ATS-GPU-BASE 4.0 is a software library developed by AlazarTech that transfers data acquired by its family of PCI Express waveform digitizers to a CUDA-enabled Graphical Processing unit (GPU) at sustained transfer rates as high as 6.9 GB/s. Data is presented in GPU memory as a buffer queue. Expert-level GPU programmers can create very high performance custom kernels to manipulate this data using an easy-to-use application programming interface (API). Fundamentally, ATS-GPU-BASE solves the problem of transferring data from a waveform digitizer to a GPU at very high speeds.

One example of such high performance kernel is the optional ATS-GPU-OCT library (sold separately), which acquires data using ATS-GPU-BASE and then implements full OCT signal processing algorithm using CUDA kernels. ATS-GPU-OCT was benchmarked at up to 950,000 4K FFTs per second, demonstrating the power and efficiency of the ATS-GMA-BASE platform.

The number of possible applications of ATS-GPU-BASE is limitless. In fact, it is limited only by application requirements and the imagination of the programmer.

AlazarTech On
Google Scholar


  • Transfer A/D data to GPU at high speed
  • Up to 6.9 GB/s transfer rate for PCIe Gen 3 digitizer boards
  • Supports CUDA compute capability 3.0+
  • Designed to work with AlazarTech PCIe waveform digitizers
  • Optional OCT Signal Processing Library: ATS-GPU-OCT
  • Compatible with 64-bit Windows and 64-bit Linux
  • Users can create custom kernels for signal processing
  • More flexible than FPGA based DSP

Create custom kernels for signal processing

ATS-GPU-BASE includes an example program that demonstrates how to use the ATS-GPU-BASE library to transfer data from a waveform digitizer to a GPU, how to do simple data processing on the GPU, and how to transfer the processed data to host memory (RAM) using CUDA kernels. Expert-level GPU programmers can use this example program as a starting point to create their own kernels to do GPU-based DSP.

Special Features

Use it with ATS-GPU-OCT for out-of-the-box OCT imaging with very high-speed floating point FFT routines that have been optimized to provide the maximum number of FFTs per second, as well as Dispersion Compensation, Zero Padding, Log, and Windowing Functions.