Gyung-Su Byun

A 1.8 V 700 Mb/s/pin 512 Mb DDR-II SDRAM with on-die termination and off-chip driver calibration

A 1.8 V 700 Mb/s/pin 512 Mb DDR-II SDRAM is JEDEC standard compliant. With the hierarchical I/O line and local sensing, t/sub AA/ /t/sub RCD//t/sub RP/ of 3/3/3 at 533 Mb/s are achieved in the design. For signal integrity at 533 Mb/s, off-chip driver calibration and on-die termination are employed.

An Energy-Efficient and High-Speed Mobile Memory I/O Interface Using Simultaneous Bi-Directional Dual (Base+RF)-Band Signaling

A fully-integrated 8.4 Gb/s 2.5 pJ/b mobile memory I/O transceiver using simultaneous bidirectionaldual band signaling is presented. Incorporating both RF-band and baseband transceiver designs, this prototype demonstrates an energy-efficient and high-bandwidth solution for future mobile memory I/O interface. The proposed amplitude shift keying (ASK) modulator/demodulator with on-chip band-selective transformer obviates a power hungry pre-emphasis and equalization circuitry, revealing a low-power, compact and standard mobile memory-compatible solution. Designed and fabricated in 65-nm CMOS technology, each RF-band and baseband transceiver consumes 10.5 mW and 11 mW and occupies 0.08 mm2 and 0.06 mm2 die area, respectively. The dual-band transceiver achieves error-free operation (BER <; 10-15 ) with 223- 1 PRBS at 8.4 Gb/s over a distance of 10 cm.

A 20-gb/s 256-mb DRAM with an inductorless quadrature PLL and a cascaded pre-emphasis transmitter

A 20GB/s 1.8V 256MB DRAM is designed and fabricated using an 80nm CMOS process. An inductorless tetrahedral oscillator generates inherent quadrant phases combined with a cascaded pre-emphasis transmitter to achieve a 10Gbit/s/pin data rate.

The DIMM tree architecture: A high bandwidth and scalable memory system

The demand for capacity and off-chip bandwidth to DRAM will continue to grow as we integrate more cores onto a die. However, as the data rate of DRAM has increased, the number of DIMMs supported on a multi-drop bus has decreased. Therefore, traditional memory systems are not sufficient to meet both these demands. We propose the DIMM tree architecture for better scalability by connecting the DIMMs as a tree. The DIMM tree architecture is able to grow the number of DIMMs exponentially with each level of latency in the tree. We also propose application of Multiband Radio Frequency Interconnect (MRF-I) to the DIMM tree architecture for even greater scalability and higher throughput. The DIMM tree architecture without MRF-I was able to scale up to 64 DIMMs with only an 8% degradation in throughput over an ideal system. The DIMM tree architecture with MRF-I was able to increase throughput by 68% (up to 200%) on a 64-DIMM system over a 4-DIMM system.

Reevaluating the latency claims of 3D stacked memories

In recent years, 3D technology has been a popular area of study that has allowed researchers to explore a number of novel computer architectures. One of the more popular topics is that of integrating 3D main memory dies below the computing die and connecting them with through-silicon vias (TSVs). This is assumed to reduce off-chip main memory access latencies by roughly 45% to 60%. Our detailed circuit-level models, however, demonstrate that this latency reduction from the TSVs is significantly less. In this paper, we present these models, compare 2D and 3D main memory latencies, and show that the reduction in latency from using 3D main memory to be no more than 2.4 ns. We also show that although the wider I/O bus width enabled by using TSVs increases performance, it may do so with an increase in power consumption. Although TSVs consume less power per bit transfer than off-chip metal interconnects (11.2 times less power per bit transfer), TSVs typically use considerably more bits and may result in a net increase in power due to the large number of bits in the memory I/O bus. Our analysis shows that although a 3D memory hierarchy exploiting a wider memory bus can increase performance, this performance increase may not justify the net increase in power consumption.

