YDEL is the ideal OS for GPU computing in servers, embedded devices, workstation, or entire clusters. It provides a reliable CUDA environment that opens new opportunities to your business while saving engineering resources.

Unparalleled GPU Computing Integration

YDEL bundles the NVIDIA X11 Driver, CUDA Toolkit, SDK, and Visual Profiler. These packages are installed and pre-configured with YDEL which will save the end user from a tedious installation and setup of these components. Everything works out of the box.

When it comes time to update the NVIDIA Driver, Toolkit, and SDK, you will no longer need to perform the same tedious installation and setup that was performed when the system was initially deployed as in YDEL will update all of the NVIDIA GPU Computing components automatically, just like the basic Linux components included in the distribution

YDEL will reduce deployment time by 30%^*1 compared to other Linux distributions while also lowering TCO after the system is in use.

Greater Reliability

YDEL is designed and intended for use with GPUs, Fixstars offers a product where defects related to GPUs and HPC workloads are simply not tolerated unlike other solutions where GPU Computing support is simply added onto another distribution of Linux.

YDEL is not only tested with the standard Linux distribution tests such as LTP, but we also test a wide variety of CUDA and OpenCL enabled software on a variety of different graphics cards.

Unlike other distributions, the end user does not need to recompile kernel modules. When an updated kernel is made available, YDEL's software update tools will also download and install verified and tested drivers for the updated kernel.

Increased Flexibility

YDEL supports using multiple versions of CUDA and OpenCL. An end user can easily use cudatoolkit 3.0, 2.3, and 2.2 on the same machine and easily switch between them using configuration files or environment variables.

Different applications that require different versions of CUDA no longer require a unique installation, simplifying deployment and maintenance. An added benefit for developers, building and testing your application against different versions of CUDA has never been simpler.

Superior Performance

YDEL uses a modified RHEL kernel that features many refinements to enhance the performance of multicore workloads resulting in up to a 9% performance improvement.

Services and software packages that are typically not used are disabled or not installed by default, saving memory and CPU cycles for computational work.
» Benchmark(PDF:57KB)

Eclipse for CUDA

YDEL includes Fixstars CUDA plugin for Eclipse, allowing developers to use tools they are familiar with improving efficiency and reducing time to market.