The São Paulo Research and Analysis Center (SPRACE) is dedicated to the study of different aspects of High-Energy Physics (HEP) – the field of physics that explores the elementary particles, which are the fundamental constituents of matter, and their interactions. We are also involved in the development of important spinoffs that go beyond the boundaries of HEP, particularly in the fields of high-performance computing and scientific instrumentation.
We are seeking highly qualified software engineers to work on the development of the High-Level Trigger for the Compact Muon Solenoid (CMS) Phase-II Upgrade. CMS is one of the experiments that analyse data from the CERN Large Hadron Collider (LHC), the most advanced particle accelerator in operation. The successful applicant is expected to work in close collaboration with the Center for Scientific Computing (NCC) of the São Paulo State University (Unesp).
The CMS experiment is equipped with a Trigger and Data Acquisition System designed to inspect the detector information at the full LHC crossing frequency of 40 MHz and to select events for archiving and later offline analysis. The system is divided into a Level-1 Trigger (L1T), an FPGA-based system that is designed to reduce the rate of events accepted for further processing to less than 100 kHz; and a High-Level Trigger (HLT), an asynchronous software-based suite of algorithms designed to reduce this maximum Level-1 Accept rate of 100 kHz to a final output rate of approximately 1 kHz.
The High-Luminosity LHC, a proposed upgrade of the LHC, will entail a twenty-fold increase of the produced data, bringing the expected L1T and HLT rates to ~750 kHz and ~7.5kHz, respectively. The current HLT architecture for that case does not scale in an acceptable manner if running on standard clusters, composed of commodity servers with CPUs. An alternative approach is the usage of heterogeneous computing techniques, identifying parts of the code that can run efficiently on "accelerators" (mainly GPUs) and porting them to the new architecture.
The successful candidate is expected to:
Generally take part on updating the HLT code for the Phase-II.
Identify and port algorithms that are most compatible with GPUs.
Take part on the optimisation of the algorithms, both in terms of efficiency and speed.
Work on a validation suite for the ported algorithms.
Integrate the algorithms in the existing software framework used by the experiment.
Be available to sporadic travels to CERN.
The successful candidate must have, at least, a Ph.D. in Computer Science / Software Engineering, or a B.Sc. and at least five years of experience in the industry.
The skills and qualifications we are looking for include, but are not restricted to:
Excellent leadership and communication skills;
Strong knowledge in object-oriented programming;
Strong knowledge in multithreaded and thread-safe programming;
Strong knowledge in software development for GPUs;
Some experience in pattern recognition problems;
Some experience in high-performance / high-throughput computing;
Proficiency in modern C++ (at least C++11);
Proficiency in English;
Experience in Python, both 2.7 and 3.X;
Strong motivation to participate in challenging R&D projects.
Earlier experiences on the following subjects are considered a bonus:
Data Acquisition Techniques;
Experimental High-Energy Physics.
The position is sponsored by the São Paulo Research Foundation (FAPESP), a public foundation, funded by the taxpayer in the State of São Paulo, with the mission to support research projects in higher education and research institutions in all fields of knowledge. This appointment is initially for fourteen months.
The monthly income is established accordingly to the academic qualifications and previous experience of the candidate, as per the FAPESP guidelines. This position qualifies for a Technical Training Fellowship V, with a yearly income of 88468.80 BRL, roughly equivalent to 23500 USD. In general, the income will be compatible to the equivalent level of a fellowship from the São Paulo Research Foundation (FAPESP).
Applications should be addressed to Prof. Thiago Tomei, with the word “HLT-TT5” in the Subject line.
The deadline is 2019-08-30. We reserve the right not to make an appointment and continue searching after the closing date. Only shortlisted candidates will be contacted. Shortlisted candidates must be available for an interview.
SPRACE is committed to the principles of equal opportunity, inclusivity, and diversity. Women remain underrepresented in STEM careers and we encourage their application for SPRACE positions. We welcome applications from qualified candidates with any sexual orientation and members of minority groups.