Timeline


and .

09/13/10:

02:58 Ticket #37 (Support for "combining" operators) created by blever
The ability to combine multiple arrays into a single one can currently …

09/11/10:

03:54 Ticket #36 (scan operations hang indefinitely on devices with Compute Capability 1.0) created by seanl
scan_intervals, in a for-loop, has a __syncthreads() and calls …

09/09/10:

03:56 WikiStart edited by tmcdonell
(diff)

09/03/10:

00:47 Ticket #29 (code generation fails for tuples as the result of a branch operation) closed by tmcdonell
fixed: […]

09/02/10:

01:13 Ticket #35 (Inclusive scans/folds) created by blever
The scan/fold functions defined within the Accelerate language are all of …

08/28/10:

05:57 Ticket #34 (CUDA backend does not support 'stencil') created by tmcdonell

08/24/10:

12:37 Ticket #33 (CUDA.run should be pure) created by chak
Currently, we have[…]

08/23/10:

08:43 Ticket #32 (OpenCL Backend) created by anonymous
Since OpenCL is a standard which works across Graphics-Hardware Vendors, a …
01:51 Ticket #31 (support concurrent kernel execution on Fermi architectures) created by tmcdonell
01:20 Ticket #30 (CUDA backend does not support 'shape') created by tmcdonell

08/22/10:

09:55 Ticket #29 (code generation fails for tuples as the result of a branch operation) created by tmcdonell
For example:[…]
08:44 Ticket #28 (reduced performance of small types) created by tmcdonell
CUDA devices do not coalesce memory transfers to global memory of 8- and …
Note: See TracTimeline for information about the timeline view.