A sub-40-ns chain FRAM architecture with 7-ns cell-plate-line drive

A nonvolatile chain FRAM adopting a new cell-plate-line drive technique was demonstrated. Two key circuit techniques, a two-way metal cell-plate line and a cell-plate line shared with 16 cells, reduce cell-plate-line delay to 7 ns and reduce plate drive area to 1/5. The total cell-plate-line delay, including cell transistor delay due to eight cells in series, is reduced to 15 μs, in contrast to 30-100-ns delay of the conventional FRAM. The die size is reduced to 86% that of the conventional FRAM by reduction of the plate driver area and sense amplifier area, assuming the same memory cell size. A prototype 16-kb chain FRAM chip was fabricated using 0.5 μm rule one-polycide and two-metal CMOS process. The memory cell size was 13.26 μm² using a 3.24-μm² capacitor. Thanks to the fast cell-plate-line drive, the chain FRAM test chip has achieved the fastest random access time, 37 ns, and read/write cycle time, 80 ns, at 3.3 V so far reported. The chain FRAM has also realized V_{dd min} of 2.3 V and 10¹⁰ read/write cycles     

100-MHz serial access architecture for 4-Mb field memory

A 4-Mb field memory with a 100-MHz serial access rate has been developed. A new architecture that significantly improves serial I/O operation speed, reduces layout area, and offers simple control is proposed. To accomplish this task, a new architectural data shifter and high-speed redundancy circuit have been developed. The field memory has a 568-line×960 pixel×8-b (4,362,240 b) memory cell array designed for high-definition television (HDTV) screens. A 1.0 μm CMOS process technology is used to produce a die size of 12.94 mm×25.9 mm. The write-read cycle time is 9 ns, the access time is 8 ns, and the active current is 170 mA at a 50-MHz cycle rate with a standby current of about 3 mA     

GaAs Schmitt trigger memory cell design

A novel GaAs five-transistor static memory cell derived from a Schmitt trigger is proposed. The memory cell overcomes MESFET subthreshold leakage loss by using a self ground-shifting technique which limits the leakage current flow to the cell. Compared with conventional GaAs SRAM cells, it offers small area and as well as fast read/write cycles. A 1 Kb prototype implemented in 1 μm nonself-aligned GaAs MESFET technology exhibited read and write access times of the order of 2.0 ns     

A 23-ns 4-Mb CMOS SRAM with 0.2-μA standby current

A 4-Mb CMOS SRAM having 0.2-μA standby current at a supply voltage of 3 V has been developed. Current-mirror/PMOS cross-coupled cascade sense-amplifier circuits have achieved the fast address access time of 23 ns. A new noise-immune data-latch circuit has attained power-reduction characteristics at a low operating cycle time without access delay. A 0.5-μm CMOS, four-level poly, two-level metal technology with a polysilicon PMOS load memory cell, yielded a small cell area of 17 μm² and the very small standby current. A quadruple-array, word-decoder architecture allowed a small chip area of 122 mm²     

Pipelined, time-sharing access technique for an integratedmultiport memory

A pipelined, time-sharing access (PTA) technique that realizes an integrated multiport memory for high-speed signal processing is described. N/2-port memory cells, with less area and a wider operating margin, are used for the N-port memory function. Memory cell access for multiple ports is performed serially within one cycle, instead of being an ordinary parallel operation. Memory operation is divided into three pipeline cycles-address selection, memory cell access, and data I/O operation-to reduce the cycle time. A 64-kb four-port memory was fabricated with conventional two-port memory cells to verify the effectiveness of this technique. A 16 ns memory operation with a wide margin was observed under a 3 V supply voltage     

