A 1.0-V 230-MHz column access embedded DRAM for portable MPEGapplications

This paper describes a 32-Mb embedded DRAM macro fabricated using 0.13-μm triple-well 4-level Cu embedded DRAM technology, which is suitable for portable equipment of MPEG applications. This macro can operate 230-MHz random column access even at 1.0-V power supply condition. The peak power consumption is suppressed to 198 mW in burst operation. The power-down standby mode, which suppresses the leakage current consumption of peripheral circuitry, is also prepared for portable equipment. With the collaboration of array circuit design and the fine Cu metallization technology, macro size of 18.9 mm² and cell efficiency of 51.3% are realized even with dual interface and triple test functions implemented     

256-Mb DRAM circuit technologies for file applications

256-Mb DRAM circuit technologies characterized by low power and high fabrication yield for file applications are described. The newly proposed and developed circuits are a self-reverse-biasing circuit for word drivers and decoders to suppress the subthreshold current to 3% of the conventional scheme, and a subarray-replacement redundancy technique that doubles chip yield and consequently reduces manufacturing costs. An experimental 256-Mb DRAM has been designed and fabricated by combining the proposed circuit techniques and a 0.25-μm phase-shift optical lithography, and its basic operations are verified. A 0.72-μm² double-cylindrical recessed stacked-capacitor (RSTC) cell is used to ensure a storage capacitance of 25 fF/cell. A typical access time under a 2-V power supply voltage was 70 ns. With the proper device characteristics, the simulated performances of the 256-Mb DRAM operating with a 1.5-V power supply voltage are a data-retention current of 53 μA and an access time of 48 ns     

A 1-V BiCMOS rail-to-rail amplifier with n-channel depletion modeinput stage

A BiCMOS rail-to-rail operational amplifier capable of operating from supply voltages as low as 1 V is presented. The folded cascode input stage uses an nMOS depletion mode differential pair to provide rail-to-rail common mode voltage range while typically requiring only 40 fA of input bias current. The bipolar transistor differential-to-single-ended conversion network employs a low-voltage base current cancellation technique which provides high input stage voltage gain from a l-V supply yet allows a 3-V/μs slew rate capability. The bipolar transistor output stage uses a low-voltage translinear loop which maintains a low impedance signal path to the output common emitter power devices. This circuit topology enables the amplifier to achieve a 4-MHz bandwidth with 60° of phase margin. The output voltage can swing to within 50 mV of each supply rail. An “on-demand” base current boost technique will be presented which can provide up to 50 mA of output drive capability from a 5-V supply, yet consumes only a few microamps when the output is in the quiescent state. A low voltage level shift technique will be described which uses an n-channel depletion mode source follower to provide isolation between the input and output stages     

A 0.25-μm, 600-MHz, 1.5-V, fully depleted SOI CMOS 64-bitmicroprocessor

A 0.25-μm, four-layer-metal, 1.5-V, 600-MHz, fully depleted (FD) silicon-on-insulator (SOI) CMOS 64-bit ALPHA1 microprocessor integrating 9.66 million transistors on a 209-mm² silicon die has been developed leveraging the existing bulk design. FD-SOI technology is used because it has better immunity for dynamic leakage current than partially depleted SOI in high speed dynamic circuits without body contact. C-V characteristics of metal-oxide-silicon-oxide-silicon with and without source-drain junctions are described to explain the behavior of FD-SOI transistor. Race, speed, and dynamic stability have been simulated to reassure the circuit operation. Key process features are shallow trench isolation, 4-nm gate oxide, 30-nm co-silicide, 46-nm silicon film, and 200-nm buried oxide. The FD-SOI microprocessor runs 30% faster than that of bulk, and it passes the reliability and system test     

High-performance large-area InGaAs metal-semiconductor-metalphotodetectors

The fabrication and characteristics of high-performance large-area InP:Fe/InGaAs:Fe/InP:Fe metal-semiconductor-metal (MSM) photodetectors are reported. With a 350-μm×350-μm active area, the detectors offer 900-MHz electrical bandwidth and 1.7-pF capacitance at 10-V bias. The respective dark current density is 20 pA/μm², an the CW responsivity is 0.4 A/W at 1.3-μm wavelength. The detectors are therefore ideally suited for applications in the long-wavelength range that require a large detection area and, at the same time, a high bandwidth and low capacitance     

