At Chicago UpDown (http://www.chupdown.com/), we are building breakthrough graph computing acceleration that can deliver 1000x performance over today's cloud-based approaches. To deliver it, we are developing novel programming tools and notations. With it real-time mapping and analytics of trillion-node graphs will become instant.
We need a talented Programming Language and Compiler Staff to refine programming notations for event-driven UpDown accelerator programs. These systems face the challenge of describing complex graph applications, managing extreme task and memory parallelism (mllion-fold). With extensive co-designed hardware support, these systems employ actor-like models, global addressing and efficient hardware task scheduling. Also critical is experience with Industry frameworks (LLVM) and code optimization, and capable with advanced AI tools to create advanced compilation tools. Key challenges include: managing thread and memory parallelism, privatization and exploitation of a scratchpad memory. We may also implement query optimization over graph data.
Key Responsibilities
* Design and extension of programming notation (actor-like) that exploits a distributed global address space, but also ability to control task and memory parallelism.
* Compiler Design and Development: Code generation and optimization, register and locality management, and data movement optimization for custom and standard hardware architectures (e.g., CPUs and Updown).
* Performance Profiling: Analyze system performance, identify bottlenecks at the instruction and memory levels, and implement compiler solutions
* Toolchain Maintenance: Maintain, extend, and upgrade existing compiler frameworks (such as LLVM, GCC, or proprietary toolchains) and ensure robust code quality.
Minimum Qualifications
* Education: Bachelors or Graduate degree in CS, CE, or related technical field.
* Experience: 5 years of experience writing production-quality multi-threaded code in C/C++.
* Compiler/Framework Experience: 5 years hands-on language/library design and work within compiler infrastructures (eg LLVM, MLIR, or GCC).
* Hardware/Architecture: Strong grasp of computer architecture, parallelism, and SIMD.
Preferred Qualifications
* Experience with domain-specific languages (DSLs), Actor models (eg. Erlang), memory hierarchy optimization, and accelerators
* Experience with parallel programming models and APIs (e.g., Actor models, CUDA, OpenCL, SYCL, MPI).
* Proficiency in C++ and scripting languages.
To apply, send a cover letter and resume to [email protected]
About Chicago UpDown Computing
Chicago UpDown Computing, Inc. is creating accelerated computing for scalable, real-time graph analytics in the government intelligence, financial fraud, graphRAG, and network cybersecurity markets. With single-node demonstrated performance of 100x times that of CPU/GPU systems, scaling to 1000x and UpDown exploits massive thread and memory parallelism to achieve breakthrough performance.
UpDown is a spinout from the University of Chicago and Purdue, capitalizing on 4 years of US Government Intelligence ARPA (IARPA) funded research. While Chicago UpDown is in stealth mode https://www.chupdown.com/ technical information on the IARPA project is available from http://updown.cs.uchicago.edu/
Similar Jobs
.png)
What you need to know about the Chicago Tech Scene
Key Facts About Chicago Tech
- Number of Tech Workers: 245,800; 5.2% of overall workforce (2024 CompTIA survey)
- Major Tech Employers: McDonald’s, John Deere, Boeing, Morningstar
- Key Industries: Artificial intelligence, biotechnology, fintech, software, logistics technology
- Funding Landscape: $2.5 billion in venture capital funding in 2024 (Pitchbook)
- Notable Investors: Pritzker Group Venture Capital, Arch Venture Partners, MATH Venture Partners, Jump Capital, Hyde Park Venture Partners
- Research Centers and Universities: Northwestern University, University of Chicago, University of Illinois Urbana-Champaign, Illinois Institute of Technology, Argonne National Laboratory, Fermi National Accelerator Laboratory
