Grant: Engage Grants Program
University of Saskatchewan
$25,000.00
Study of low power IEEE754 compliant arithmetic adder for scientific applications
FY 2011-12
Summary
Computer arithmetic is an important technology that affects many aspect of computing power. Addition is
often the timing critical path of modern microprocessors. A number of high-performance adders have been
proposed in the past. All of them implement a parallel adder structure using dynamic logic to achieve its high
performance. Those techniques all result in significantly more power consumption. Although the area may not
be a major issue as the transistor's size is scaled down, the heat dissipation could be unacceptable if the power
of the chip increases. The demand for low power floating point adder cores has been on the rise in recent years.
These units are essential building blocks of microprocessors and floating point DSP (data signal processing)
data paths. Since the hardware implementation of floating point addition involves the realization of a multitude
of distinct data processing sub-units that endure a series of power consuming transitions during the course of
their operations, the power consumption of floating point adders are, in general, quite significant in comparison
to that of their integer counterparts. Owing to the presence of a relatively high traffic of floating point additions
in microprocessors and DSPs, the power/performance implications of floating point adders directly impact the
power/performance desirability of the target application.
It is planned to use three data paths and transition activity scaling of functional units. We could provide
alternate path for common cases when there is no operation expected in a part of the unit. We also plan to
implement novel techniques (e.g. operand isolation and input-vector control) to reduce dynamic leakage and
static power dissipation. All the design is to be synthesized and optimized using Synopsys tools such as Design
compiler and Power compiler. Evaluation metric could be either a power usage or the power-delay product.
This research project investigates power and complexity-aware floating point adder design, analysis, and
validation. The output of this research project should be of critical and immediate importance to the IBM
Power series processor focus.
Program
Engage Grants Program
Recipient
Location
Saskatchewan, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
428446-2011
Name nom
Ko, SeokBum
Department département
Electrical & Computer Engineering
Organization id
16
Institution établissement
University of Saskatchewan
Province en
Saskatchewan
Province fr
Saskatchewan
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2011
Competition year année de concours
2011
Award amount
25000
Program id
EGP
Program nane en
Engage Grants Program
Program name fr
Programme de subventions d'engagement partenariat
Group en
Research partnerships programs
Group fr
Programmes de partenariat
Committee code
1555
Committee name en
Prairie Internal Decision Committee
Committee name fr
Comité de décision interne pour les Prairies
Area of application code
1207
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Information, computer and communication technologies
Area of application fr
Technologie de l'information, des ordinateurs et des réseaux de communications
Research subject code
2701
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Computer hardware
Research subject fr
Matériel informatique
Partie
1
Nb partie
1
Application title
Study of low power IEEE754 compliant arithmetic adder for scientific applications
Keyword
NSERC
Application summary
Computer arithmetic is an important technology that affects many aspect of computing power. Addition is
often the timing critical path of modern microprocessors. A number of high-performance adders have been
proposed in the past. All of them implement a parallel adder structure using dynamic logic to achieve its high
performance. Those techniques all result in significantly more power consumption. Although the area may not
be a major issue as the transistor's size is scaled down, the heat dissipation could be unacceptable if the power
of the chip increases. The demand for low power floating point adder cores has been on the rise in recent years.
These units are essential building blocks of microprocessors and floating point DSP (data signal processing)
data paths. Since the hardware implementation of floating point addition involves the realization of a multitude
of distinct data processing sub-units that endure a series of power consuming transitions during the course of
their operations, the power consumption of floating point adders are, in general, quite significant in comparison
to that of their integer counterparts. Owing to the presence of a relatively high traffic of floating point additions
in microprocessors and DSPs, the power/performance implications of floating point adders directly impact the
power/performance desirability of the target application.
It is planned to use three data paths and transition activity scaling of functional units. We could provide
alternate path for common cases when there is no operation expected in a part of the unit. We also plan to
implement novel techniques (e.g. operand isolation and input-vector control) to reduce dynamic leakage and
static power dissipation. All the design is to be synthesized and optimized using Synopsys tools such as Design
compiler and Power compiler. Evaluation metric could be either a power usage or the power-delay product.
This research project investigates power and complexity-aware floating point adder design, analysis, and
validation. The output of this research project should be of critical and immediate importance to the IBM
Power series processor focus.