Rising Stars of HPC: Meet the SCC Connect Teams

SAPIENZA UNIVERSITY OF ROME

Team PlumJuice

Italy

PlumJuice got its start at IndySCC during SC25 under Professor Daniele De Sensi.  Although last year’s members have moved on, the team kept going. The six students from this year’s lineup come from across the computer science (CS) spectrum, including applied CS, and artificial intelligence. They represent four different countries, including Georgia, Macedonia, Pakistan, and Italy, and half the team are women. The team also placed third in the HPL benchmark at SBCC26. To get the newer members up to speed, the team paired each one with a more experienced student for regular sessions, covering everything from parallel programming to job scheduling on real systems. They also are part of the HLC research lab and run a Raspberry Pi cluster with a blog, where they document builds, competition lessons, and performance tuning, mostly for fun but partly so future SCC Connect teams do not have to start from scratch. 

Loyola Marymount University

Team LMU Lions

United States

LMU Lions are the first team ever to represent Loyola Marymount University at a Student Cluster Competition event. The team came together through research under advisor Dr. Jared Coleman, working on HPC-related topics, such as task scheduling and AI for distributed systems. To prepare, they have been running twice-weekly meetings that double as strategy sessions and member-led workshops, where pairs of students learn a topic then teach it back to the group. Loyola Marymount faculty have also put together custom HPC courses for the team. On the practical side, they set up a two-node cluster from university-provided servers to run practice workloads. As a primarily undergraduate institution, the team sees their participation as a way to show that similar schools can compete meaningfully at the national level, and they hope to open the door for future LMU teams and similar programs at other undergrad-focused institutions.

Jason Chamorro

Quinn Austin

Diego Cuadros

Simon Hernandez

Sebastian Lange

Nicholas Laus

Dr. Jared Coleman

(Advisor)

Tennessee Technological University

Team Exascale Eagles

United States

The Exascale Eagles are first-time SCC Connect competitors, but they come in with a solid foundation through Tennessee Tech’s H.P.C. Club, an on-campus organization focused on making HPC skills accessible through competitions and hands-on learning. The team members were selected by their own clubmates based on their skills and contributions. Preparation happens through semi-weekly club meetings that mix lectures with hands-on exercises plus day-long rehearsal sessions that simulate what the actual competition “feels” like. They get guidance from HPC professionals at the ASCEND Center, Tennessee Tech’s central HPC research hub, along with guest speakers from other departments and alumni. The majority of the team are seniors who have been taking classes and experiencing college together for the past four years, bringing a strong chemistry to the competition.

University of new Mexico

Team UNM Roadrunners

United States

The UNM Roadrunners are proud to represent a university with a remarkable legacy in high-performance computing, University of New Mexico researchers played foundational roles in technologies such asLAPACK and the HPL benchmark. Two of the team’s members previously competed in IndySCC, and four are new to the competition this year. The team is preparing through the summer and fall with training sessions and independent studies on the university’s Center for Advanced Research Computing (CARC) clusters, guided by an HPC curriculum developed by faculty advisor Matthew Fricke that covers cluster administration, software environments, benchmarking, and scientific applications. The UNM Roadrunners are also learning from HPC professionals at CARC, Sandia National Laboratories, and Tesla. Building on UNM’s history, the team hopes to create opportunities for new students to enter the field and contribute to the future of high-performance computing.

University of California, Santa Cruz

Team Not-So-Slow Slugs

United States

Not-So-Slow Slugs is a small but dedicated club at the University of California, Santa Cruz. This year the team built its own cluster, “Haydean”, giving them practical experience with networking, schedulers, and system administration. They meet regularly to run workloads, write scripts, and track down performance bottlenecks, working across processor architectures, instruction sets, and node communication. They also have access to NERSC for testing at production scale. Four members are returning from IndySCC at SC25, joined by two new competitors. All six earned their spots through strong showings in local student competitions. Outside of HPC, the team shares a love for rock climbing, affording a good excuse to step away from the screen for a few hours before getting back to work.

