Synergetic Growth of AI and Silicon Age 4.0 through Heterogeneous Integration of Technologies
11:00 - 12:30 pm
IAS Lecture Theater, Lo Ka Chung Building, Lee Shau Kee Campus, HKUST

Abstract

Over the last five decades, the semiconductor industry has grown tremendously, from its infancy to now reaching over US$350 billion in annual revenue. This growth has been underpinned by two intelligent rules: one is an effective Return-on-Investment economic rule based on Moore's Law, which dictates that the number of transistors on a silicon chip should double every two years. The other rule is a design guideline, namely, scaling transistor line width to achieve enough productivity in lower power-delay products. Now that the line width has shrunk from several micrometers to 10 nanometers, the critical question the integrated-circuit industry faces is how far can both rules sustain? This talk proposes two points: firstly, a new observation on how the industry can continue to be very efficient and keep Moore's Law effectively for the 10 nanometer technology node. Secondly, this talk makes a bold prediction that we will see a new guideline, which can be called Virtual Moore's Law Economy (VME) and achieve effective economics beyond the 5 nanometer node or even with an equivalent 3 nanometer node through a new scaling method described as 3D x 3D Microsystem Volume-Scaling for Heterogeneous Integration!

During these past five decades, the IC industry has aggressively pushed for progress and fortunately has also been pulled forward by killer applications, from mainframe computers and telecom switch machines, to PCs and notebooks, to mobile phones and today's smartphones. Now worries have arisen on whether any killer applications will emerge that can sustain semiconductor demand in the future. This talk describes more new killer applications which will continue healthy growth for the semiconductor industry, with values created by heterogeneous integration technology scaling. An emerging era named as Semiconductor-Intelligence Paradigm (SIP), however, will not only be created by scaling devices down, but also by generating more smarter integrations whereby more new technologies will focus on scaling versatile functions up!

 

Image removed.

Related Link:

-https://www.seng.ust.hk/web/eng/events_detail.php?id=307

日期
時間
11:00 - 12:30 pm
地點
IAS Lecture Theater, Lo Ka Chung Building, Lee Shau Kee Campus, HKUST
活動形式
講者/ 表演者:
Dr. Nicky Lu
語言
英文
主辦單位
電子及計算機工程學系
聯絡方法

For enquiries, please email to sengpr@ust.hk or call at 2358 8045.

新增活動
請各校內團體將活動發布至大學活動日曆。