A 1-Mb 2-Tr/b nonvolatile CAM based on flash memory technologies

This paper describes the circuit technologies and the experimental results for a 1 Mb flash CAM, a content addressable memory LSI based on flash memory technologies. Each memory cell in the flash CAM consists of a pair of flash memory cell transistors. Additionally, four new circuit technologies have been developed: a small-size search sense amplifier; a highly parallel search management circuit; a high-speed priority encoder; and word line/bit line redundancy circuits for higher production yields. A cell size of 10.34 μm² and a die size of 42.9 mm² have been achieved with 0.8 μm design rules. Read access time and search access time are 115 ns and 135 ns, respectively, with a 5 V supply voltage. Power dissipation in 3.3 MHz operations is 210 mW in read and 140 mW in search access     

A 250-MHz skewed-clock pipelined data buffer

A synchronous dual-port memory employing a three-transistor (3T) dynamic cell has been designed for use as a high throughput embedded data buffer in digital switching and signal processing applications. Skewed-clock pipelining is used to achieve operation at frequencies as high as 250 MHz with a low register element count. The 3T cell provides separate read and write access ports while occupying less than half the area of a conventional dual-port SRAM cell. On-chip Hamming error correction coding (ECC) is used to enhance the fault tolerance of the memory, A 25-kb experimental prototype has been integrated in a 0.8-μm CMOS technology; it occupies a die area of 3800 μm×1600 μm and dissipates 420 mW while operating at 250 MHz     

A 1-V TFT-load SRAM using a two-step word-voltage method

A 1-V SRAM using a TFT load cell was developed. Key circuits for obtaining the low-voltage operation are a two-step word-voltage (TSW) method, a submicroampere boosted-level generator using a multivibrator, and a sense amplifier using low-threshold MOSFETs. An access time of 250 ns and a standby current of 0.23 μA were achieved for a 4-kb test chip using a 10.2-μm² TFT-load cell. This technology is applicable for high-density and single-battery operational SRAMS     

A 7-ns/850-mW GaAs 4-kb SRAM with little dependence on temperature

A GaAs 1 K×4-kb SRAM designed using a novel circuit technology is described. To reduce the temperature dependence and the scattering of the access time, it was necessary to increase the signal voltage swing and to reduce the leakage current in access transistors of unselected memory cells. In the 4-kb SRAM, source-follower circuits were adopted to increase the voltage swing, and the storage nodes of unselected memory cells were raised by about 0.6 V to reduce the subthreshold leakage current in the access transistors. The 4-kb SRAM was fabricated using 1.0-μm self-aligned MESFETs with buried p-layers beneath the FET regions. A maximum address access time of 7 ns and a power dissipation of 850 mW were obtained for the galloping test pattern at 75°C. Little change in the address access time was observed between 0 and 75°C     

A new low-power GaAs two-single-port memory cell

This paper describes an experimental static memory cell in GaAs MESFET technology. The memory cell has been implemented using a mix of several techniques already published in order to overcome some of their principal drawbacks related to ground shifting, destructive readout, and leakage current effects. The cell size is 36×37 μm² using a 0.6-μm technology. An experimental 32 word × 32 bit array has been designed. From simulation results, an address access time of 1 ns has been obtained. A small 8 word×4 bit protoype was fabricated. The cell can be operated at the single supply voltage from 1 up to 2 V. The evaluation is provided according to the functionality and power dissipation. Measured results show a total current consumption of 14 μA/cell when operated at 1 V     

Advanced TFT SRAM cell technology using a phase-shift lithography

An advanced TFT memory cell technology has been developed for making high-density and high-speed SRAM cells. The cell is fabricated using a phase-shift lithography that enables patterns with spaces of less than 0.25 μm to be made using the conventional stepper. Cell area is also reduced by using a small cell-ratio and a parallel layout for the transistor. Despite the small cell-ratio, stable operation is assured by using advanced polysilicon PMOS TFT's for load devices. The effect of the Si₃N₄ multilayer gate insulator on the on-current and the influence of the channel implantation are also investigated. To obtain stable operation and extremely low stand-by power dissipation, a self-aligned offset structure for the polysilicon PMOS TFT is proposed and demonstrated. A leakage current of only 2 fA/cell and an on-/off-current ratio of 4.6×10⁶ are achieved with this polysilicon PMOS TFT in a memory cell, which is demonstrated in a experimental 1-Mbit CMOS SRAM chip that has an access time of only 7 ns     

