A 250-Mb/s 32×32 CMOS crosspoint LSI for ATM switchingsystems

A 32×32 crosspoint LSI and a time-slot controlled asynchronous-transfer-mode (ATM) switch architecture utilizing the LSI are presented. The ATM switch, which is classified as an input-buffer-type ATM switch, enables 99% throughput and broadcasting capability. The crosspoint LSI is characterized by the bit-map oriented and pipelined connection control method which can switch and broadcast 160-Mb/s ATM cells, 32×32 switch cells which have less parasitic capacitance, and emitter-coupled-logic (ECL) compatible interfaces which are compatible with a 160-MHz broadband ISDN data rate. The LSI has been fabricated by a 1-μm CMOS process. The chip size is 7.4 mm×7.4 mm. According to the evaluation, operation at 250 Mb/s is confirmed. 1.2-W power consumption is observed at 160-Mb/s operating condition     

A 622-Mb/s 8×8 ATM switch chip set with shared multibufferarchitecture

An asynchronous transfer mode (ATM) switch chip set, which employs a shared multibuffer architecture, and its control method are described. This switch architecture features multiple-buffer memories located between two crosspoint switches. By controlling the input-side crosspoint switch so as to equalize the number of stored ATM cells in each buffer memory, these buffer memories can be treated as a single large shared buffer memory. Thus, buffers are used efficiently and the cell loss ratio is reduced to a minimum. Furthermore, no multiplexing or demultiplexing is required to store and restore the ATM cells by virtue of parallel access to the buffer memories via the crosspoint switches. Access time for the buffer memory is thus greatly reduced. This feature enables high-speed switch operation. A three-VLSI chip set using 0.8-μm BiCMOS process technology has been developed. Four aligner LSIs, nine bit-sliced buffer-switch LSIs, and one control LSI are combined to create a 622-Mb/s 8×8 ATM switching system that operates at 78 MHz. In the switch fabric, 155-Mb/s ATM cells can also be switched on the 622-Mb/s port using time-division multiplexing     

32×32 shared buffer type ATM switch VLSIs for B-ISDNs

A set of 0.8 μm CMOS VLSIs developed for shared buffer switches in asynchronous transfer mode (ATM) switching systems is described. A 32×32 unit switch consists of eight buffer memory VLSIs, two memory control VLSIs, and two commercially available first in first out (FIFO) memory LSIs. Using the VLSIs, the switch can be mounted on a printed board. To provide excellent traffic characteristics not only under random traffic conditions but also under burst traffic conditions, this switch has a 2-Mb shared buffer memory, the largest reported to date. which can save 4096 cells among 32 output ports. This switch has a priority control function to meet the different cell loss rate requirements and switching delay requirements of different service classes. A multicast function and a 600 Mb/s link switch architecture, which are suitable for ATM network systems connecting various media, and an expansion method using the 32×32 switching board to achieve large-scale switching systems such as 256×256 or 1024×1024 switches are discussed     

Scalable 10 Gbit/s 4×20.25 μm CMOS/SIMOX ATM switch LSIcircuit based on distributed contention control

A scalable 10 Gbit/s 4×2 ATM switch LSI circuit has been fabricated. It employs a new distributed contention control technique that makes the LSI circuit expandable. To increase the LSI circuit throughput, 0.2 μm CMOS/SIMOX (separation by implanted oxygen) technology is used. It allows the LSI circuit to offer 221 I/O pins, an operating speed of 1.25 Gbit/s and 7 W power consumption     

BiCMOS circuit technology for a 704 MHz ATM switch LSI

This paper describes BiCMOS level-converter circuits and clock circuits that increase VLSI interface speed to 1 GHz, and their application to a 704 MHz ATM switch LSI. An LSI with a high speed interface requires a BiCMOS multiplexer/demultiplexer (MUX/DEMUX) on the chip to reduce internal operation speed. A MUX/DEMUX with minimum power dissipation and a minimum pattern area can be designed using the proposed converter circuits. The converter circuits, using weakly cross-coupled CMOS inverters and a voltage regulator circuit, can convert signal levels between LCML and positive CMOS at a speed of 500 MHz. Data synchronization in the high speed region is ensured by a new BiCMOS clock circuit consisting of a pure ECL path and retiming circuits. The clock circuit reduces the chip latency fluctuation of the clock signal and absorbs the delay difference between the ECL clock and data through the CMOS circuits. A rerouting-Banyan (RRB) ATM switch, employing both the proposed converter circuits and the clock circuits, has been fabricated with 0.5 μm BiCMOS technology. The LSI, composed of CMOS 15 K gate logic, 8 Kb RAM, I Kb FIFO and ECL 1.6 K gate logic, achieved an operation speed of 704-MHz with power dissipation of 7.2 W     

