Opis pracy

Project Description

Are you enjoying Machine Learning? Do you believe you can challenge state-of-the-art of AI technology applications? See the below call!
Luxoft, in partnership with AMD, is searching for outstanding, talented, experienced software architects and developers with AI and machine learning on the GPU experience with hands-on in GPU performance profiling to join the rapidly growing team in Gdansk. As a ML GPU engineer, you will participate in creation of real-time AI application supporting GPU compute, 3D graphics and video pipelines, build a solid background in Machine Learning with the most recent, state-of-the-art GPU and AI architectures, and collaborate with industry leaders.


Develop pytorch or tensor flow AI DL models, then port then to the GPU to DirectML or similar API and optimize for performance
Train ML models and assess the AI network accuracy in problem solving
Build GPU processing or rendering pipelines supported by the AI inferencing on the state-of-the-art hardware
Perform GPU profiling with tools for state-of-the-art hardware, investigate and fix performance & functional defects
Work with the recent research papers and publications, and propose state-of-the-art advances in the research area
Work with a distributed team in a CI environment
Share expert knowledge through trainings and internal documentation


Must have

Senior level experience in design and development of C/C++ applications
Extensive experience in Machine Learning using Python & one of Tensor Flow, Keras or PyTorch
CPU/GPU architecture knowledge, extensive experience with one of the 3D rendering APIs
Generic data processing, data structures and 3D algorithms
System/Linux programming, resource management
Basic knowledge of video encoding algorithms and state-of-the-art solutions
Good understanding of multi-threaded programming techniques and SIMD architecture

Nice to have

Experience in HW/SW performance analysis and optimization, best if including GPU performance profiling
GPU shader coding experience
Computer Vision or Text-to-speech experience and background is helpful
Experience/ know how of game engine solutions like Unreal Engine 4 or Unigine
Experience with ML APIs like Caffe, DirectML, ONNX, good understanding of Deep Learning process
Experience with GPU profiling tools, Linux performance tuning and VTune
Demonstrated ability to investigate software defects with debuggers like gdb
Experience working with GPU 3D or compute environments, like for example OpenGL, OpenCL, Vulcan, DirectX
Experience working with large volume high throughput data applications
Experience with GPU and CPU performance optimizations and debugging
Solid 3D math skills (linear algebra, vector math)
Understanding of video encoding technologies, H264, H265, AV1 is a huge plus


English: C1 Advanced