https://flagos-ai.github.io/FlagGems/overview/Recent content in Overview on FlagGemsHugoenhttps://flagos-ai.github.io/FlagGems/overview/features/Mon, 01 Jan 0001 00:00:00 +0000https://flagos-ai.github.io/FlagGems/overview/features/<!-- Copyright 2026 FlagOS Contributors Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. --> <h1 id="features-overview">Features Overview<a class="anchor" href="#features-overview">#</a></h1> <h2 id="rich-collection-of-operators">Rich Collection of Operators<a class="anchor" href="#rich-collection-of-operators">#</a></h2> <p>FlagGems features a large collection of PyTorch compatible operators. Refer to <a href="https://flagos-ai.github.io/FlagGems/references/operators/">operator supported</a> and <a href="https://flagos-ai.github.io/FlagGems/references/experimental/">experimental operators</a> for list of formally supported operators and experimental operators.</p>https://flagos-ai.github.io/FlagGems/overview/platforms/Mon, 01 Jan 0001 00:00:00 +0000https://flagos-ai.github.io/FlagGems/overview/platforms/<!-- Copyright 2026 FlagOS Contributors Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. --> <h1 id="platforms-supported">Platforms Supported<a class="anchor" href="#platforms-supported">#</a></h1> <p><em>FlagGems</em> supports 10+ backends. The currently supported platforms and their capabilities are listed blow:</p>https://flagos-ai.github.io/FlagGems/overview/pointwise-dynamic/Mon, 01 Jan 0001 00:00:00 +0000https://flagos-ai.github.io/FlagGems/overview/pointwise-dynamic/<!-- Copyright 2026 FlagOS Contributors Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. --> <h1 id="pointwise-dynamic-operators">Pointwise Dynamic Operators<a class="anchor" href="#pointwise-dynamic-operators">#</a></h1> <h2 id="1-pointwise-operations">1. Pointwise operations<a class="anchor" href="#1-pointwise-operations">#</a></h2> <p>Pointwise operators are trivial to parallelize. Most parallel programming guides begin with pointwise addition between 2 contiguous vectors. For <a href="https://triton-lang.org/main/getting-started/tutorials/01-vector-add.html#sphx-glr-getting-started-tutorials-01-vector-add-py"><code>vector_add</code> in Triton</a>, it is simple to implement a task partitioning schema that each CTA reads a contiguous range from each input vector and writes to a contiguous range of the output vector.</p>