A 433-MHz 64-b quad-issue RISC microprocessor

A quad-issue custom VLSI microprocessor is described. This microprocessor implements the Alpha architecture and achieves an estimated performance of 13.3 SPECint9S and 18.4 SPECfp95 at 433 MHz. The 9.6 million transistor die measures 14.4 mm×14.5 mm, and is fabricated in a 0.35-μm, four-metal layer CMOS process. This chip dissipates less than 25 W at 433 MHz using a 2.0 V internal power supply. The design was leveraged from a prior 300-MHz, 3.3-V, 0.50-μm CMOS design. It includes several significant architectural enhancements and required circuit solutions for operation at 2.0 V. The chip will operate at nominal internal power supply voltages up to 2.5 V allowing improved performance at the cost of increased power consumption. At 2.5 V, the chip operates at 500 MHz and delivers 15.4 SPECint95 (est) and 21.1 SPECfp95 (est). This paper describes the chip implementation details and the strategy for efficiently migrating the existing design to the 0.35-μm technology     

A current-sensed high-speed and low-power first-in-first-out memoryusing a wordline/bitline-swapped dual-port SRAM cell

First-in-first-out (FIFO) data storages are in great demand for telecommunication LSIs. This paper presents high-speed and low-power CMOS memory techniques specialized for FIFO operation. A size-configurable architecture using the tile methodology is employed to customize the word counts and/or data bits with a. short time of less than 30 min. Four flag bits are introduced to inform the internal state of FIFO memories. To obtain a higher operating speed, an SRAM-like memory cell with current-sense readout is used. The critical-path delay of the Gray-code up/down counter, indicating the stored data volume, is shortened to 6.0 ns (66%) by using a double-rail single-stage XOR circuit. As to the low-power techniques, a wordline/bitline-swapped dual-port memory-cell architecture is proposed to cut off the static power-supply current of unselected columns. By using the hidden blanket-precharged bitline scheme, the power dissipation of the writing circuitry is minimized without degrading the operating speed. A new data-driven gated-shift-pulse architecture is also proposed to reduce the power dissipation of shift-register-type address pointers (1.5 mW at 100 MHz). A 2K-words × 8-bits FIFO memory test chip, fabricated with a 0.6-μm CMOS process (a short effective channel length of 0.35 μm is available for both the nMOS and pMOS), has demonstrated the 140-MHz operation at a typical 3.3-V power supply. The power dissipation in standby is less than 0.1 μW and that at 100-MHz dual-port operation with single fan-out loads is in the range from 28 mW (in the best case with the M-scan test pattern) to 46 mW (in the worst case with the checkerboard test pattern)     

A 600-MHz superscalar floating-point processor

The floating-point unit of a 600-MHz, out-of order, superscalar RISC Alpha microprocessor is described. The unit achieves 59 SpecFP95 and can transfer register data at up to 9.6 GB/s. It has two independent pipelines for multiply and add/subtract operations, with iterative divide and square-root circuits, and is fabricated in a 2.2-V, 0.35-μm CMOS process     

A micropower class-AB CMOS log-domain filter for DECT applications

This paper presents a micropower second-order low-pass filter using the log-domain principle and integrated in a 0.35-μm CMOS process. It has been designed as an antialiasing filter for a DECT transceiver with a 45-kHz nominal cutoff frequency. The circuit uses transistors biased in weak inversion without requiring separate wells. It operates at 1.5-V supply voltage and its current consumption is 8 μA in idle mode. The log-domain filter is implemented with an on-chip conditioner which allows class-AB operation. It can process input currents at 5 kHz that are 25 times larger than the 200-nA bias current. Measurements up to 500 times the bias current have been done, since at 1 kHz the input current is only limited by the supply voltage     

A low-power 170-MHz discrete-time analog FIR filter

A 170-MHz analog finite impulse response (FIR) filter operating from a single 3.3-V supply is described. The design has been fabricated in the HP 1.2-μm CMOS process and has an area of 2.35 mm by 1.97 mm including bonding pads. This 9-tap filter dissipates 70 mW when operating at 170 MHz. The multipliers are implemented using multiplying digital-to-analog converters (MDAC's) with 6-b resolution     

