A digital-optical system-on-package module architecture for high performance multiprocessors

Gee Kung Chang; Daniel Guidotti; Zhaoran Huang; Fuhan Liu; Yin Jung Chang; Jianjun Yu; Rao R. Tummala

A digital-optical system-on-package module architecture for high performance multiprocessors

Gee Kung Chang, Daniel Guidotti, Zhaoran Huang, Fuhan Liu, Yin Jung Chang, Jianjun Yu, Rao R. Tummala

Department of Optics and Photonics

Research output: Contribution to journal › Conference article › peer-review

1 Scopus citations

Abstract

Data transport bottlenecks namely memory access delay (MAD) and node crossing delay (NCD) were analyzed using a digital-optical system-on-package module architecture. The MAD bottleneck was minimized by using a 3-D distributed shared memory field in which high speed OIs delivered data to and from each processor. The NCD bottleneck was resolved by the use of non-blocking, high speed, optical switching, each switch having a local controller that was optically synchronized and programmed by an optical header signal. It was found that the coupling efficiency between laser and waveguide was around 45% and the coupling efficiency between waveguide with a 45° end mirror and the PiN PD with a36 μm diameter active area was around 35%.

Original language	English
Article number	WO2
Pages (from-to)	581-582
Number of pages	2
Journal	Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS
Volume	2
State	Published - 2004
Event	2004 IEEE LEOS Annual Meeting Conference Proceedings, LEOS 2004 - Rio Grande, Puerto Rico Duration: 7 Nov 2004 → 11 Nov 2004

Cite this

@article{fc80e4b67e4d45e1b64ec001f2b94a1c,

title = "A digital-optical system-on-package module architecture for high performance multiprocessors",

abstract = "Data transport bottlenecks namely memory access delay (MAD) and node crossing delay (NCD) were analyzed using a digital-optical system-on-package module architecture. The MAD bottleneck was minimized by using a 3-D distributed shared memory field in which high speed OIs delivered data to and from each processor. The NCD bottleneck was resolved by the use of non-blocking, high speed, optical switching, each switch having a local controller that was optically synchronized and programmed by an optical header signal. It was found that the coupling efficiency between laser and waveguide was around 45% and the coupling efficiency between waveguide with a 45° end mirror and the PiN PD with a36 μm diameter active area was around 35%.",

author = "Chang, {Gee Kung} and Daniel Guidotti and Zhaoran Huang and Fuhan Liu and Chang, {Yin Jung} and Jianjun Yu and Tummala, {Rao R.}",

year = "2004",

language = "???core.languages.en_GB???",

volume = "2",

pages = "581--582",

journal = "Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS",

issn = "1092-8081",

note = "2004 IEEE LEOS Annual Meeting Conference Proceedings, LEOS 2004 ; Conference date: 07-11-2004 Through 11-11-2004",

}

TY - JOUR

T1 - A digital-optical system-on-package module architecture for high performance multiprocessors

AU - Chang, Gee Kung

AU - Guidotti, Daniel

AU - Huang, Zhaoran

AU - Liu, Fuhan

AU - Chang, Yin Jung

AU - Yu, Jianjun

AU - Tummala, Rao R.

PY - 2004

Y1 - 2004

N2 - Data transport bottlenecks namely memory access delay (MAD) and node crossing delay (NCD) were analyzed using a digital-optical system-on-package module architecture. The MAD bottleneck was minimized by using a 3-D distributed shared memory field in which high speed OIs delivered data to and from each processor. The NCD bottleneck was resolved by the use of non-blocking, high speed, optical switching, each switch having a local controller that was optically synchronized and programmed by an optical header signal. It was found that the coupling efficiency between laser and waveguide was around 45% and the coupling efficiency between waveguide with a 45° end mirror and the PiN PD with a36 μm diameter active area was around 35%.

AB - Data transport bottlenecks namely memory access delay (MAD) and node crossing delay (NCD) were analyzed using a digital-optical system-on-package module architecture. The MAD bottleneck was minimized by using a 3-D distributed shared memory field in which high speed OIs delivered data to and from each processor. The NCD bottleneck was resolved by the use of non-blocking, high speed, optical switching, each switch having a local controller that was optically synchronized and programmed by an optical header signal. It was found that the coupling efficiency between laser and waveguide was around 45% and the coupling efficiency between waveguide with a 45° end mirror and the PiN PD with a36 μm diameter active area was around 35%.

UR - http://www.scopus.com/inward/record.url?scp=14344264051&partnerID=8YFLogxK

M3 - 會議論文

AN - SCOPUS:14344264051

SN - 1092-8081

VL - 2

SP - 581

EP - 582

JO - Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS

JF - Conference Proceedings - Lasers and Electro-Optics Society Annual Meeting-LEOS

M1 - WO2

T2 - 2004 IEEE LEOS Annual Meeting Conference Proceedings, LEOS 2004

Y2 - 7 November 2004 through 11 November 2004

ER -

A digital-optical system-on-package module architecture for high performance multiprocessors

Abstract

Fingerprint

Cite this