Southern University of Science and Technology

Team SUSTech Supercomputing

China

The SUSTech Supercomputing Team brings together a mix of sophomores, juniors, and seniors, some competing for the second time and others stepping into their first HPC competition. SC26 will mark the first time all six are teaming up for SCC Connect. The team benefits from its solid hardware access, including the Qiming and Taiyi clusters with more than 200 NVIDIA V100 GPUs, plus an experimental cluster with A100s and FPGAs. Preparation includes dedicated coursework in distributed and cloud computing; hands-on CUDA training; and weekly sessions tackling challenges in cluster design, CPU acceleration, and heterogeneous computing. The team’s technical backgrounds range from distributed machine learning to reinforcement learning (RL)-based EDA. They are guided by mentorship from Professor Li, Dr. Fan and experienced former competitors who help refine tactics and streamline workflows.

Australian National university

Team ANUHPC

Australia

The SUSTech Supercomputing Team brings together a mix of sophomores, juniors, and seniors, some competing for the second time and others stepping into their first HPC competition. SC26 will mark the first time all six are teaming up for SCC Connect. The team benefits from its solid hardware access, including the Qiming and Taiyi clusters with more than 200 NVIDIA V100 GPUs, plus an experimental cluster with A100s and FPGAs. Preparation includes dedicated coursework in distributed and cloud computing; hands-on CUDA training; and weekly sessions tackling challenges in cluster design, CPU acceleration, and heterogeneous computing. The team’s technical backgrounds range from distributed machine learning to reinforcement learning (RL)-based EDA. They are guided by mentorship from Professor Li, Dr. Fan and experienced former competitors who help refine tactics and streamline workflows.

Shanxi university

Team Ascent

China

Although the entire roster is new, this is Shanxi University’s second appearance at SCC Connect. The six members come from different majors and have divided tasks to play to each member’s strengths. Under the guidance of Professor Jiawei Gao, the team runs weekly technical seminars where members take turns presenting on parallel programming, performance profiling, and heterogeneous optimization backed by code samples, past competition lessons, and hands-on exercises. The lab provides 24/7 access to A800 clusters and high-speed HDR InfiniBand networks, while experienced alumni regularly share practical supercomputing tuning techniques. One thing that makes this team stand out is the liberal arts students who wandered into the team’s open-source club workshops and voluntarily took on HPC coding challenges. The team’s only female member leads compilation tuning and all-English technical communication, and her involvement has drawn more women into the lab. Most members are first-generation college students from small inland cities, and the team name draws from the university’s historic motto, symbolizing the pursuit of excellence.

EAFIT University

Team SCAR

Columbia

The SCAR team is part of the SCAR research group at EAFIT University, where all six members first crossed paths through APOLO, the university’s scientific computing center. What started as curiosity about high-performance computing, a field still largely unknown in Colombia, grew into real work through introductory courses and hands-on projects. This is the team’s first time at SC. However, its members have competed in ASC25, ASC26, and ISC25. Preparation began with benchmarking suites, including HPL, HPL-MxP, and MLPerf, with members working in pairs to improve performance. The team also attended the ninth edition of the HPC Summer School, an event that brings together universities across Colombia to learn HPC trends and fundamentals. As a pipeline into the SCAR research group, they teach introductory classes on operating systems and distributed computing fundamentals to newer students. The team spans students from their third to eighth semesters, includes both women and men, and several members are scholarship students. Because SCAR is tightly connected to APOLO, the team learns from people who work on real-world cases with researchers across Latin America every day.

