast and does not need a high supply voltage for read/write operations, and is compatible with the CMOS technology. It can also endure almost unlimited read/write cycles. These combined advantages of RAM and flash memory make it a potential choice for SOC. In this paper, we present the write disturbance fault (WDF) model for MRAM, i.e., a fault that affects the data stored in the MRAM cells due to excessive magnetic field during the write operation. We also construct the SPICE macro model for the magnetic tunneling junction (MTJ) device of the toggle MRAM to obtain circuit simulation results. We then present an MRAM fault simulator called RAMSES-M, based on which we derive the shortest test for the proposed WDF model. The test is shown to be better and more robust as compared with the conventional March C-test algorithm. We also present a March 17 N diagnosis algorithm for identifying WDF. A 1 Mb MRAM chip has been designed and fabricated using a CMOS-based 0.18-mum technology. The proposed WDF model is justified by chip measurement results, with the march test results reported. Finally, specific MRAM fault behavior and test issues are discussed.     

High-speed split-emitter I/SUP 2/L/MTL memory cell

Describes a novel circuit/device approach that overcomes the performance drawback of the injection-sensed I/SUP 2/L/MTL memory cell cited in a 16-kbit static MTL RAM (see IEEE ISSCC Dig. Tech. Papers, p.222-4, 1980). As a result, a compact memory cell with extremely low DC standby power in the nanowatt range and with read/write times below 5 ns is achieved. This has been verified by experimental investigations on small test arrays. They have been fabricated with an advanced process featuring a p-polysilicon-base self-alignment scheme and a double-diffused p-n-p structure. In addition, computer circuit simulations have been performed that show the read delay sensitivities in large arrays. Based on these results, an access time of less than 25 ns is projected for a 16-kbit MTL RAM.     

Hybrid crossbar architecture for a memristor based cache

This paper describes a new memristor crossbar architecture that is proposed for use in a high density cache design. This design has less than 10% of the write energy consumption than a simple memristor crossbar. Also, it has up to 3 times the bit density of an STT-MRAM system and up to 11 times the bit density of an SRAM architecture. The proposed architecture is analyzed using a detailed SPICE analysis that accounts for the resistance of the wires in the memristor structure. Additionally, the memristor model used in this work has been matched to specific device characterization data to provide accurate results in terms of energy, area, and timing. The proposed memory system was analyzed by modeling two different devices that vary in resistance range and switching time. This system does not require that the memristor devices have inherent diode effects which limit alternate current paths. Therefore this system is capable of utilizing a much broader class of devices.An architectural analysis has also been completed that shows how the memory system may perform as a cache memory. A hybrid cache structure was used to alleviate the long write latencies of memristor devices. This approach consisted of the tag array being made of SRAM cells while the data array was made of the memristor circuit proposed. This hybrid scheme allows multiple reads and writes to concurrently access different sub-arrays within a cache. The performance of these novel memristor based caches was compared to SRAM and STT-MRAM based caches through detailed simulations. The results show that the memristor caches are denser and allow better performance along with lower system power when compared to the STT-MRAM and SRAM caches.     

A four-phase handshaking asynchronous static RAM design forself-timed systems

The motivation of designing asynchronous memory arises from the recent development of asynchronous processors. As different from the conventional design, the proposed asynchronous static RAM can: (1) communicate with other asynchronous systems based on a four-phase handshaking control protocol; and (2) generate the read/write completion signals with increased average speed by the variable bit-line load concept. The techniques investigated include (1) dual-rail voltage sensing completion detection for read operation and (2) multiple delays completion generation for write operation. In this paper, the performances of these techniques are evaluated for 1 Mb memory with four regions of bit-line segmentation. The simulated and measured results are presented and compared     

Spin-transfer torque RAM technology: Review and prospect

Non-volatile RAM (NV-RAM) enables instant-on/off computing, which drastically reduces power consumption. One of the most promising candidates for NV-RAM technology is the spin-transfer torque RAM (SPRAM) based on magnetic tunnel junction (MTJ) device technology. This paper reviews the development of MTJ device technology and formulates considerations regarding its memory application, including SPRAM memory cell structure and operation, write voltage limitation, and thermal stability. At the circuit level, a disruptive read operation for future large integration scale is described. A 4F² memory cell and a multi-bit cell approach are also presented. Finally, the potential value of instant-on/off computing through NV-RAM and its impact are explored.     

新型4T SRAM的读辅助电路设计

SoC芯片的很大一部分面积被存储器占据,而静态随机存储器SRAM为主要部分,因此高密度的SRAM研究引起更多重视。随着半导体工艺的不断发展,SRAM存储器的读写性能愈发重要。研究和分析了两种高密度、低功耗、高速的SRAM读辅助电路,即降低字线电压电路和增大供电电压电路。针对存储密度提升的4T SRAM,通过使用读辅助电路,增强了数据读取的稳定性,同时可以保证SRAM的数据写能力。在55 nm CMOS工艺条件下,相对传统6T SRAM,4T存储单元的面积减小20%。仿真结果表明,通过在外围电路中设计辅助电路,4T SRAM的读稳定性改善了134%。     

基于NVMe的超高速多通道遥感相机模拟源设备研制

针对当前遥感卫星电荷耦合器件(CCD)相机幅宽越来越大,速率越来越高,现有相机模拟源设备数据输出带宽不足的问题,提出并实现了一种基于非易失性存储器Express(NVMe)的超高速多通道遥感相机模拟源设备。该设备利用现场可编程逻辑门阵列(FPGA)实现4组NVMe SSD主机控制器,完成对固态硬盘(SSD)的读写操作;同时利用DMA控制器读取DDR4中缓存数据,数据经封装处理后通过光纤接口输出。实验结果表明：NVMe主机控制器的写平均速率可以达到1.7 GBps,读平均速率达到3.2 GBps。模拟源系统整体存储容量8 TB,对外输出带宽高达80 Gbps,支持8路光纤接口输出。该模拟源具有较强的稳定性及良好的可扩展性,已成功应用在某遥感卫星CCD相机模拟源系统中,为数传等设备的测试以及调试提供了充分保障。     

一种基于位交错结构的亚阈值10管SRAM单元

提出了一种基于位交错结构的亚阈值10管SRAM单元,实现了电路在超低电压下能稳定地工作,并降低了电路功耗。采用内在读辅助技术消除了读干扰问题,有效提高了低压下的读稳定性。采用削弱单元反馈环路的写辅助技术,极大提高了写能力。该10管SRAM单元可消除半选干扰问题,提高位交错结构的抗软错误能力。在40 nm CMOS工艺下对电路进行了仿真。结果表明,该10管SRAM单元在低压下具有较高的读稳定性和优异的写能力。在0.4 V工作电压下,该10管SRAM单元的写裕度为传统6管单元的14.55倍。     

嵌入式DSP访问片外SDRAM的低功耗设计研究

为了降低DSP外部SDRAM存储系统的功耗,针对DSP访问片外SDRAM的功耗来源特点,提出了基于总线利用率动态监测的读写归并方案。该方案动态监测外部存储器接口(EMIF)总线的利用率,根据总线利用率的不同选择开放的页策略、封闭的页策略或休眠模式;设计了简化的指令Cache(I-Cache),采用块读的方法取指令;设计了写后数据缓冲区,由EMIF对同一行的读写进行归并。经计算,根据EMIF总线利用率的不同(10%~40%),该方案相比单纯采用开放的页策略,功耗可减少5%~20%左右。