Graduate Certificate in High Performance Computing
The Graduate Certificate in High Performance Computing is intended for graduate students in any discipline who have significant prior computational experience. The purpose of this certificate is to complement students' degree programs with a set of courses that provide students with a broad exposure to parallel computing methodology, and give them experience with computational and data science challenges that require parallel and/or high-performance computing in order to solve effectively. Students that have completed this certificate will be able to:
- Demonstrate a high-level understanding of functional and object-oriented computer programming as applied to a range of problems in computational and data science.
- Analyze problems in terms of the algorithms and pre-existing computational tools required to solve a range of problems in computational and data science, and write a program to efficiently solve the problem on modern parallel computers and specialized hardware (e.g., graphics processing units).
- Construct and implement models of a variety of systems using modern parallel programming techniques and software development techniques, and use these models/simulations to gain understanding of these systems.
- Apply some subset of discipline-focused or methodology-focused topics in computational and data science to solve problems in the student’s primary discipline.
The Graduate Certificate in High Performance Computing consists of at least three courses comprising a minimum of 9 credit hours, taken from the two categories listed below. The targets of the certificate program are graduate students in any discipline with interest in applying computational and data science approaches that require parallel and/or high-performance computing to their research problems, or who generally desire an education in parallel computational methodology. To facilitate this goal, in addition to there being no restriction on graduate student discipline, students can apply for the certificate at any time prior to receiving their degree (either Master’s or PhD), and can apply for the certificate after taking all the necessary courses.
Note that credit from courses whose focus is largely or primarily an introduction to programming and/or basic numerical methods (i.e., CMSE 801, CMSE 802, CSE 801, or other comparable courses) will not count for credit toward this certificate. In addition, 400-level computational coursework may not count for credit toward this certificate without the prior permission of the CMSE graduate director. The primary circumstance where a 400-level course may be acceptable for credit toward this certificate program is when an equivalent 800-level course is unavailable (e.g., a highly specialized 400-level combined undergraduate and graduate course.) Students that have questions about any particular course are strongly encouraged to consult the CMSE Graduate Director.
The two categories of courses include:
- CMSE/CSE-822, Parallel Computing (3 credits)
- Two or more additional courses, which may include further CMSE courses at the 800 level or above, courses from an approved list of non-CMSE courses (found in the CMSE graduate handbook), or any other 800- or 900-level computational science or data science-focused courses as approved by the CMSE graduate advisor (6 or more credits).
Note: In order to obtain this graduate certificate the student must have at least a 3.0 average in the courses that are applied to the certificate.
If you have any questions, please contact the CMSE Graduate Director.