An 8 ns 4 Mb serial access memory

A new architecture for serial access memory is described that enables a static random access memory (SRAM) to operate in a serial access mode. The design target is to access all memory address serially from any starting address with an access time of less than 10 ns. This can be done by all initializing procedure and three new circuit techniques. The initializing procedure is introduced to start the serial operation at an arbitrary memory address. Three circuit techniques eliminate extra delay time caused by an internal addressing of column lines, sense amplifiers, word lines, and memory cell blocks. This architecture was successfully implemented in a 4-Mb CMOS SRAM using a 0.6 μm CMOS process technology. The measured serial access time was 8 ns at a single power supply voltage of 3.3 V     

A 3.3-V single power supply 16-Mb nonvolatile virtual DRAM using aNAND flash memory technology

A 3.3-V 16-Mb nonvolatile memory having operation virtually identical to DRAM with package pin compatibility has been developed. Read and write operations are fully DRAM compatible except for a longer RAS precharge time after write. Fast random access time of 63 ns with the NAND flash memory cell is achieved by using a hierarchical row decoder scheme and a unique folded bit-line architecture which also allows bit-by-bit program verify and inhibit operation. Fast page mode with a column address access time of 21 ns is achieved by sensing and latching 4 k cells simultaneously. To allow byte alterability, nonvolatile restore operation with self-contained erase is developed. Self-contained erase is word-line based, and increased cell disturb due to the word-line based erase is relaxed by adding a boosted bit-line scheme to a conventional self-boosting technique. The device is fabricated in a 0.5-μm triple-well, p-substrate CMOS process using two-metal and three-poly interconnect layers. A resulting die size is 86.6 mm², and the effective cell size including the overhead of string select transistors is 2.0 μm²     

A 2-ns 16K bipolar ECL RAM with reduced word-line voltage swing

A reduced word-line voltage swing (RWS) circuit configuration that results in a high-speed bipolar ECL (emitter coupled logic) RAM is proposed. The write operation can be performed with the configuration in the condition of reduced word-line voltage swing, which causes write operation error in conventional circuit configurations. The proposed configuration cuts off the hold current of the selected memory cell, and then the low-voltage node is charged up through the load p-n-p transistor. A 16-kb ECL RAM with a p-n-p loaded memory cell was fabricated by advanced silicide-base transistor (ASBT) process technology. A 2-ns access time was obtained with 1.8-W power consumption in which the word-line voltage swing was reduced by 0.7 V from a conventional case. Simulation results show that the access time is improved by 25% compared with a conventional case. Simulation results also show that writing time becomes comparable with the conventional time of 1.7 ns when the load p-n-p transistor has a saturation current of 5.0× 10¹⁷ A and a current gain of 1.0. The saturation current is 5 times larger and the current gain is 5 times smaller than those of the standard lateral p-n-p transistor     

A 1-Mbit CMOS EPROM with enhanced verification

A high-speed 1-Mb EPROM (erasable programmable read-only memory) with an enhanced verify mode to insure adequate threshold shift after programming has been developed. The sense circuitry uses an offset current to shift the sense point to require higher threshold shift during verification. The access time is improved by a clocking scheme that balances the sensing circuitry between column accesses and by a chip architecture optimized for speed. The chip features an access time of 70 ns and an active current of 20 mA. A typical programming time of 50 μs has been measured. The device is processed in a 1-μm L _eff CMOS process with silicides     

An experimental 1-Mbit CMOS SRAM with configurable organization andoperation

A 1-Mb SRAM (static random-access memory) configurable as a 128-kb×8, 256-kb×4, or 1-Mb×1 memory featuring asynchronous operation with static-column and chip-enable-access speedup modes or synchronous operation with a fast-page (toggle) or static-column mode is described. It has been fabricated in a double-metal, double-polysilicon CMOS process with 0.7-μm geometry and special SRAM structures. The measured synchronous access of 29 ns with a fast-page mode access of 22 ns. Measured asynchronous access is 34 ns with a static-column access of 33 ns and a chip-select speedup access of 29 ns. The SRAMs six-transistor CMOS memory cell is 58.24 μm²     

