Re: Compiler optimizations on GPU?

"Giridhar S" <>
Tue, 29 Apr 2008 20:51:30 -0400

          From comp.compilers

Related articles
Compiler optimizations on GPU? (shrey) (2008-04-29)
Re: Compiler optimizations on GPU? (Giridhar S) (2008-04-29)
Re: Compiler optimizations on GPU? (2008-04-30)
Compiler optimizations on GPU? (Inderaj Bains) (2008-04-30)
Re: Compiler optimizations on GPU? (Anton Lokhmotov) (2008-05-06)
| List of all articles for this month |

From: "Giridhar S" <>
Newsgroups: comp.compilers
Date: Tue, 29 Apr 2008 20:51:30 -0400
Organization: Compilers Central
References: 08-04-109
Keywords: optimize
Posted-Date: 30 Apr 2008 10:01:39 EDT

If you're talking of general-purpose computations on GPUs (GPGPU),
then many of the traditional optimizations are indeed useful and
necessary, since the code structure is general and not specialized
graphics code. Optimizations related to performance could be quite
important - strength reduction, induction variable optimizations, code
motion and so on. IPA may not be very important because most of the
compute-intensive "kernels" that run on the GPU are loops with a high
degree of data-parallelism and minimal control-flow branching (like in
function calls).

In fact, NVIDIA's compiler for CUDA uses the high-level optimizer from
the Open64 compiler framework for optimizing the general purpose code
before performing register allocation, scheduling and other
optimizations specific to the underlying GPU.

On Tue, Apr 29, 2008 at 1:03 PM, shrey <> wrote:
> Can any one give me some broad hints on what sort of compiler
> optimizations are important for GPU's especially those on mobile
> platforms? Do traditional heavy weight optimizations/analysis like
> IPA, PRE, loop transformations hold any usefulness ?


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.