A 1.4 Gb/s DLL using 2nd order charge-pump scheme with low phase/duty error for high-speed DRAM application

A technique for reducing the phase error of DLL/PLLs, due to non-ideal characteristics of the charge pump, is proposed. It makes the output of the charge pump virtually grounded, to eliminate the current mismatch and to seamlessly convert the locking information into digital form. A DLL is designed and fabricated to exhibit duty-cycle corrector performance with a speed of 1.4 Gb/s.

Analysis of Noncoherent ASK Modulation-Based RF-Interconnect for Memory Interface

A noncoherent amplitude shift keying (ASK)-based RF-interconnect (RF-I) system design for off-chip communication is analyzed. The proposed RF-I system exploits the simple architecture and characteristics of noncoherent ASK modulation. This provides an efficient way of increasing interconnect bandwidth by transmitting an RF-modulated data stream simultaneously with a conventional baseband counterpart over a shared off-chip transmission line. Both analysis and tested results prove that the performance of the proposed dual-band (RF+baseband) interconnect system is not limited by thermal noise interference. Therefore, a more sophisticated modulation scheme and/or coherent receiving scheme becomes unnecessary within the scope of system requirements. In addition, it confirms that the proposed inductive coupling network is able to support simultaneous bidirectional communications without using complicated replica circuits or additional filters to isolate simultaneous baseband and RF-band data streams.

An 8Gb/s/pin 4pJ/b/pin Single-T-Line dual (base+RF) band simultaneous bidirectional mobile memory I/O interface with inter-channel interference suppression

The demand for higher power efficiency and bandwidth is increasing as mobile devices keep enhancing its graphic computing and media processing capabilities. Current memory interfaces with single-wire signaling operate at 5Gb/s/pin [1] and 6Gb/s/pin [2] with the power efficiency of 17.4pJ/b/pin and 15.8pJ/b/pin, respectively. Mobile DDR memory I/O with differential signaling has better power efficiency of 6.4pJ/b/pin [3], and so does the prior dual-band interconnect (DBI) [4] with the efficiency of 5pJ/b/pin at 4.2Gb/s/pin for simultaneous bidirectional (SBD) mobile memory I/O interface. However, DBIs differential signaling is incompatible with existing standards, and it also occupies large die area for using differential transmission lines and an LC-oscillator for generating RF-carrier. To alleviate these concerns, we propose to use a Single-Transmission-Line DBI (STL-DBI) with the best figure-of-merit (FoM) defined as data rate per pin divided by the I/O-interface die area and power consumption.

A

This letter presents a divide-by-five injection-locked frequency divider (ILFD) using a near-threshold (NT) VCO. The input frequency is injected in the proposed ILFD by a VCO employing a noise-cancelling technique. Both the proposed ILFD and VCO are based on a single compact coil which consists of a distributed and center-tapped inductor for an enhanced ILFD locking range of 20.2 to 21.9 GHz without varactors. The proposed VCO and ILFD are fabricated in a 0.13 μm CMOS process and consume 0.3 and 0.55 mW, respectively, from a 0.45 V supply.

K

The demand for capacity and off-chip bandwidth to dynamic random-access memory (DRAM) will continue to grow as we integrate more cores onto a die. However, as the data rate of DRAM has increased, the number of dual in-line memory modules (DIMMs) supported on a multi-drop bus has decreased. Therefore, traditional memory systems are not sufficient to meet both these demands. We propose the DIMM tree architecture for better scalability by connecting the DIMMs as a tree. The DIMM tree architecture is able to grow the number of DIMMs exponentially with each level of latency in the tree. We also propose application of multiband radio-frequency interconnect (MRF-I) to the DIMM tree architecture for even greater scalability and higher throughput. The DIMM tree architecture without MRF-I was able to scale up to 64 DIMMs with only an 8% degradation in throughput over an ideal system. The DIMM tree architecture with MRF-I was able to increase throughput by 68% (up to 200%) on a 64-DIMM system over a 4-DIMM system. Finally, we propose the partitioned DIMM tree, which allows the scaling of a main memory system to a many-DIMM memory system while still maintaining high throughput. The partitioned DIMM tree is able to improve throughput by an average of 19% up to 35% over the DIMM tree with 256 DIMMs on a single channel.