A 16×622 Mb/s ATM switch: PRELUDE switch architectureintegrated into a 6-million transistor monochip

This paper presents the integration of the Prelude switch architecture into a monochip ATM switch, COM16M, capable of handling 16 multiplexes carrying ATM cells at 622 Mb/s. It is a fully autonomous switch, i.e., the chip includes clock adaptation, routing, and cell buffering as well as header translation and control capabilities. The switch is integrated into one single chip containing 6000000 transistors implemented in a 0.5-μm CMOS process     

Design and implementation of Abacus switch: a scalable multicastATM switch

Describes a new architecture for a multicast ATM switch scalable from a few tens to a few thousands of input ports. The switch, called the Abacus switch, has a nonblocking switch fabric followed by small switch modules at the output ports. It has buffers at input and output ports. Cell replication, cell routing, output contention resolution, and cell addressing are all performed in a distributed way so that it can be scaled up to thousands of input and output ports. A novel algorithm has been proposed to resolve output port contention while achieving input buffers sharing, fairness among the input ports, and call splitting for multicasting. The channel-grouping mechanism is also adopted in the switch to reduce the hardware complexity and improve the switch's throughput, while the cell sequence integrity is preserved. The switch can also handle multiple priority traffic by routing cells according to their priority levels. The performance study of the Abacus switch in throughput, average cell delay, and cell loss rate is presented. A key ASIC chip for building the Abacus switch, called the ARC (ATM routing and concentration) chip, contains a two-dimensional array (32×32) of switch elements that are arranged in a crossbar structure. It provides the flexibility of configuring the chip into different group sizes to accommodate different ATM switch sizes. The ARC chip has been designed and fabricated using 0.8 μm CMOS technology and tested to operate correctly at 240 MHz     

An ATM routing and concentration chip for a scalable multicast ATMswitch

We have proposed a new architecture for building a scalable multicast ATM switch from a few tens to a few thousands of input/output ports. The switch, called the Abacus switch, employs input and output buffering schemes. Cell replication, cell routing, and output contention resolution are all performed in a distributed way so that the switch can be scaled up to a large size. The Abacus switch adopts a novel algorithm to resolve the contention of both multicast and unicast cells destined for the same output port (or output module). The switch can also handle multiple priority traffic by routing cells according to their priority levels. This paper describes a key ASIC chip for building the Abacus switch. The chip, called the ATM routing and concentration (ARC) chip, contains a two-dimensional array (3×32) of switch elements that are arranged in a cross-bar structure. It provides the flexibility of configuring the chip into different group sizes to accommodate different ATM switch sizes. The ARC chip has been designed and fabricated using 0.8-μm CMOS technology and tested to operate correctly at 240 MHz, Although the ARC chip was designed to handle the line rate at OC-3 (155 Mb/s), the Abacus switch can accommodate a much higher line rate at OC-12 (622 Mb/s) or OC-48 (2.5 Gb/s) by using a bit-sliced technique or distributing cells in a cyclic order to different inputs of the ARC chip. When the latter scheme is used, the cell sequence is retained at the output of the Abacus switch     

Measurement-based performance evaluation of an ATM switch withexternal multicasting engine and multiple-priority classes

This paper uses measurement-based traffic models to evaluate a shared-memory ATM switch with 32×32 155 Mbit/s ports and an external multicasting engine; this is the design of Cisco System's next-generation ATM switch, the LightStream-1010 (LS-1010). Assuming that the multicast traffic can take approximately 30% of the total switch load, we find that an external multicasting engine requires a 32 (8) cell buffer at a replication rate of 16 (64) cells per cell service time. We discover that in a multimedia environment, the shared-memory architecture requires 10-30 times less total memory than the bus architecture; a 64 K cell buffer is sufficient to handle 90% utilization with the nonuniform traffic that we investigated. Multiple-priority classes are considered     

