Amber Group is a global leading digital asset company providing crypto financial services to both institutional and high-net-worth investors globally. 

We offer best-in-class liquidity solutions and cutting-edge trading infrastructure across major exchanges, applications, and networks. With over $1 trillion in cumulative trading volume, our deep liquidity helps power the digital asset ecosystem. 

Beyond trading, our full-suite of offerings includes wealth management, lending and investing products. But at our core, we focus on building strong relationships and delivering personalized service to help clients navigate this fast-growing industry.

At Amber, security is our #1 priority. We have invested years of effort and millions of dollars in cybersecurity, crypto-security, and operational security across the firm, with industry-leading certifications like SOC 2 Type II and ISO 27001.

Powered by a 400+ team of traders, technologists and engineers operating 24/7 globally, our technology and research capabilities are world-class. Yet we remain entrepreneurial, always seeking fresh ideas and risks worth taking. We are always interested in people who have an appetite for taking calculated risk, demonstrate a high level of original thinking and intellectual curiosity.  

Responsibilities:

• Design and develop high-frequency quantitative trading systems using GPU acceleration, and build a high-performance AI strategy platform.
• Optimize the strategy model performance on GPUs to minimize latency and ensure efficient real-time execution of trading strategies.
  
• Research and apply advanced GPU parallel computing technologies to enhance trading strategy performance and accelerate system iterations.

Qualifications:

• Bachelor's degree or higher in Computer Science, Software Engineering, or related fields, with a solid foundation in computer systems.
• Proficient in at least two programming languages and expert in one, e.g., C++, Python.
  
• Practical experience with CUDA programming and GPU architecture, including performance optimization and familiarity with related tools, e.g., CUDA libraries and debugging/analysis tools.
  
• Understanding of deep learning principles, familiarity with common deep learning frameworks (such as TensorFlow, PyTorch), and experience in model inference deployment is preferred.
  
• Strong problem-solving and communication skills, excellent team collaboration spirit, and a passion for technological innovation.