Solo testo/Text only

Microarchitecture: foundations and advanced topics

Prof. Yale N. Patt
Electrical and Computer Engineering Department, The University of Texas at Austin, Austin, Texas

15 hours, 4 credits

July 8 - July 13, 2009

Dipartimento di Ingegneria dell'Informazione: Elettronica, Informatica, Telecomunicazioni, Largo Lucio Lazzarino (formerly via Diotisalvi), meeting room

Contacts: Ing. Pierfrancesco Foglia

Abstact

Using computers to solve problems requires starting with a natural language formulation of the problem and systematically transform it until one has a machine language (ISA) specification of the problem (i.e., a program). This then is executed on the implementation hardware. Application specialists continue to think up more applications for computers. Moore's Law continues to provide more transistors on a chip (50 billion transistors running at a clock speed of greater than 10 GHz in a few years). The ISA is the interface between the software that produces the program and the hardware that carries it out. The ISA is implemented by a microarchitecture that is constrained by trade-offs such as performance, power consumption, cost, reliability, availability, etc. In this series of lessons, we will examine these tradeoffs, in light of the increasing transistor count and the fundamentals of microarchitecture. Our study will be grounded in fundamental principles of microarchitecture, but will involve current state-of-the-art approaches.

Syllabus

Architecture vs. microarchitecture
Alternatives in the ISA
Alternatives in the microarchitecture
Alternatives in the computer system (including memory trade-offs)

Pipelining

Branch prediction

Trace cache

SMT and SSMT
L2-miss activities (run-ahead execution, cache enhancements, etc.)
Other paradigms

Block-structured ISA
Predicated execution
Wish branches

Approaches (Larrabee, Barcelona, ACMP)
Tradeoffs
Case studies and their implications