Research Buffet
Monday, August 27, 2012
3147 MEB
Lecture 3:40 p.m.
Matt Might
University of Utah
Title: Securing Software with Static Analysis
Monday, August 27, 2012
3147 MEB
Lecture 3:40 p.m.
Matt Might
University of Utah
Title: Securing Software with Static Analysis
Abstract
The cost of cyber-insecurity grows exponentially each year. Neither government nor industry is immune. Even the U.S. military reports tens of thousands of intrusions into its networks daily. Businesses are routinely compromised, with secrets and sensitive data exfiltrated by the terabyte. Best practices in security--the "patch and pray" model--have failed. Many problems in cybersecurity can be traced to the fact that most software ships with security vulnerabilities. So, why *is* software shipped with vulnerabilities? Software is vulnerable because programmers make mistakes. Each vulnerability is a programmer mistake, e.g., failure to check a pointer, failure to check a bound, failure to check a result, failure to sanitize an input, failure to catch an exception. Clearly, the world needs a tool that can report whether or not a programmer made such a mistake. Unfortunately, Turing's halting problem prevents this "silver bullet" from existing. Fortunately, Turing's halting problem holds only in the limit. For many programs of human interest, the solution to the halting problem can be *approximated* with reasonable precision. The field within computer science charged with producing these "silver-plated bullets" is static analysis. Come learn more about how it's done, what the open challenges are and how it's impacting the future of cybersecurity.
Erik Brunvand
University of Utah
Title: Two Computing Degrees: Computer Engineering and Digital Media
The cost of cyber-insecurity grows exponentially each year. Neither government nor industry is immune. Even the U.S. military reports tens of thousands of intrusions into its networks daily. Businesses are routinely compromised, with secrets and sensitive data exfiltrated by the terabyte. Best practices in security--the "patch and pray" model--have failed. Many problems in cybersecurity can be traced to the fact that most software ships with security vulnerabilities. So, why *is* software shipped with vulnerabilities? Software is vulnerable because programmers make mistakes. Each vulnerability is a programmer mistake, e.g., failure to check a pointer, failure to check a bound, failure to check a result, failure to sanitize an input, failure to catch an exception. Clearly, the world needs a tool that can report whether or not a programmer made such a mistake. Unfortunately, Turing's halting problem prevents this "silver bullet" from existing. Fortunately, Turing's halting problem holds only in the limit. For many programs of human interest, the solution to the halting problem can be *approximated* with reasonable precision. The field within computer science charged with producing these "silver-plated bullets" is static analysis. Come learn more about how it's done, what the open challenges are and how it's impacting the future of cybersecurity.
Erik Brunvand
University of Utah
Title: Two Computing Degrees: Computer Engineering and Digital Media
Abstract
I'll describe two different Computing Degrees - the track in Computer Engineering, and the brand new track in Digital Media.
Computer Engineering is a discipline that combines elements of both Electrical Engineering and Computer Science. Computer engineers design and study computer systems at many levels from the circuits that make up computers, to the architecture of processors and subsystems, to the programming interfaces of those processors. Often there is a strong focus on embedded computing systems. This usually involves an interesting mix of software and hardware skills and the integration of both skills.
The Digital Media track is for students who are interested in tailoring their MS program to include some digital media content. This generally means allowing a broader range of electives from other departments that involve some sort of digital media area, where the term "digital media" is interpreted fairly broadly in this context. Students interested in the Digital Media track may be interested in computer games, graphics, character animation, kinetic art, HCI, information-visualization, computer animation, experimental media, and many other areas that have a media focus.
I'll describe two different Computing Degrees - the track in Computer Engineering, and the brand new track in Digital Media.
Computer Engineering is a discipline that combines elements of both Electrical Engineering and Computer Science. Computer engineers design and study computer systems at many levels from the circuits that make up computers, to the architecture of processors and subsystems, to the programming interfaces of those processors. Often there is a strong focus on embedded computing systems. This usually involves an interesting mix of software and hardware skills and the integration of both skills.
The Digital Media track is for students who are interested in tailoring their MS program to include some digital media content. This generally means allowing a broader range of electives from other departments that involve some sort of digital media area, where the term "digital media" is interpreted fairly broadly in this context. Students interested in the Digital Media track may be interested in computer games, graphics, character animation, kinetic art, HCI, information-visualization, computer animation, experimental media, and many other areas that have a media focus.