Southwest Petroleum University

Team Dreambrook

China

Dreambrook is carrying forward Southwest Petroleum University’s IndySCC legacy to SCC Connect after senior students made the university’s debut at SC25 last year. This year’s team blends veterans from competitions like ASC with newcomers inspired by coursework and the university’s HPC club. All members have completed a parallel computing course centered on CUDA programming, and they hold weekly meetings where members take turns sharing technical insights. The team captain dynamically adjusts roles with dedicated leads for system administration, algorithm optimization, and data processing. The team has also gained practical experience through collaboration with the university’s Network and Information Center, giving them a bottom-up understanding of the system. What makes the team special is its mix of cultural backgrounds. Members come from different fields, bringing varied angles to problem-solving. Rooted in the multicultural city of Chengdu, the team actively recruits students from underrepresented groups through club activities and cross-year mentoring.

Fudan University

Team FDUSC

China

FDUSC is a newly established supercomputing team at Fudan University, founded and supported by the university’s Natural Language Processing Laboratory. Every member is new to SCC Connect, and the team’s interest comes from wanting to bridge cutting-edge artificial intelligence research with large-scale computing systems. As AI models continue to grow in size and complexity, the team sees understanding HPC technologies as essential for the next generation of AI engineers and researchers. They hold weekly meetings to discuss progress, share knowledge, and work through technical challenges together with frequent collaboration outside of formal sessions on topics related to AI infrastructure and HPC systems. Beyond competing, the team wants to do more. As Fudan’s first dedicated supercomputing team, they hope to spark interest in HPC across campus and create opportunities for more students to explore AI infrastructure and supercomputing through workshops, technical discussions, and community activities.

National University of Singapore

Team Kent Ridge

Singapore

Team Kent Ridge is the National University of Singapore (NUS)’s student HPC team. The team grew out of NUS’ HPC Interest Group, a community of more than 20 students and staff.This is their third time applying to SCC. Some members have competed at SCC@ISC26; APAC HPC-AI; and the Single Board Cluster Competition, where they placed first in 2026. Seniors returning from SC25 now mentor the group, while many on the team are entering their first HPC competition, bringing backgrounds in hardware building, quantum computing benchmarking, and physics. The team meets in person every week for deep-dive presentations on specific applications and benchmarks with most sessions built around live demonstrations so expertise remains cross-disciplinary and anyone can step into another teammate’s role. On the systems side, the team rehearses full deployments on staging setups, including an Orange Pi single board cluster, using Ansible, Lmod, Spack, and optimized SLURM configurations. They also attend workshops run by NSCC Singapore, NVIDIA, and AMD. Team members come from across Southeast Asia and are pursuing internships in fields ranging from AI applications to high-frequency trading.

Shanghaitech university

Team GeekPie_HPC

China

GeekPie_HPC has been part of the SC Conference since 2019, and this year’s team carries that legacy forward. The team is led by veterans who competed in IndySCC at SC25, joined by new members from diverse academic and research backgrounds who are stepping into competition for the first time. This year’s team is made up mostly of freshmen supported by a few sophomores, bringing a surge of fresh energy and new perspectives. The team has access to a dedicated EPYC cluster on campus, which they use year round to hone skills and simulate real-world scenarios. Preparation includes on-campus HPC races and short-term training sessions designed to screen and mentor new members with experienced competitors sharing knowledge from prior SCC events. The team acknowledges their relative lack of experience but believes their enthusiasm, creativity, and willingness to take on challenges will carry them throughout the competition.

National Yang Ming Chiao Tung University

Team LISΛRDGΓY

Taiwan

The name LISΛRDGΓY is “Yggdrasil” spelled backwards, reflecting the idea that in computer science, trees grow with the root on top and leaves below. All six team members came out of the first-ever HPC talent training program run by Taiwan’s National Center for High-performance Computing (NCHC). Twelve students were selected nationwide, and after competing in the ISC26 Student Cluster Competition virtual round as two separate teams, the team merged. The six who could commit to the coming season formed LISΛRDGΓY. They come from different universities across Taiwan, connected entirely through the NCHC ecosystem of training programs, internships, and domestic competitions, like HiPAC. As this is the program’s inaugural year, there are no senior competitors to pass down knowledge, so the team learns directly from the engineers who operate Taiwan’s national supercomputing infrastructure through regular online sessions. For hands-on practice, NCHC provides servers with NVIDIA V100 GPUs, and, since the completion of NCHC’s new Nano 4 system, the team can also practice on H200-based nodes. During training, they ran HPL across 32 nodes and 256 H200 GPUs, reaching 11.48 PFLOPS. Members are also active open-source contributors with work upstreamed to AMD’s rocHPL, GitHub Linguist, and vLLM-Metal. They also have earned a Best Paper Award.

