A 300-MHz 25-/spl mu/A/Mb-leakage on-chip SRAM module featuring process-variation immunity and low-leakage-active mode for mobile-phone application processor

An on-chip 1-Mb SRAM suitable for embedding in the application processor used in mobile cellular phones was developed. This SRAM supports three operating modes - high-speed active mode, low-leakage low-speed active mode, and standby mode - and uses a subdivisional power-line control (SPC) scheme. The combination of three operating modes and the SPC scheme realizes low-power operation under actual usage conditions. It operates at 300 MHz, with leakage of 25 /spl mu/A/Mb in standby mode, and 50 /spl mu/A/Mb at the low-leakage active mode. This SRAM also uses a self-bias write scheme that decreases of minimum operating voltage by about 100 mV.     

A 21-ns 32 K/spl times/8 CMOS static RAM with a selectively pumped p-well array

The design and performance of a 32 K/spl times/8-b CMOS static RAM (SRAM) are presented. The design features a selectively pumped p-well array. Using this array technology, high-impedance polysilicon resistor loads can be used to reduce the array standby current by three orders of magnitude, and the device characteristics can be optimized to achieve high speed in the peripheral circuits. A unique divided-word-line architecture with shared sense amplifiers is used to achieve high-speed read operation. The read speed is further enhanced by a novel quasistatic equalization that minimizes peak current. A high-speed write circuit with a write-to-read transition detection is used to achieve a fast write operation. An advanced 1.2-/spl mu/m double-level-metal CMOS technology was used to fabricate the devices. The access time is 21 ns and the active power is 330 mW at 22 MHz.     

A 0.5-V 25-MHz 1-mW 256-kb MTCMOS/SOI SRAM for solar-power-operated portable personal digital equipment - sure write operation by using step-down negatively overdriven bitline scheme

Multithreshold-voltage CMOS (MTCMOS) technology has a great advantage in that it provides high-speed operation with low supply voltages of less than 1 V. A logic gate with low-V/sub th/ MOSFETs has a high operating speed, while a low-leakage power switch with a high-V/sub th/ MOSFET eliminates the off-leakage current during sleep time. By using MTCMOS circuits and silicon-on-insulator (SOI) devices, the authors have developed a 256-kb SRAM for solar-power-operated digital equipment. A double-threshold-voltage MOSFET (DTMOS) is adopted for the power switch to further reduce the off leakage. As regards the SRAM core design, we consider a hybrid configuration consisting of high-V/sub th/ and low-V/sub th/ MOSFETs (that is, multi-V/sub th/ CMOS). A new memory cell with a separate read-data path provides a larger readout current without degrading the static noise margin. A negatively overdriven bitline scheme guarantees sure write operation at ultralow supply voltages close to 0.5 V. In addition, a charge-transfer amplifier integrated with a selector and data latches for intrabus circuitry are installed to enhance the operating speed and/or reduce power dissipation. A 32K-word /spl times/ 8-bit SRAM chip, fabricated with the 0.35-/spl mu/m multi-V/sub th/ CMOS/SOI process, has successfully operated at 25 MHz under typical conditions with 0.5-V (SRAM core) and 1-V (I/O buffers) power supplies. The power dissipation during sleep time is less than 0.4 /spl mu/W and that for 25-MHz operation is 1 mW, excluding that of the I/O buffers.     

A 200-mW GaAs 1 K SRAM with 2-ns cycle time

A very high-speed and low-power 1024/spl times/1 SRAM has been designed and fabricated using a normally-off recessed-gate FET technology. Minimum gate length is 0.7 /spl mu/m. A minimum access time of 1.4 ns has been obtained with a power dissipation of 210 mW. The memory cell area is 1197 /spl mu/m/SUP 2/ and the chip size is 1.91/spl times/2.21 mm/SUP 2/. The output voltage swing across a 50-/spl Omega/ load is 700 mV. The maximum simulated yield for 1 K SRAMs is discussed theoretically. A mean standard deviation in threshold voltage less than 15 mV is required to obtain 100% design yield. The SRAM has been shown to be fully operational using the march and checkerboard tests and exhibits read and write cycle times of 2 ns.     

SRAM design on 65-nm CMOS technology with dynamic sleep transistor for leakage reduction

A 70-Mb SRAM is designed and fabricated on a 65-nm CMOS technology. It features a 0.57-/spl mu/m/sup 2/ 6T SRAM cell with large noise margin down to 0.7 V for low-voltage operation. The fully synchronized subarray contains an integrated leakage reduction scheme with dynamically controlled sleep transistor. SRAM virtual ground in standby is controlled by programmable bias transistors to achieve good voltage control with fine granularity under process skew. It also has a built-in programmable defect "screen" circuit for high volume manufacturing. The measurements showed that the SRAM leakage can be reduced by 3-5/spl times/ while maintaining the integrity of stored data.     