The ATM layer chip: an ASIC for B-ISDN applications

The authors describe the architecture of an experimental research prototype application specific integrated circuit (ASIC) designed to serve as a generic building block of the future broadband integrated services digital network (B-ISDN). The chip performs common asynchronous transfer mode (ATM) layer functions such as cell assembly and cell disassembly. A new media access control (MAC) protocol developed for a broadband customer premises network is also integrated in the chip. The chip interfaces to the B-ISDN through a synchronous optical network (SONET) synchronous transmission signal-3c (STS-3c) framer chip. The ATM layer chip has been designed using 1.2 μm CMOS technology with a die area of 5.4×5.4 mm² and approximately 27000 transistors. Experimental results are described. At the user network interface, the chip can be used to implement broadband terminal adaptors and the network termination. At the broadband local exchange, the chip can be used in the implementation of ATM statistical multiplexers, ATM switch port controllers, etc     

A 622-Mb/s bit/frame synchronizer for high-speed backplane datacommunication

A compact 622-Mb/s/port bit/frame synchronizer is presented. Sampling equally-phased clocks from a phase-locked loop (PLL) at the data transition edges, the bit synchronizer selects the optimum one as the extracted clock. An elastic serial-to-parallel converter is used for the frame synchronization. The circuit is designed for a 32-port ATM switch chip, achieving 622-Mb/s port capacity by four parallel 156-Mb/s bits. Using 0.5-μm CMOS technology, the circuit was verified by simulations. The bit synchronizer consumes only 15 mW under typical conditions     

A 20 kbit associative memory LSI for artificial intelligencemachines

A 20 kb (512 words×40 b) CMOS associative-memory LSI is described. This LSI performs large-scale parallelism for highly efficient associative operations in artificial intelligence machines. Relational search, large-bit-length data treatment, and quick garbage collection are realized on the single-chip associative-memory LSI. A cell array structure has been designed in order to reduce the chip area. A newly designed simple accelerator circuit allows for high-speed search operations. The LSI is fabricated using 1.2 μm double-aluminium-layer CMOS process technology. 284000 devices have been integrated on a 5.3×7.9 mm² chip. The measured minimum cycle time and power dissipation at 10 MHz operation are 85 ns and 250 mW, respectively. The associative memory, with its highly efficient associative operation capabilities, promises to be a large step toward the development of high-performance artificial intelligence machines     

Memory array architecture and decoding scheme for 3 V only sectorerasable DINOR flash memory

A memory array architecture and row decoding scheme for a 3 V only DINOR (divided bit line NOR) flash memory has been designed. A new sector organization realizes one word line driver per two word lines, which is conformable to tight word line pitch. A hierarchical negative voltage switching row decoder and a compact source line driver have been developed for 1 K byte sector erase without increasing the chip size. A bit-by-bit programming control and a low threshold voltage detection circuit provide a high speed random access time at low V_cc and a narrow program threshold voltage distribution. A 4 Mb DINOR flash memory test device was fabricated from 0.5 μm, double-layer metal, triple polysilicon, triple well CMOS process. The cell measures 1.8×1.6 μm² and the chip measures 5.8×5.0 mm ². The divided bit line structure realizes a small NOR type memory cell     

A 286 mm2 256 Mb DRAM with ×32 both-ends DQ

This paper describes a 256 Mb DRAM chip architecture which provides up to ×32 wide organization. In order to minimize the die size, three new techniques: an exchangeable hierarchical data line structure, an irregular sense amp layout, and a split address bus with local redrive scheme in the both-ends DQ were introduced. A chip has been developed based on the architecture with 0.25 μm CMOS technology. The chip measures 13.25 mm×21.55 mm, which is the smallest 256 Mb DRAM ever reported. A row address strobe (RAS) access time of 26 ns was obtained under 2.8 V power supply and 85°C. In addition, a 100 MHz×32 page mode operation, namely 400 M byte/s data rate, in the standard extended data output (EDO) cycle has been successfully demonstrated     

A 1.5-ns 256-kb BiCMOS SRAM with 60-ps 11-K logic gates

A 1.5-ns address access time, 256-kb BiCMOS SRAM has been developed. To attain this ultra-high-speed access time, an emitter-coupled logic (ECL) word driver is used to access 6-T CMOS memory cells, eliminating the ECL-MOS level-shifter time delay. The RAM uses a low-power active pull down ECL decoder. The chip contains 11-K, 60-ps ECL circuit gates. It provides variable RAM configurations and general logic functions. RAM power consumption is 18 W; chip power consumption is 35 W. The chip is fabricated by using a 0.5-μm BiCMOS process. The memory cell size is 58 μm² and the chip size is 11×11 mm     