Durham university

Team ClusDur

United Kingdom

ClusDur is Durham University’s 2026 SCC Connect team, bringing together students from computer science and mathematics. While this is a new team, it builds on Durham’s established SCC tradition and knowledge passed down by previous cohorts. The only returning competition member, Kiami Vemba, represented Durham at CIUK 2025, giving the team firsthand experience of optimizing workloads under real-time constraints. The rest are first-time competitors but collectively bring experience in Linux systems administration, compiler and kernel development, CUDA, and open-source contributions. Preparation runs through weekly sessions covering Slurm, MPI, benchmarking, and other essentials with every member exploring key areas before specializing. Between meetings, members independently study documentation, complete tutorials, attend webinars, and share findings through student-led presentations. The team also plans timed mock competition sessions to practice decision-making under pressure. One thing that makes ClusDur unique is the breadth of low-level experience across the team. Between them, they have built an HPC cluster using Slurm and Ansible, developed compilers, debugged the Linux kernel, contributed to GNOME Mutter, and built high-performance multi-GPU systems.

Peking University

Team Radiance of Weiming

China

The team is part of PKUSC, which has a long history of participation in cluster competitions, but the organization makes a point of giving new students the chance to take on real challenges. Every member of this year’s Peking University team is a first-time SCC Connect competitor. The members come from different academic backgrounds, including one teammate who majors in information management, bringing perspectives that connect data, systems thinking, and user needs to the group’s approach. The team has built and maintains its own cluster with Intel Xeon CPUs, A800 GPUs, and 100G InfiniBand with members rotating through software deployment, system configuration, and performance benchmarking. Training runs through hands-on workshops, covering CUDA, OpenMP, OpenACC, Linux management, and AI infrastructure, alongside regular talks from university and industry speakers on HPC system management and optimization. Each competition task is assigned a lead and co-lead, and the team’s advisor works closely with them on debugging and environment setup.

Indiana University

Team Hoosier Performance Linpack

United States

The Indiana University (IU) team came together in spring 2026. At SC25, two of the current members competed as part of a joint team with Purdue University, and they now have leadership roles as the four newer members learn the ropes. This year, the team has been competing virtually, which is a new experience for everyone. Preparation blends parallel programming education on IU HPC clusters with cluster administration skill development on the Jetstream2 cloud system also hosted at IU Bloomington. The team has been meeting virtually over the summer and will come together on campus as the fall semester begins. With a nearly uniform background in computer science, the team is focused on learning how supercomputing applies to physical simulations. Their role model is Marshal Wrubel, an IU astronomy professor and pioneering computer programmer who used early digital computers to model the physics of stellar interiors and atmospheres. Wrubel partnered with the chemistry department to secure the university’s first digital computer, an IBM 650, and was deeply committed to the idea of computers as a shared resource with democratized access for scholarly work.

Mississippi State University

Team GIGADAWGS

United States

The GIGADAWGS came together through a shared interest in HPC and the Mississippi State University (MSU) SIGHPC student chapter. Every member participated in MSU SIGHPC’s first entry to the Single Board Cluster Competition in spring 2026, and the team includes both newcomers and returning members from IndySCC. The team collaborates through weekly meetings and impromptu work sessions, both in person and online, using Muscadine, MSU SIGHPC’s five-node student cluster that mirrors aspects of the university’s production environment. They also have access to HPC staff for technical support and faculty expertise for potential competition workloads. Since 2024, the team has improved with each competition, and the broader mission goes beyond competing. Knowledge gained during preparation and competition is shared with the wider MSU SIGHPC group to build a sustainable HPC learning community at Mississippi State.

