previous
next

A weak ILO enables near-constant JTB.

Presented at VLSI 2011 by Sang-Hye Chung

This paper presents a forwarded-clock I/O receiver with wide range constant jitter tracking bandwidth (JTB). A JTB interpolator in the receiver enables wide JTB using a weak injection-locked oscillator (ILO) instead of a strong ILO, so narrow VCO tuning is possible. By using the JTB interpolator, the proposed receiver compensates for the JTB reduction caused by deskew-JTB dependency in conventional ILOs. An 8Gb/s receiver has been implemented to verify 70M~1GHz constant JTB and one UI deskew with only 1.94~2GHz VCO tuning. The 0.13um test chip consumes 17.2~21.4mW and occupies 0.04mm2. Keywords: Source synchronous link, injection-locked oscillator, receiver.

news

Multimedia Processor on Si-Interposer platform

Presented at ISSCC 2011 by Hyo-Eun Kim

A multimedia processor embedding a reconfigurable transceiver pool is presented for IC-Stacking on Si-interposer. Configurable vector processing units for frame-level parallelism, a unified filtering unit with memory-access-efficient texturing algorithm, and a programmable shader integrating multiple cores are used to achieve 140GOPS/W, 5.8GFLOPS/W in full operation. This processor achieves 1.7x higher frame rate, 8x higher memory bandwidth compared to the previous design by using architecture/circuit-level performance/energy optimization and IC-Stacking technology. The proposed filtering unit and algorithm reduces the energy-delay-product 83.8% compared to the previous design. The proposed graphics core architecture reduces the execution latency by 53% compared to the conventional SIMD architecture.

news

Mobile Ray Tracing Processor

Accepted in JSSC (Feb. 2012) by Hong-Yun Kim

This processor presents a mobile ray tracing processor (MRTP) with reconfigurable stream multi-processors (RSMPs) for high datapath utilization. The MRTP includes three RSMPs that operate in multiple instruction multiple data (MIMD) mode asynchronously to exploit instruction-level parallelism. Each RSMP is based on single instruction multiple thread (SIMT) architecture to exploit thread-level parallelism. An RSMP consists of twelve scalar processing elements (SPEs) that run multiple threads in parallel synchronously. A low datapath utilization caused by a branch divergence is improved by 19.9% on average by reconfiguring twelve SPEs between scalar and vector SIMT with 0.1% area overheads. MRTP achieves a peak performance of 673K rays per second while consuming 156mW at 100MHz with VDD=1.2V

news

Adaptive Equalizer

Accepted in TVLSI (2012) by Won-Young Lee

An adaptive equalizer with the capacitance multiplication for DisplayPort main link has been proposed. The proposed equalizing filter is based on Miller’s theorem and composed of MIM capacitors and a sub-amplifier. The active source degeneration capacitor achieves low cost and area saving with the capacitance multiplication. The equalizer satisfies the specification of DisplayPort version 1.1a. The measured eye widths of 2.7 Gb/s data are 0.6 UI and 0.5 UI for 5 m and 8 m cables, respectively. The core area is 286×380 μm2 and power consumption is 22.3 mW at 2.7 Gb/s at 1.8 V.

news

Recently Accepted Papers

Journal (to be published)

Won-Young Lee, Lee-Sup Kim
A 5.4 Gb/s Clock and Data Recovery Circuit Using Seamless Loop Transition Scheme with Minimal Phase Noise Degradation
IEEE Transactions on Circuits and Systems-I:Regular Papers, 2012
 
Jae-Sung Yoon, Jeonghyun Kim, Hyo-Eun Kim, Won-Young Lee, Seok-Hoon Kim, Kyusik Chung, Jun-Seok Park, Lee-Sup Kim
A Unified Graphics and Vision Processor with a 0.89µW/fps Pose Estimation Engine for Augmented Reality
IEEE Transactions on Very Large Scale Integration Systems, 2012
 
Young-Jun Kim, Hyo-Eun Kim, Seok-Hoon Kim, Jun-Seok Park, Seungwook Paek, Lee-Sup Kim
Homogeneous Stream Processors with Embedded Special Function Units for High-Utilization Programmable Shaders
IEEE Transactions on Very Large Scale Integration Systems, 2011
 

Conference (to be presented)

Sang-Hye Chung, Lee-Sup Kim
A 1.22mW/Gb/s 9.6Gb/s Data Jitter Mixing Forwarded-Clock Receiver Robust against Power Noise with 1.92ns Latency Mismatch between Data and Clock in 65nm CMOS
IEEE Symposium on VLSI Circuits, 2012
 
Seungwook Paek, Seok-Hwan Moon, Wongyu Shin, Jaehyeong Sim, Lee-Sup Kim
PowerField: A Transient Temperature-to-Power Technique based on Markov Random Field Theory
ACM/IEEE Design Automation Conference, 2012
 
Yong-Hun Kim, Lee-Sup Kim
A 20 Gbps 1-Tap Decision Feedback Equalizer with Unfixed Tap Coefficient
IEEE International Symposium on Circuits and Systems, 2012