A leakage reduction methodology for distributed MTCMOS

Multithreshold CMOS (MTCMOS) circuits reduce standby leakage power with low delay overhead. Most MTCMOS designs cut off the power to large blocks of logic using large sleep transistors. Locally distributing sleep devices has remained less popular even though it has several advantages described in this paper. However, locally placed sleep devices are only feasible if sneak leakage currents are prevented. This paper makes two contributions to leakage reduction. First, we examine the causes of sneak leakage paths and propose a design methodology that enables local insertion of sleep devices for sequential and combinational circuits. A set of design rules allows designers to prevent most sneak leakage paths. A fabricated 0.13-/spl mu/m, dual V/sub T/ test chip employs our methodology to implement a low-power FPGA architecture with gate-level sleep FETs and over 8/spl times/ measured standby current reduction. Second, we describe the implementation and benefits of local sleep regions in our design and examine the interfacing issues for this technique. Local sleep regions reduce leakage in unused circuit components at a local level while the surrounding circuits remain active. Measured results show that local sleep regions reduce leakage in active configurable logic blocks (CLBs) on our chip by up to 2.2/spl times/ (measured) based on configuration.     

Locally switched and limited source-body bias and other leakage reduction techniques for a low-power embedded SRAM

A low-power embedded SRAM for a large range of applications has been implemented in a standard digital 0.18-/spl mu/m process. The leakage current in the cells is reduced by using a source-body bias not exceeding the value that guaranties safe data retention, and less leaking nonminimum length transistors. Locally short-circuiting this bias, speed and noise margin loss in active mode is avoided, especially for low supply voltages. The bias is generated internally at the carefully designed equilibrium between cell, switch, and diode limiter leakages averaged over the array. The leakage of the full SRAM, including an optimized periphery, is reduced more than 20 times. Used in an industrial RF transceiver, the measurements confirm its performances.     

A Low-Power Embedded SRAM for Wireless Applications

This paper introduces a novel ultra-low-power SRAM. A large power reduction is obtained by the use of four new techniques that allow for a wider and better trade-off between area, delay and active and passive energy consumption for low-power embedded SRAMs. The design targets wireless applications that require a moderate performance at an ultra-low-power consumption. The implemented design techniques consist of a more efficient memory databus, the exploitation of the dynamic read stability of SRAM cells, a new low-swing write technique and a distributed decoder. An 8-KB 5T SRAM was fabricated in a 0.18-mum technology. The measurement results confirm the feasibility and the usefulness of the proposed techniques. A reduction of active power consumption with a factor of 2 is reported as compared to the current state of the art. The results are generalized towards a 32-KB SRAM.     

A Low-Power SRAM Using Bit-Line Charge-Recycling

Low-power SRAM design is crucial since it takes a large fraction of total power and die area in high-performance processors. Reducing voltage swing of the bit-line is an effective way to save the power dissipation in write cycles. Voltage swing reduction of bit-lines is, however, limited due to possible write-failures. We propose a new low-power SRAM using bit-line charge recycling (CR-SRAM) for the write operation. In the proposed write scheme, differential voltage swing of a bit-line is obtained by recycled charge from its adjacent bit-line capacitance, instead of the power line. Applying such a charge recycling technique to the bit-line significantly reduces write power. A test chip with 32 Kbits (256 rows x 128 columns) is fabricated and measured in 0.13 mum CMOS to demonstrate operation of the proposed SRAM. Measurement results show 88% reduction in total power during write cycles compared to the conventional SRAM (CON-SRAM) at V_DD = 1.5 V and f = 100 MHz.     

PVT-aware leakage reduction for on-die caches with improved read stability

Effectiveness of previous SRAM leakage reduction techniques vary significantly as the leakage variation gets worse with process and temperature fluctuation. This paper proposes a simple circuit technique that adaptively trades off overhead energy for maximum leakage savings under severe leakage variations. The proposed run-time leakage reduction technique for on-die SRAM caches considers architectural access behavior to determine how often the SRAM blocks should enter a sleep mode. A self-decay circuit generates a periodic sleep pulse with an adaptive pulse period, which puts the SRAM array into a sleep mode more frequently at high leakage conditions (fast process, high temperature) and vice versa. An 0.18-/spl mu/m 1.8-V 16-kbyte SRAM testchip shows 94.2% reduction in SRAM cell leakage at a performance penalty less than 2%. Measurement results also indicate that our proposed memory cell improves SRAM static noise margin by 25%.     

