DSL's: Difference between revisions
From Modelado Foundation
imported>Dquinlan No edit summary |
imported>Dquinlan No edit summary |
||
| Line 17: | Line 17: | ||
|- style="vertical-align:top;" | |- style="vertical-align:top;" | ||
|DSL 1 | |DSL 1 | ||
| | |Halide | ||
| | |D-TEC | ||
| | |http://halide-lang.org | ||
| | |Image processing algorithms | ||
| | |Cloverleaf, miniGMG, boxlib | ||
| | |Uses C++ | ||
| | |Custom IR | ||
| | |Stencil optimizations (fusion, blocking, parallelization, vectorization) Schedules can produce all levels of locality, parallelism and redundant computation. OpenTuner for automatic schedule generation. | ||
| | |LLVM | ||
| | |X86 multicores, Arm and GPU | ||
| | |Working system. Used by Google and Adobe. | ||
| | |Local laplacian filter: Adobe top engineer took 3 months and 1500 loc to get 10x over original. Halide in 1-day, 60 lines 20x faster. In addition 90x faster GPU code in the same day (Adobe did not even try GPUs). Also, all the pictures taken by google glass is processed using a Halide pipeline. | ||
|- style="vertical-align:top;" | |- style="vertical-align:top;" | ||
|''Associated X-Stack Project | |''Associated X-Stack Project | ||
Revision as of 16:07, April 29, 2014
Sonia requested that Saman Amarasinghe and Dan Quinlan initiate this page. For comments, please contact them. This page is still in development.
| DSLs | Name of the DSL | Associated X-Stack Project | URL | Target domain | Miniapps supported | Front-end technology used | Internal representation used | Key Optimizations performed | Code generation technology used | Processors computing models targeted | Current status | Summary of the best results |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DSL 1 | Halide | D-TEC | http://halide-lang.org | Image processing algorithms | Cloverleaf, miniGMG, boxlib | Uses C++ | Custom IR | Stencil optimizations (fusion, blocking, parallelization, vectorization) Schedules can produce all levels of locality, parallelism and redundant computation. OpenTuner for automatic schedule generation. | LLVM | X86 multicores, Arm and GPU | Working system. Used by Google and Adobe. | Local laplacian filter: Adobe top engineer took 3 months and 1500 loc to get 10x over original. Halide in 1-day, 60 lines 20x faster. In addition 90x faster GPU code in the same day (Adobe did not even try GPUs). Also, all the pictures taken by google glass is processed using a Halide pipeline. |
| Associated X-Stack Project | ||||||||||||
| URL | ||||||||||||
| Target domain | ||||||||||||
| Miniapps supported | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) | ||
| Xstack projects involved | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) | ||
| Internal representation used | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) | ||
| Key Optimizations performed | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) | ||
| Code generation technology used | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) | ||
| Processors/computing models targeted | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) | ||
| Current status | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) | ||
| Summary of the best results | (EXPRESS) | (TG) | (DEGAS) | (D-TEC) | (DynAX) | (X-TUNE) | (GVR) | (CORVETTE) | N/A | (PIPER) |