Takayuki Gyohten

A 312-MHz 16-Mb random-cycle embedded DRAM macro with a power-down data retention mode for mobile applications

An embedded DRAM macro with a self-adjustable timing control (STC) scheme, a negative edge transmission scheme (NET), and a power-down data retention (PDDR) mode is developed. A 13.98-mm/sup 2/ 16-Mb embedded DRAM macro is fabricated in 0.13 /spl mu/m logic-based embedded DRAM process. Co-salicide word lines and MIM capacitors are used for high-speed array operation. The delay timing variation of 36 % for an RC delay can be reduced to 3.8% by using the STC scheme. The NET scheme transfers array control signals to local array blocks with high accuracy. Thereby, the test chip achieves 1.2-V 312-MHz random cycle operation even in the low-power process. 73-/spl mu/W data retention power is realized by using the PDDR mode, which is 5% of conventional schemes.

A Configurable Enhanced TTRAM Macro for System-Level Power Management Unified Memory

A twin-transistor random access memory (TTRAM) can provide high speed, low power and high density with CMOS compatible SOI process. However it is difficult to handle as the unified memory required for advanced SoC because it needs the simple control sensing operation for memory compiler, higher cell efficiency, and lower voltage operation for dynamic frequency and voltage control. Enhanced TTRAM (ET2RAM) applies the actively body-bias control technique to realize the low voltage array operation, and never require the negative voltage source. The ET2RAM can realize both 263 MHz at 0.8 V and 10.2 mW at 0.5 V random-cycle operation, higher cell efficiency, and process scalability. It also provides the simple control method suitable for the unified macro for system-level power management SoC with keeping the merits of TTRAM as CMOS compatibility

A Capacitorless Twin-Transistor Random Access Memory (TTRAM) on SOI

We propose a novel capacitorless twin-transistor random access memory (TTRAM). The 2Mb test device has been fabricated on 130nm SOI-CMOS process. We demonstrate the TTRAM cell has two data-storage states and confirm the data retention time of 100 ms at 80°C. TTRAM process is compatible with the conventional SOI-CMOS and never requires any additional processes. A 6.1 ns row-access time is achieved and 250 MHz operation can be realized by using 2 bank 8 b-burst mode.

Compact associative-memory architecture with fully parallel search capability for the minimum Hamming distance

An associative-memory architecture for a fully parallel minimum Hamming distance search is proposed, which uses digital circuitry for bit comparison and fast analog circuitry for word comparison as well as winner-take-all (WTA) functionality. Following this original approach allows compact and high-performance integration in conventional CMOS technology. First, static encoding of word-comparison results as a current-sink capability reduces word-comparison circuitry to the theoretical minimum, namely, one transistor per bit and one signal line per word. Second, a new WTA principle, which we call self-adapting winner line-up amplification (WLA), regulates the winner row output automatically into the narrow maximum-gain region of a distance amplifier. Third, winner search circuit complexity scales linear with reference-word number and not quadratic as inevitable for digital approaches. Due to static distance encoding and WLA regulation, transient noise and fabrication process variations are largely tolerated. Only relative chip-internal transistor-parameter variations, creating effective mismatch of matched transistors, limit winner search result correctness. Practical feasibility is verified.

A capacitorless twin-transistor random access memory (TTRAM) on SOI

We propose a novel capacitorless twin-transistor random access memory (TTRAM). The 2Mb test device has been fabricated on 130nm SOI-CMOS process. We demonstrate the TTRAM cell has two data-storage states and confirm the data retention time of 100ms at 80/spl deg/C. TTRAM process is compatible with the conventional SOI-CMOS and never requires any additional processes. A 6.1ns row-access time is achieved and 250MHz operation can be realized by using 2bank 8b-burst mode.

The Design and Implementation of the Massively Parallel Processor Based on the Matrix Architecture

This paper describes the design and implementation of the massively parallel processor based on the matrix architecture which is suitable for portable multimedia applications. The proposed architecture in this paper achieves the high performance of 40 GOPS in the case of consecutive fixed-point 16-bit additions at 200MHz clock frequency and the small power dissipation of 250mW. In addition, 1Mbit SRAM for data registers and 2048 2-bit-grained processing elements connected by a flexible switching network are integrated in the small area of 3.1 mm 2 in 90nm CMOS low standby technology. These design techniques and architectures described in this paper are attractive for realizing area-efficient, energy-efficient, and high-performance multimedia processors

A 322 MHz random-cycle embedded DRAM with high-accuracy sensing and tuning

A 16 Mb embedded DRAM macro in a fully CMOS logic compatible 90 nm process with a low noise core architecture and a high-accuracy post-fabrication tuning scheme has been developed. Based on the proposed techniques, 61% improvement of the sensing accuracy is realized. Even with the smallest 5 fF/cell capacitance, a 322 MHz random-cycle access while 32 ms data retention time which contributes to save the data retention power down to 60 /spl mu/W are achieved.

An architecture for compact associative memories with deca-ns nearest-match capability up to large distances

Associative-memory architecture for Hamming-distance search, compact implementation, and short nearest-matches times up to large distances are proposed. The main ideas are fast analog word comparison and self-adaptive winner-line-up amplification. An implementation in a 0.6 /spl mu/m 2-poly 3-metal CMOS technology with 32 rows and 128 columns verifies the key concepts. Search time is <38 ns.

A 312MHz 16Mb random-cycle embedded DRAM macro with 73/spl mu/W power-down mode for mobile applications

An embedded DRAM macro with self-adjustable timing control and a power-down data retention scheme is described. A 16Mb test chip is fabricated in a 0.13/spl mu/m low-power process and it achieves 312MHz random cycle operation. Data retention power is 73/spl mu/W, which is 5% compared to a conventional one.

Integration Architecture of Content Addressable Memory and Massive-Parallel Memory-Embedded SIMD Matrix for Versatile Multimedia Processor

This paper presents an integration architecture of content addressable memory (CAM) and a massive-parallel memory-embedded SIMD matrix for constructing a versatile multimedia processor. The massive-parallel memory-embedded SIMD matrix has 2,048 2-bit processing elements, which are connected by a flexible switching network, and supports 2-bit 2,048-way bit-serial and word-parallel operations with a single command. The SIMD matrix architecture is verified to be a better way for processing the repeated arithmetic operation types in multimedia applications. The proposed architecture, reported in this paper, exploits in addition CAM technology and enables therefore fast pipelined table-lookup coding operations. Since both arithmetic and table-lookup operations execute extremely fast, the proposed novel architecture can realize consequently efficient and versatile multimedia data processing. Evaluation results of the proposed CAM-enhanced massive-parallel SIMD matrix processor for the example of the frequently used JPEG image-compression application show that the necessary clock cycle number can be reduced by 86% in comparison to a conventional mobile DSP architecture. The determined performances in Mpixel/mm2 are factors 3.3 and 4.4 better than with a CAM-less massive-parallel memory-embedded SIMD matrix processor and a conventional mobile DSP, respectively.