Received: 26 March 2019 Revised: 2 August 2019 Accepted: 8 August 2019
DOI: 10.1002/ett.3747
RESEARCH ARTICLE
An expandable topology with low wiring congestion for
silicon interposer-based network-on-chip systems
Sajed Dadashi
1
Midia Reshadi
1
Akram Reza
2
Ahmad Khademzadeh
3
1
Department of Computer Engineering,
Science and Research Branch, Islamic
Azad University, Tehran, Iran
2
Department of Computer Engineering,
Shahr-e-Qods Branch, Islamic Azad
University, Tehran, Iran
3
Iran Telecommunication Research
Center, Tehran, Iran
Correspondence
Midia Reshadi, Department of Computer
Engineering, Science and Research
Branch, Islamic Azad University, Tehran
14778 93855, Iran.
Email: reshadi@srbiau.ac.ir
Abstract
In 2.5D stacking technology, multiple chips have stacked side-by-side on a sili-
con interposer layer. The network-on-chip in the central processing unit (CPU)
layer makes it possible to connect processing cores to each other. The interposer
layer prepares the connection between the CPU cores and other chips such as
memory chip. The memory chip usually contains several segments stacked ver-
tically. The network-on-chip can be extended to the interposer layer to make use
of unused routing resources of the interposer. Applying an efficient topology and
deadlock-free routing algorithm on the CPU layer and interposer layer is essen-
tial. In this paper, a new topology and routing algorithm are proposed to use on
the CPU layer as well as on the interposer layer for creating a uniform inter-
connection network to decrease delay and power consumption. This topology is
scalable and can be used simply for extensive networks. Most of interposer layer
topologies are not scalable and have a lot of crossed links. Moreover, it is required
to increase the degree of routers connected to memory segments. The proposed
topology uses a few crossed links compared to other proposed interposer layer
topologies. There will be some similar small sub-networks that fairly connected
together with intermediate routers. Furthermore, this topology can be extended
in a hierarchical manner.
1 INTRODUCTION
There are two types of stacking technologies to integrate multiple silicon chips.
1
Three-dimensional (3D) stacking tech-
nology makes it possible to stack various dies together.
2
It takes multiple silicon chips and places one on top of the other.
3,4
The 3D stacking has some problems such as heat dissipation.
5
Another stacking approach is two-and-a-half-dimensional
(2.5D) stacking technology that places multiple chips side-by-side on a silicon interposer layer.
6
In fact, an interposer
is on top of the package of a 2.5D system. Different chips can be mounted on the interposer layer. The 2.5D approach
allows functional blocks to be designed separately.
7
This technology can decrease metal connection routes and can solve
many of 3D stacking problems. It prepares the capability for system components to be designed simultaneously.
8
Silicon
interposers can be used in many devices such as FPGAs to reduce design complexity. Applying this type of stacking can
improve efficiency and decrease the implementation cost.
9
Intermediate stages of packaging are no longer needed and
the system occupies less space. It is feasible to design the chips independently with reusable components.
10
The inter-
poser is actually a silicon chip with upward metal layers. Using an interposer layer has many benefits such as considerable
electrical property, low cost and high wiring capacity.
11
A many-core central processing unit (CPU) alongside a vertically
stacked memory chip is a common integration.
12
For example, the memory segments can be positioned on the right and
left sides of the CPU chip on the interposer layer.
13
Trans Emerging Tel Tech. 2019;e3747. wileyonlinelibrary.com/journal/ett © 2019 John Wiley & Sons, Ltd. 1 of 20
https://doi.org/10.1002/ett.3747