HPC Boston 2020
Boston's First HPC-centric student competition
HPC Boston is a 2-day competition for undergraduate students. Student teams design and build a small computing cluster that can execute typical HPC workloads. Teams will be scored on their ability to successfully run the provided workloads, as well as achieving at the best price/performance.
- Teams are comprised of 2-4 currently-enrolled undergraduate students. Each student team may have one faculty or graduate student advisor. Once the competition starts, students may not consult with advisors for any technical questions.
- All team members must be available to attend from 8AM to 8PM on February 15th and 8AM to 12PM on February 16th
- During the the competition, food will be provided to students. Teams are strongly encouraged to find sponsorship for their hardware and transportation costs.
Teams can choose to either bring their own hardware to HPC Boston or to use a provided Discovery cluster reservation.
If you are interested in using the Discovery cluster, please email firstname.lastname@example.org for the instance specifications.
If the team chooses to bring their own hardware, there is a $5,000 cap on hardware cost for each team's hardware. All hardware (including racks and cabling) must be under this budget cap. Hardware must be purchased new and from retailers that are accessible to everyone.
During the application process, teams will have an opportunity to submit a hardware specification. This hardware specification list may change up until two weeks before the start of the competition. At this time, a final specification list must be submitted along with a detailed budget showing receipts. All hardware purchased for the final system (including but not limited to cases and cables) must be included in this budget.
This budget must include the following information for all components used in the final competition system:
- Where purchased
- Price of equipment
Judging and Prizes
Teams will be scored on their ability to complete the workloads, as well as their HPCG score. A awards will be given for the best price/performance. Each team member will receive the prize
First Place: AMD Radeon RX 5700 XT per team member
Second Place: NVIDIA Jetson Nano per team member
During the competition, students will be running a diverse set of workloads.
Teams will be required to run the HPCG Benchmark, an industry standard benchmark designed to profile parallel systems.
In addition, teams will be required to implement a set of three workloads ahead of the competition. On the day of the competition, teams will also implement an additional mystery application.Learn More
Unfortunately, due to a low application submission rate, the HPC Boston planning team has decided to cancel this event. We hope to revisit this competition in the future when there is more interest.
Along with basic contact information, a 1 page proposal must be attached.
Each proposal must include the following information:
- Preliminary hardware information
- Basic hardware specifications, such as processor types, RAM, and number of nodes.
- NOTE: If you are planning on using the Discovery cluster, you will need to describe how your team will be taking full advantage of the hardware provided.
- Team member information
- Relevant past experience of team members (including internships, co-ops, and research experience)
- Vendor/institution support
Once accepted, there is a $20 registration fee for each student participant. This fee covers meals for the duration of the competition as operation costs for the hackaton. This fee is waived for community college students.
Directions to campus and related information will be posted closer to the start of the competition.
This year's competition will be hosted at 140 The Fenway.
Thank you to the following organizations for supporing HPC Boston 2020:
Questions? Email email@example.com
HPC Boston is being run by students from Northeastern University’s Computer Architecture Laboratory, also known as NUCAR. NUCAR conducts research in areas such as computer security, GPU computing, and embedded systems. The lab is overseen by Professor David Kaeli.