The MainConcept EVA (Easy Video API) ecosystem unifies hardware-accelerated and software-based video processing across commercial off-the-shelf (COTS) hardware through one API. Our longstanding collaborations with major CPU and GPU designers offer benefits for broadcasters, OTT streaming service providers, and production environments that work with media processing on a daily basis. With Codec SDK 17.0, EVA now extends support for full end-to-end transcoding pipelines to NVIDIA GPUs on Linux Arm and NETINT VPUs on Linux x86.
These extensions bring decoding, scaling, color conversion, and encoding stages together in complete hardware-accelerated workflows, giving developers more flexibility when building and deploying video processing applications.
Why MainConcept EVA?
MainConcept Easy Video API (EVA) combines hardware acceleration and software codec capabilities through a single encoder and/or decoder library via a unified API. Through partnerships with NVIDIA, Intel, AMD, Qualcomm, and NETINT, EVA provides access to supported hardware encoders and decoders in the user’s pipeline without requiring separate vendor SDK integrations, reducing development costs and maintenance effort. EVA enables MainConcept software and hardware codecs for AVC/H.264, HEVC/H.265, JPEG XS, and AV1, as well as color conversion and scaling, through a unified API on Windows and Linux. These capabilities vary by platform and configuration. A key advantage of EVA is complete on-GPU transcoding pipelines. In supported NVIDIA GPU workflows on Linux x86, GPU-based color conversion and scaling are integrated with the corresponding encoders and decoders, minimizing frame copies and transfers between CPU and GPU memory.
Full End-to-End Processing for NVIDIA and NETINT in Codec SDK 17.0
MainConcept’s AVC/H.264 and HEVC/H.265 SDKs extend EVA’s end-to-end hardware-accelerated transcoding to NVIDIA Blackwell GPUs on Linux Arm and NETINT VPUs on Linux x86. Available decoding, scaling, color conversion, and encoding stages run on the hardware, minimizing memory transfers. For unsupported profiles, applications can select MainConcept software codecs for professional formats such as Sony XAVC and Panasonic P2 AVC-ULTRA, subject to configuration. The setup is quick and simple: developers integrate one MainConcept low-level library through a unified API, with MainConcept as their single point of contact for integration and support.
NVIDIA GPUs on Linux Arm
Codec SDK 17.0 extends MainConcept’s existing full on-GPU AVC/H.264 and HEVC/H.265 pipelines on Linux x86 to Linux Arm platforms with NVIDIA Blackwell GPUs. This expands deployment options for cloud and on-premises transcoding infrastructure, including systems built around Arm processors such as Ampere Altra and AWS Graviton, as well as NVIDIA DGX Spark systems featuring integrated Blackwell graphics. For professional formats, hardware capability alone does not guarantee compliance. Although NVIDIA NVENC on Blackwell supports 10-bit 4:2:2 encoding, this does not by itself establish compliance with Sony XAVC or Panasonic P2 AVC-ULTRA. Where the required format is unsupported by hardware, an application can switch to the MainConcept software encoder, provided the required configuration is supported. Existing MainConcept AVC/H.264 and HEVC/H.265 SDK licensees can extend supported pipelines to Linux Arm without an additional EVA charge, subject to the applicable licensing terms. Exposing your workflows to MainConcept EVA technology opens up new business opportunities, as products and services can be easily extended to new hardware and platforms, making them more flexible.
NETINT VPUs on Linux x86
With MainConcept’s latest release, NETINT Quadra VPU boards now cover the full transcoding pipeline for AVC/H.264 and HEVC/H.265, including scaling and color conversion on Linux x86. Quadra VPUs offer high encoding density and low power consumption for cloud and on-premises broadcast delivery and OTT streaming. Hardware supports 4:2:0 8-bit and 10-bit processing up to 8K, while MainConcept software codecs support 4:2:2 10-bit configurations. All these benefits add up to remarkable cost savings and make MainConcept EVA a no-brainer for companies that already deploy MainConcept SDK-based use cases and want to minimize their ongoing costs at no additional charge. Running on Linux x86 platforms, MainConcept EVA and NETINT Quadra boards are a perfect match for next-generation broadcast delivery and OTT streaming pipelines.
What’s Next?
Media workflows prioritize different needs, from performance in live and VOD applications to output quality and format compliance. MainConcept EVA keeps your workflow highly flexible and addresses these needs through a unified API, combining supported NVIDIA, AMD, Intel, Qualcomm and NETINT hardware with MainConcept software codecs when hardware lacks the required capabilities.
The benefits of MainConcept EVA are obvious: integrating a single library using a single API saves your development team a significant amount of time that can be spent on adding features or generally improving your product or service. This is great for your budget because, by incorporating the MainConcept SDK instead of one SDK from each GPU or VPU vendor, developers can reduce integration effort and costs by more than 80% and focus instead on product innovation.
And more support is on the way! With full end-to-end on-GPU/VPU transcoding pipelines now available for NVIDIA and NETINT hardware, support for AMD and Intel GPU is next on the EVA roadmap. MainConcept’s goal is to simplify hardware integration so developers can focus on their products while EVA handles the complexity.
To evaluate the latest extensions of MainConcept EVA within Codec SDK 17.0, download the free evaluation version from the MainConcept website and start testing today.
