| EPSRC Reference: |
GR/R55252/01 |
| Title: |
Differentiation Enabled Compiler Technology |
| Principal Investigator: |
Naumann, Professor U |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Computer Science |
| Organisation: |
University of Hertfordshire |
| Scheme: |
Standard Research (Pre-FEC) |
| Starts: |
02 January 2002 |
Ends: |
01 July 2003 |
Value (£): |
51,612
|
| EPSRC Research Topic Classifications: |
| Fundamentals of Computing |
|
|
| EPSRC Industrial Sector Classifications: |
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
Numerous numerical methods in simulation and optimisation of real-world processes require sensitivity (derivative) information for their outputs with respect to certain parameters. Nowadays most scientists and engineers use numerical differentiation by evaluating finite difference quotients to get this information. This approach is often inefficient and its accuracy depends strongly on the choice of the corresponding step-size. In the case where the underlying mathematical modell is given as a computer program (e.g. written in Fortran 95) Automatic Differentiation (AD) may be used to compute derivatives efficiently with machine accuracy. Existing AD-tools require the user to play a considerable role in the process of setting up the necessary configuration. The aim of this project is to explore ways of easing the burden which AD places upon the user, as a vehicle for exploring the extent to which recent theoretical results can be exploited to ensure maximal efficiency.A Differentiation Enabled Compiler will limit the work to be carried out by the user to a minimum by providing new datatypes, automatic type propagation, exploitation of both structural and runtime sparsity,and easy-to-use interface routines for extracting the derivatives. Consequently, the user will no longer have to care about how derivatives are computed any more than he should be concerned about how the evaluation of an intrinsic function (such as cosine or exponential) is actually implemented.
|
|
Key Findings |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Potential use in non-academic contexts |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Impacts |
|
| Description |
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk |
| Summary |
|
| Date Materialised |
|
|
|
|
Sectors submitted by the Researcher |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
| Project URL: |
|
| Further Information: |
|
| Organisation Website: |
http://www.herts.ac.uk |