Novel low-power and stable SRAM cells for sub-threshold operation at 45 nm

In scaled technologies with lower supply voltage, conventional Static Random Access Memory (SRAM) cell suffers from unsuccessful read & write operation due to high off state current in sub-threshold region at nanometre technologies. This work proposes new functional low-power designs of SRAM cells with 7, 8, 9 and 12 transistors which operate at only 0.4V power supply in sub-threshold operation at 45 nm technology. Stability analysis is carried out using static noise margins as well as N-curve cell stability metrics. For performance measurement, read/write access time and leakage power consumption in hold mode are analysed. The comparison with published designs shows that two new proposed designs namely M8T, MPT8T have 30% less leakage power consumption along with 2× read stability, 2× write ability, more than 60% faster read & write operation.     

Single Miller capacitor frequency compensation technique for low-power multistage amplifiers

Due to the rising demand for low-power portable battery-operated electronic devices, there is an increasing need for low-voltage low-power low-drop-out (LDO) regulators. This provides motivation for research on high-gain wide-bandwidth amplifiers driving large capacitive loads. These amplifiers serve as error amplifiers in low-voltage LDO regulators. Two low-power efficient three-stage amplifier topologies suitable for large capacitive load applications are introduced here: single Miller capacitor compensation (SMC) and single Miller capacitor feedforward compensation (SMFFC). Using a single Miller compensation capacitor in three-stage amplifiers can significantly reduce the total capacitor value, and therefore, the overall area of the amplifiers without influencing their stability. Pole-splitting and feedforward techniques are effectively combined to achieve better small-signal and large-signal performances. The 0.5-/spl mu/m CMOS amplifiers, SMC, and SMFFC driving a 25-k/spl Omega///120-pF load achieve 4.6-MHz and 9-MHz gain-bandwidth product, respectively, each dissipates less than 0.42 mW of power with a /spl plusmn/1-V power supply, and each occupies less than 0.02 mm/sup 2/ of silicon area.     

A low-power SRAM using hierarchical bit line and local sense amplifiers

This paper proposes a low power SRAM using hierarchical bit line and local sense amplifiers (HBLSA-SRAM). It reduces both capacitance and write swing voltage of bit lines by using the hierarchical bit line composed of a bit line and sub-bit lines with local sense amplifiers. The HBLSA-SRAM reduces the write power consumption in bit lines without noise margin degradation by applying a low swing signal to the high capacitive bit line and by applying a full swing signal to the low capacitive sub-bit line. The HBLSA-SRAM reduces the swing voltage of bit lines to V/sub DD//10 for both read and write. It saves 34% of the write power compared to the conventional SRAM. An SRAM chip with 8 K/spl times/32 bits is fabricated in a 0.25-/spl mu/m CMOS process. It consumes 26 mW read power and 28 mW write power at 200 MHz with 2.5 V.     

A 30 ns 16K/spl times/1 fully static RAM

A fully static 16K/spl times/1 random access memory (SRAM) with significantly improved speed is discussed. Design innovations using conservative 2.5 /spl mu/m transistors and state-of-the-art double level poly (DLP) scaled NMOS technology were utilized to accomplish 30 ns address and chip select access times with an active power of 550 mW and standby power of 75 mW. A cost effective DLP process was developed using `shared' contacts in the cell. These `shared' contacts utilize second level poly to provide connection between the first poly level and moat, reduced the number of contacts per cell to four. The DPL cell size is 1.6 mil/SUP 2/ (1000 /spl mu/m/SUP 2/) which yields a bar size of 158/spl times/264 mil/SUP 2/ (4.0/spl times/6.7 mm/SUP 2/). In this fully static design a novel architecture was used to power down half of the X-decoders in the active mode using the AO address buffer signals. This technique allowed the use of power saved in the X-decoder to be distributed throughout the circuit to improve overall access times. One of the other major speed improvements came from utilizing column sense amps. The use of the column sense amp improves the overall speed by more than 20 percent. A write cycle of 30 ns has been achieved with a typical write pulse width of 10 ns.     