A 1.2- to 3.3-V wide voltage-range/low-power DRAM with acharge-transfer presensing scheme

A charge-transfer presensing scheme (CTPS) for 0.8-V array operation with a 1/2 V_cc bit-line precharge achieves a five times larger readout voltage and 40% improvement in sensing speed compared with conventional sensing schemes. Operation over a 1.2- to 3.3-V range is achieved. A nonreset row block control scheme (NRBC) for power-consumption improvement in data-retention mode is proposed which decreases the charge/discharge number of the row block control circuit. By combining CTPS and NRBC, the data-retention current is reduced by 75%     

A 5-GHz-band multifunctional BiCMOS transceiver chip for GMSKmodulation wireless systems

This paper reports the first multifunctional 0.4-μm BiCMOS-based transceiver chip developed for 5-GHz-band Gaussian minimum-shift keying modulation wireless systems. The chip integrates a low-noise radio-frequency amplifier, a down-mixer, and an intermediate-frequency (IF) amplifier in the down-converter path; an IF amplifier, a limiter, an up-mixer, and a buffer amplifier in the up-converter path; and a frequency doubler and a local oscillator amplifier in the local oscillator path. The chip featuring gain attenuation as well as standby mode operation uses a single 2.6-5.2-V bias voltage and dissipates 56 mW in receive mode and 66 mW in transmit mode. The transceiver chip size is 3.0×2.4 mm²     

A fundamental limitation for bipolar transistor scaling

A microscopic model of minority-carrier diffusion in a heavily doped emitter is proposed. Monte Carlo simulation demonstrates that statistical fluctuation in the base current is one of the fundamental limitations in high-speed applications of scaled bipolar transistors. For the transistor presently investigated, with 5.0-μm² emitter area, 0.1-μm junction depth, 8.5-ps measurement time, and 0.75-V emitter/base bias, the base current deviation is 43%. This sets up the maximum operating frequency for the transistor. More lightly doped emitters (such as for heterojunction bipolar transistors) will relax this limitation, but at a cost of increased contact resistance, especially when poly-emitters are utilized. Increasing the emitter/base bias will also make the base current rate more deterministic, but the other limitations such as power dissipation and contact resistance will become more obvious     

A 0.5-V MTCMOS/SIMOX logic gate

This paper proposes a multithreshold CMOS (MTCMOS) circuit that uses SIMOX process technology. This MTCMOS/SIMOX circuit combines fully depleted low-threshold CMOS logic gates and partially depleted high-threshold power-switch transistors. The low-threshold CMOS gates have a large noise margin for fluctuations in operating temperature in addition to high-speed operation at the low supply voltage of 0.5 V. The high-threshold power-switch transistor in which the body is connected to the gate through the reverse-diode makes it possible to obtain large channel conductance in the active mode without any increase of the leakage current in the sleep mode. The effectiveness of the MTCMOS/SIMOX circuit is confirmed by an evaluation of a gate-chain test element group (TEG) and an experimental 0.5-V, 40-MHz, 16-b ALU, which were designed and fabricated with 0.25-μm MTCMOS/SIMOX technology     

High-voltage-tolerant I/O buffers with low-voltage CMOS process

This paper presents high-voltage-tolerant I/O buffer designs for a 1.9-V external cache interface and a 3.3-V system interface using 1.9-V MOS transistors in a 0.21-μm process with 40-Å gate-oxide thickness. Various circuit techniques are used for 1.9- and 3.3-V I/O buffers to ensure that the voltage across the gate oxide of every MOS element is below specified limits of 2.2 V for transient (short duty cycle) and 1.9 V for steady state. Only one PMOS pullup driver transistor between the bond pad and the power supply, and one NMOS pulldown driver transistor between the bond pad and ground, are used for the 1.9-V I/O buffer design, while cascoded MOS transistors between the bond pad and power supply or ground terminals are used for the 3.3-V I/O buffer design. The primary design goal is to ensure the reliability of MOS elements by avoiding excessive gate oxide stress due to high electric fields. However, due to differences in requirements for speed, power-supply voltage, and tristate leakage current, completely different circuit techniques have been used for the two designs. Both of the designs have been successfully implemented in a 400-MHz UltraSPARC microprocessor     