An ultrahigh-density high-speed loadless four-transistor SRAM macrowith twisted bitline architecture and triple-well shield

We have developed two schemes for improving access speed and reliability of a loadless four-transistor (LL4T) SRAM cell: a dual-layered twisted bitline scheme, which reduces coupling capacitance between adjacent bitlines in order to achieve highspeed READ/WRITE operations, and a triple-well shield, which protects the memory cell from substrate noise and alpha particles. We incorporated these schemes in a high-performance 0.18-μm-generation CMOS technology and fabricated a 16-Mb SRAM macro with a 2.18-μm² memory cell. The macro size of the LL4T-SRAM is 56 mm², which is 30% to 40% smaller than a conventional six-transistor SRAM when compared with the same access speed. The developed macro functions at 500 MHz and has an access time of 2.0 ns. The standby current has been reduced to 25 μA/Mb with a low-leakage nMOSFET in the memory cell     

A 76-mm2 8-Mb chain ferroelectric memory

This paper demonstrates the first 8-Mb chain ferroelectric RAM (chain FeRAM) with 0,25-μm 2-metal CMOS technology. A small die of 76 mm² and a high average cell/chip area efficiency of 57.4 % have been realized by introducing not only chain architecture but also four new techniques: 1) a one-pitch shift cell realizes small cell size of 5.2 μm²; 2) a new hierarchical wordline architecture reduces row-decoder and plate-driver areas without an extra metal layer; 3) a small-area dummy cell scheme reduces dummy capacitor size to 1/3 of the conventional one; and 4) a new array activation scheme reduces dataline and second amplifier areas. As a result, the chain architecture with these new techniques reduces die size to 65% of that of the conventional FeRAM. Moreover a ferroelectric capacitor overdrive scheme enables sufficient polarization switching, without overbias memory cell array. This scheme lowers the minimum operation voltage by 0.23 V, and enables 2.5-V V_dd operation. Thanks to fast cell plateline drive of chain architecture, the 8-Mb chain FeRAM has achieved the fastest random access time, 40 ns, and read/write cycle time, 70 ns, at 3.0 V so far reported     

An experimental DRAM with a NAND-structured cell

An experimental 256-Mb dynamic random access memory using a NAND-structured cell (NAND DRAM) has been fabricated. The NAND-structured cell has four memory cells connected in series, which reduces the area of isolation between the adjacent cells and also reduces the bit-line contact area. The cell area per bit measures 0.962 μm², using 0.4-μm CMOS technology, which is 63% in comparison with the conventional cell. In order to reduce the die size, time division multiplex sense-amplifier (TMS) architecture, in which a sense amplifier is shared by four bit lines, has been newly introduced. The chip area is 464 mm², which is 68% compared with the DRAM using the current cell structure. The data can be accessed by a fast-block-access mode up to 512 bytes as well as a random access mode. Typical 112-ns access time of the first data in a block and 30-ns serial cycle time are achieved     

High-speed sensing techniques for ultrahigh-speed SRAMs

Two high-speed sensing techniques suitable for ultrahigh-speed SRAMs are proposed. These techniques can reduce a 64-kb SRAM access time to 71~89% of that of conventional high-speed bipolar SRAMs. The techniques use a small CMOS memory cell instead of the bipolar memory cell that has often been used in conventional bipolar SRAMs for cache and control memories of mainframe computers. Therefore, the memory cell size can also be reduced to 26~43% of that of conventional cells. A 64-kb SRAM fabricated with one of the sensing techniques using 0.5-μm BiCMOS technology achieved a 1.5-ns access time with a 78-μm² memory cell size. The techniques are especially useful in the development of both ultrahigh-speed and high-density SRAMs, which have been used as cache and control memories of mainframe computers