New jersey institute of technology

Team HPC Highlanders

United States

The HPC Highlanders are New Jersey’s first SCC Connect team, drawing students from several of the state’s top universities: New Jersey Institute of Technology (NJIT), Princeton, and Rutgers. After their debut at IndySCC in 2025, this marks the team’s second year competing. The team was started by a single student passionate about HPC who found like-minded people across campuses. Members now bring backgrounds in neuroscience, programming languages, computer engineering, and mechanical engineering. Four members are returning this year, joined by two who are brand new to HPC. The team is advised by Dr. David Bader, one of the inventors of the first Linux supercomputer, and supported by the staff behind NJIT’s Wulver cluster. The organization meets twice weekly and has a dedicated room reserved as a workshop for assembling, testing, and tuning full-size clusters, which doubles as an HPC classroom for introducing new students to servers. They also maintain decommissioned machines from their campuses to build Beowulf clusters as experimental platforms for testing different software stacks. The students have worked so closely with faculty and administration that NJIT created what appears to be the only undergraduate course in the country focused entirely on CUDA-aware MPI. The HPC student organization now has more than 100 members, and the university is developing two more HPC courses.

Zhejiang University

Team ZJUSCT

China

ZJUSCT is the HPC interest group at Zhejiang University, and, this year, all six members are first-time SCC Connect participants. The team brings together students from physics, electronic information, computer science, and artificial intelligence. While new to SCC, all members have attended ASC, giving them useful exposure to international HPC events. Zhejiang University provides a robust academic foundation through courses in computer organization, machine learning, numerical analysis, scientific computing, and AI4Science. In addition, the team runs its own two-week HPC-101 summer course featuring professors and industry experts. To prepare for competition, the team studies tasks from previous supercomputing contests and practices with classic HPC problems. Senior members share past competition experience and technical guidance through weekly meetings, where the team tracks progress, discusses challenges, and shares findings. Each member also takes on practical responsibilities, such as cluster maintenance, tool usage, and application testing. The team describes itself as friendly and supportive with senior members who are always willing to share and newer members actively bringing in fresh ideas.

University of birmingham

Team Beariables

United Kingdom

The Beariables qualified for the Student Cluster Challenge at Computing Insight UK (CIUK) 2025 and won the in-person finals as HPC newcomers, marking the first time the University of Birmingham won a national HPC competition. The same team then went on to compete in the ISC26 Virtual Challenge. It started when Harry Vicaradge and Adam Wilson competed in the university’s internal BEAR challenge, discovered CIUK, and pulled together a team. For SCC Connect, four original members return alongside two new additions, Sriya Musunuru and Efe Halilov. The Beariables work best in small groups with each member focused on different areas based on specific interests and strengths. This approach proved its worth at CIUK 2025, where Arthur Baverstock and Ewan Hallatt focused exclusively on benchmarking and took first place in that portion of the competition. The team is largely self taught. Most of their experience comes from competing in cluster challenges and personal projects, and all members are now entering their third year as undergraduates.

Portland State university

Team Oregon

United States

Team Oregon is a cross-institutional effort returning for its second year, bringing together students from five campuses across the state: the University of Oregon, Oregon State University, Willamette University, Portland State University, and Portland Community College. The team is united by a practical interest in high-performance computing and a focus on statewide cyberinfrastructure collaboration, education, and workforce development. HPC professionals and scientists from major institutions, including Oregon Health & Science University, serve as mentors and lead specialized workshops. The team is training directly on ORCA, the Oregon Regional Computing Accelerator, and, this year, the students will also work on deploying their own physical high-performance cluster alongside peers from the Oregon Supercomputing club. The broader goal extends past the competition itself. By connecting students, HPC staff, and researchers from multiple campuses, the initiative is building a durable network that enhances collaborative research capacity across Oregon for future cohorts.