An 8.8-ns 54×54-bit multiplier with high speed redundantbinary architecture

A high speed redundant binary (RB) architecture, which is optimized for the fast CMOS parallel multiplier, is developed. This architecture enables one to convert a pair of partial products in normal binary (NB) form to one RE number with no additional circuit. We improved the RB adder (RBA) circuit so that it can make a fast addition of the RB partial products. We also simplified the converter circuit that converts the final RE number into the corresponding NE number. The carry propagation path of the converter circuit is carried out with only multiplexer circuits. A 54×54-bit multiplier is designed with this architecture. It is fabricated by 0.5 μm CMOS with triple level metal technology. The active area size is 3.0×3.08 mm² and the number of transistors is 78,800. This is the smallest number for all 54×54-bit multipliers ever reported. Under the condition of 3.3 V supply voltage, the chip achieves 8.8 ns multiplication time. The power dissipation of 540 mW is estimated for the operating frequency of 100 MHz. These are, so far, the fastest speed and the lowest power for 54×54-bit multipliers with 0.5-μm CMOS     

A 10-Gb/s (1.25 Gb/s×8)4×2 0.25-μm CMOS/SIMOX ATMswitch based on scalable distributed arbitration

This paper presents the design and implementation of a scalable asynchronous transfer mode switch. We fabricated a 10-Gb/s 4×2 switch large-scale integration (LSI) that uses a new distributed contention control technique that allows the switch LSI to be expanded. The developed contention control is executed in a distributed manner at each switch LSI, and the contention control time does not depend on the number of connected switch LSI's. To increase the LSI throughput and reduce the power consumption, we used 0.25-μm CMOS/SIMOX (separation by implanted oxygen) technology, which enables us to make 221 pseudo-emitter-coupled-logic I/O pins with 1.25-Gb/s throughput. In addition, power consumption of 7 W is achieved by operating the CMOS/SIMOX gates at -2.0 V. This consumption is 36% less than that of bulk CMOS gates (11 W) at the same speed at -2.5 V. Using these switch LSI's, an 8×8 switching multichip module with 80-Gb/s throughput was fabricated with a compact size     

A 7.5-ns 32 K×8 CMOS SRAM

A 256 K (32 K×8) CMOS static RAM (SRAM) which achieves an access time of 7.5 ns and 50-mA active current at 50-MHz operation is described. A 32-block architecture is used to achieve high-speed access and low power dissipation. To achieve faster access time, a double-activated-pulse circuit which generates the word-line-enable pulse and the sense-amplifier-enable pulse has been developed. The data-output reset circuit reduces the transition time and the noise generated by the output buffer. A self-aligned contact technology reduces the diffused region capacitance. This RAM has been fabricated in a twin-tub CMOS 0.8-μm technology with double-level polysilicon (the first level is polycide) and double-level metal. The memory cell size is 6.0×11.0 μm² and the chip size is 4.38×9.47 mm ²     

A self-learning neural network chip with 125 neurons and 10 Kself-organization synapses

A learning neural network LSI chip is described. The chip integrates 125 neuron units and 10K synapse units with the 1.0 μm double-poly-Si, double-metal CMOS technology. Most of this integration has been realized by using a mixed design architecture of digital and analog circuits. The fully feedback connection network LSI can memorize at least 15 patterns with 50 μs learning time for each pattern. Under the condition that each test vector keeps a Hamming distance of 6 from memorized pattern, a correct association rate of 98% is obtained. The relaxation time is 1 to 2 μs. This chip consumes less than 7.5 W     

Design of a second-level cache chip for shared-busmultimicroprocessor systems

The design of a second-level cache chip with the most suitable architecture for shared-bus multiprocessing is described. This chip supports high-speed (160-MB/s) burst transfer between multilevel caches and a newly proposed cache-consistency protocol. The chip, which supports a 50-MHz CPU and uses 0.8 μm CMOS technology, includes a 32 kB data memory, 42 kb tag memory. and 21.7 K-gate logic