v0.9.8 - cudnn half fixes

What's Changed

• Add ordinal to CudaDevice struct by @tthebst in #126
• added complex build workflow example by @l3utterfly in #131
• Fixing feature guard for cudnn half data types by @coreylowman in #134
• Fixing scalar parameters for f16 by @coreylowman in #136

New Contributors

• @tthebst made their first contribution in #126
• @l3utterfly made their first contribution in #131

Full Changelog: v0.9.7...v0.9.8

v0.9.6

What's Changed

• Adding Conv2d APIs to safe cuDNN by @coreylowman in #116
• Add cudnn error msg by @coreylowman in #118
• Add cudnn reductions apis by @coreylowman in #119

Full Changelog: v0.9.5...v0.9.6

v0.9.3 - Static linking, view slicing, runtime module names, multi threading usage

What's Changed

• Static linking (with feature flag). by @Narsil in #105
• Add inline tags to launches by @coreylowman in #107
• Fix RangeHelper::bounds's handling of zero-length ranges by @nkoppel in #109
• Support runtime names of modules by @coreylowman in #110
• Enable slicing on CudaView/CudaViewMut by @coreylowman in #111
• Adds CudaDevice::bind_to_thread and example of multi-threaded use by @coreylowman in #112
• Improving docs by @coreylowman in #114
• Adds CudaTypeName by @coreylowman in #115

New Contributors

• @nkoppel made their first contribution in #109

Full Changelog: v0.9.2...v0.9.3

v0.9.2 - transmutes & cudnn sys api

What's Changed

• Cudnn_base2 by @M1ngXU in #97
• Adding transmute/transmute_mut by @coreylowman in #100
• Adds DevicePtr for CudaViewMut by @coreylowman in #103

Full Changelog: v0.9.1...v0.9.2

v0.9.0 - Feature flags, better streams, bug fixes

What's Changed

• Removing dep on find-cuda-helper. Moving no-std behind feature flag by @coreylowman in #92
• Adds feature flags for each of the parts of cudarc by @coreylowman in #93
• Reverting free stream, putting free_async calls on default stream by @coreylowman in #94
• More sound CudaStream by @coreylowman in #95

Full Changelog: v0.8.0...v0.9.0

v0.8.0 - spring cleaning

What's Changed

• Reorganizing cudarc::driver::safe by @coreylowman in #83
• Adding DeviceRepr, removing AsKernelParam by @coreylowman in #84
• Consistent & clear naming convention by @coreylowman in #85
• Using DevicePtrMut and DevicePtr for copies by @coreylowman in #86
• Removing CudaDeviceBuilder by @coreylowman in #87
• Simplifying Ptx functions by @coreylowman in #88

Full Changelog: v0.7.5...v0.8.0

v0.7.5 - safe multi stream support

What's Changed

• Adding CudaStream & LaunchAsync::par_launch_async by @coreylowman in #82

Full Changelog: v0.7.4...v0.7.5

v0.7.4

Forgot to impl ValidAsZeroBits for i128/u128 😅

Full Changelog: v0.7.3...v0.7.4

v0.7.3

What's Changed

• Adding cuda Profiler API. by @Narsil in #76
• Adding event apis by @coreylowman in #78
• Adding a free stream to concurrently free memory by @coreylowman in #79
• Adding memset_zeros_async by @coreylowman in #81

New Contributors

• @Narsil made their first contribution in #76

Full Changelog: v0.7.2...v0.7.3

v0.7.2 - default stream usage

What's Changed

• Using default null stream instead of creating a stream by @coreylowman in #75

Full Changelog: v0.7.1...v0.7.2