3.3-V BiCMOS circuit techniques for a 120-MHz RISC microprocessor

This paper describes 3.3-V BiCMOS circuit techniques for a 120-MHz RISC microprocessor. The processor is implemented in a 0.5-μm BiCMOS technology with 4-metal-layer structure. The chip includes a 240 MFLOPS fully pipelined 64-b floating point datapath, a 240-MIPS integer datapath, and 24 KB cache, and contains 2.8 million transistors. The processor executes up to four operations at 120 MHz and dissipates 17 W. Novel BiCMOS circuits, such as a 0.6-ns single-ended common base sense amplifier, a 0.46-ns 22-b comparator, and a 0.7-ns path logic adder are applied to the processor. The processor with the proposed BiCMOS circuits has a 11%-47% shorter delay time advantage over a CMOS microprocessor     

A 950-MHz IF second-order integrated LC bandpass delta-sigmamodulator

A second-order integrated LC bandpass delta-sigma modulator is presented. This modulator is implemented in a 0.5-μm bipolar process and can be used for digitizing radio frequency or high intermediate frequency signals. It employs an integrated LC resonator with active Q-enhancement and two nonreturn-to-zero digital-to-analog pulse-shaping feedback loops. The modulator test chip achieves a signal-to-noise ratio of 57 dB over a 200-kHz bandwidth for converting a 950-MHz signal, and dissipates 135 mW with a 5-V supply     

A 64-b quad-issue CMOS RISC microprocessor

The HP-PA8000 is a 180-MHz quad-issue custom VLSI implementation of the HP-PA 2.0 64-b architecture delivering 11.84 SPECint95 and 20.18 SPECfp95 with 3.8 million transistors integrated on a 17.68 mm×19.1 mm die in a 3.3-V, 0.5-μm CMOS process. Specialized clock circuits and extensive use of dynamic logic are key factors in this microprocessor's performance. Attention to clock analysis and distribution resulted in a 170 ps clock skew between any two clock nodes. This microprocessor utilizes a 56-entry instruction reorder buffer (IRE), register renaming, and dual functional units to fully exploit instruction level parallelism     

A 0.25-μm CMOS 0.9-V 100-MHz DSP core

This paper describes a 0.25-μm CMOS 0.9-V 100-MHz DSP core which is composed of a 2-mW 16-b multiplier-accumulator and a 1.5-mW 8-kb SRAM. High-speed operation with a supply of less than 1 V has been achieved by developing 0.25-μm CMOS technology, reducing threshold voltage to 0.3 V, developing tristate inverter 3-2/4-2 adders for the multiplier, realizing small bit-line swing operation for the SRAM, and so on. The adder circuits operate faster than conventional adders at low supply voltages. In addition, short-circuit current and area for diffusion contact are reduced. Small bit-line swing operation has been realized by using a device-deviation immune sense amplifier. Leakage current during sleep mode was reduced by the use of high threshold voltage MOSFETs     

A 900-MHz 1-V CMOS frequency synthesizer

A 900-MHz 1-V frequency synthesizer has been fabricated in a standard 0.35-μm CMOS technology. The frequency synthesizer consists of a divide-by-128/129 and 64/65 dual-modulus prescaler, phase-frequency detector, charge pump, and voltage-doubler circuit with an external voltage-controlled oscillator (VCO) and passive loop filter. The on-chip voltage-doubler circuit converts the 1-V supply voltage to the higher voltage which supplies the prescaler internally. In this way, the 900-MHz 1-V frequency synthesizer with an external VCO can be achieved. The measured phase noise is -112.7 dBc/Hz at a 100-kHz offset from the carrier, and the synthesizer dissipates 3.56 mW (not including VCOs) from a single 1-V supply when the switching frequency of the on-chip voltage doubler is 200 kHz and the power efficiency of the voltage doubler is 77.8%. The total chip area occupies 0.73 mm²