![Template vector](https://kumkoniak.com/110.jpg)
Experimental support for ARM AVE and an older version of the backend for ARM Neon can be found in the other branches. Currently we support the most popular Intel instruction sets (sse, avx2, and avx512) as well as the SX Aurora Tsubasa accelerator (intrinsic based backend). The mapping to a specific hardware is done at compile time depending on a template argument, which we call processing style. The TVL is a C++ library, which offers hardware-oblivious SIMD-programming. However, the plan is to use the code from this repository in MorphStore, too, in the not so distant future. This repository may be slightly behind either of these branches. MorphStore still has an integrated version in the master branch and a TVL branch for testing purposes. This is a stand-alone version of the Template Vector Library. The old sources are available in the old_master branch Welcome to the Template Vector Library (TVL)! If you rely on the old interfaces or on the neon support, do not checkout the main branch.
![template vector template vector](https://static.vecteezy.com/system/resources/previews/000/276/750/original/vector-template-set.jpg)
Neon support temporarily not available (we are working on bringing it back).Added support for NEC SX Aurora Tsubasa.More base types and mask-primitives supported.Old default branch renamed to old_master.Any hints to unclear explanations are welcome.
![template vector template vector](https://static.vecteezy.com/system/resources/previews/000/219/597/original/vector-template.jpg)
![template vector template vector](https://i.pinimg.com/736x/96/ee/34/96ee34eef6ab08382683f98395fa0074--abstract-template-templates-free.jpg)
() Readme UPDATE! Please read this, especially if you are migrating from the old interface.
![Template vector](https://kumkoniak.com/110.jpg)