The design and analysis of a Miller-divider-based clock generator for MBOA-UWB application

A Miller-divider-based clock generator is proposed for Multi-Band OFDM Alliance (MBOA) ultrawideband (UWB) application. Employing closed-loop operation, the clock generator can produce three different carrier frequencies with negligible in-band spurs. The settling time of the proposed clock generator is analyzed based on a linear feedback system. A transistor sizing optimization technique for active inductors with a current-reusing technique is used to achieve low-power operation and area saving. Fabricated in a 0.18-/spl mu/m technology, the clock generator achieves less than 9.5-ns settling time while dissipating less than 47 mW from a 1.8-V power supply.     

90% write power-saving SRAM using sense-amplifying memory cell

This paper describes a low-power write scheme which reduces SRAM power by 90% by using seven-transistor sense-amplifying memory cells. By reducing the bitline swing to V/sub DD//6 and amplifying the voltage swing by a sense-amplifier structure in a memory cell, the charging and discharging component of the power of the bit/data lines is reduced. A 64-kb test chip has been fabricated and correct read/write operation has been verified. It is also shown that the scheme can also have the capability of leakage power reduction with small modifications. Achievable leakage power reduction is estimated to be two orders of magnitude from SPICE simulation results.     

A dual voltage-frequency VLSI chip for image watermarking in DCT domain

In this brief, we present a new VLSI architecture that can insert invisible or visible watermarks in images in the discrete cosine transform domain. The proposed architecture incorporates low-power techniques such as dual voltage, dual frequency, and clock gating to reduce the power consumption and exploits pipelining and parallelism extensively in order to achieve high performance. The supply voltage level and the operating frequency are chosen for each module so as to maintain the required bandwidth and throughput match among the different modules. A prototype VLSI chip was designed and verified using various Cadence and Synopsys tools based on TSMC 0.25-/spl mu/m technology with 1.4 M transistors and 0.3 mW of estimated dynamic power.     

An embedded DRAM with a 143-MHz SRAM interface using a sense-synchronized read/write

This paper describes a 4-Mb embedded DRAM macro using novel fast random cycle architecture with sense-synchronized read/write (SSR/SSW). The test chip has been fabricated with a 0.15-/spl mu/m logic-based embedded DRAM process and the 1.5-V 143-MHz no-wait row random access operation has been confirmed. Data retention power is suppressed to 92 /spl mu/W owing to the hierarchical power supply and SSR. The macro size is 4.59 mm/sup 2/. The cell occupation ratio of the macro is 46%, which is the same as that of a conventional embedded DRAM macro. The macro size and the data retention power are 30% and 4.6%, respectively, of a 4-Mb embedded SRAM macro fabricated by an identical process.     

A 25-ns 1-Mbit CMOS SRAM with loading-free bit lines

A 128 K/spl times/8-b CMOS SRAM is described which achieves a 25-ns access time, less than 40-mA active current at 10 MHz, and 2-/spl mu/A standby current. The novel bit-line circuitry (loading-free bit line), using two kinds of NMOSFETs with different threshold voltages, improves bit-line signal speed and integrity. The two-stage local amplification technique minimizes the data-line delay. The dynamic double-word-line scheme (DDWL) allows the cell array to be divided into 32 sections along the word-line direction without a huge increase in chip area. This allows the DDWL scheme to reduce the core-area delay time and operating power to about half that of other conventional structures. A double-metal 0.8-/spl mu/m twin-tub CMOS technology has been developed to realize the 5.6/spl times/9.5-/spl mu//SUP 2/ cell size and the 6.86/spl times/15.37-mm/SUP 2/ chip size.     

A complete single-chip GPS receiver with 1.6-V 24-mW radio in 0.18-/spl mu/m CMOS

We have developed a complete single-chip GPS receiver using 0.18-/spl mu/m CMOS to meet several important requirements, such as small size, low power, low cost, and high sensitivity for mobile GPS applications. This is the first case in which a radio has been successfully combined with a baseband processor, such as SoC, in a GPS receiver. The GPS chip, with a total size of 6.3 mm /spl times/ 6.3 mm, contains a 2.3 mm /spl times/ 2.0 mm radio part, including RF front end, phase-locked loops, IF functions, and 500 K gates of baseband logic, including mask ROM, SRAM, and dual port SRAM . It is fabricated using 0.18-/spl mu/m CMOS technology with a MIM capacitor and operates from a 1.6-2.0-V power supply. Experimental results show a very low power consumption of, typically, 57 mW for a fully functional chip including baseband, and a high sensitivity of -152dBm. Through countermeasures against substrate coupling noise from the digital part, the high sensitivity was successfully achieved without any external low-noise amplifier.