Hangzhou City university

Team HZCU-HPCT

China

This is Hangzhou City University’s first time participating in SCC Connect. Most of the team are active members of the university’s HPC Lab and Supercomputing Team, where they regularly compete in events, such as ASC and IndySCC. The team came together through late-night debugging sessions, hardware tuning experiments, and post-competition reflections that turned classmates into close collaborators. To prepare, the team holds weekly meetings where each member reports progress in their assigned focus area, whether benchmarking, system tuning, application optimization, or HPC theory. Outside of formal meetings, they study past competition problems; work through open courses, like CMU 15-418; and simulate real competition scenarios. The team has also run internal workshops on MPI; OpenMP; schedulers; and common applications, such as HPL, HPCG, and GROMACS. The university’s Supercomputing Center provides access to local HPC infrastructure for testing software stacks, practicing deployment pipelines, and early performance tuning. The team members come from computer science, software engineering, and mathematics. Outside of competition work, some of them run a student laptop repair team on campus.

Texas state university

Team Bobcats

United States

The Bobcats came together through shared work at Texas State University under the mentorship of Dr. Damien Valles, where they built and benchmarked LEAP2, the university’s flagship HPC cluster and developed a web portal for job scheduling with Open OnDemand. This is the first time Texas State has fielded its own Student Cluster Competition team. Only two members have attended SC before: Kaylee Scanlin, who participated in SC24 and returned as a student volunteer for SC25, and Dr. Valles. The team meets weekly for technical workshops, lectures, and collaborative sessions, following a structured training timeline that covers Linux, cloud computing, networking, distributed systems, SLURM, MPI, benchmarking, and performance optimization. Every member contributes to a shared Git repository by building cloud HPC projects, documenting experiments, and benchmarking different system configurations. Beyond competition preparation, the team is assisting with a new NVIDIA GPU-based cluster for the College of Engineering and continuing development on the Open OnDemand portal. Six of the seven members are international students, and two began their academic journeys at community colleges before transferring to Texas State. Several also are interning at HPC centers and national labs over the summer, gaining valuable production experience.

University of warwick

Team Warwick SegFaults

United Kingdom

Warwick SegFaults is the first team ever to represent the University of Warwick in this competition. The members have a shared passion for systems programming and high-performance computing, recognizing an opportunity to introduce more undergraduates to HPC at a university where exposure to the field has traditionally been limited to a single optional master’s-level module. The team draws from computer science, computer systems engineering, and mathematics. Rather than looking for existing HPC experience, they prioritized finding people with real curiosity and the self-taught drive to learn. Preparation runs through weekly meetings with structured tasks, starting with individual work to build fundamentals before shifting to paired tasks as training progresses. Ahead of competition applications being released, the team trains on benchmarks and applications from previous years to get hands-on practice. The team is guided by advisor Joseph Glynn, an SCC 2025 alumnus, who sees this competition as the starting point of a lasting HPC community at Warwick that they hope will grow and draw in more students in the years to come.

Mackenzie Presbyterian University

Team HPC Brazil Force

Brazil

HPC Brazil Force brings together a unique collection of team members as each of the six study at different universities across Brazil: UNICAMP, Federal Rural University of Rio de Janeiro, Federal University of Rio Grande do Sul, UFSM, Federal University of Rio de Janeiro, and Federal University of Pampa. They each came to HPC through their own interest and met each other virtually during HPC competitions. After participating together in IndySCC last year, they are back with knowledge and notes from that experience. Each member studies individually about their core interests, and the team meets virtually to discuss updates and SCC-specific topics. They also participate in similar events when possible and stay in touch with the friends they made at last year’s competition. The diversity of universities gives the team wide visibility and reach across Brazil, and it also creates an opportunity for cultural exchange among the members themselves. The team is advised by Dr. Calebe Bianchini from Mackenzie Presbyterian University.

Sun Yat-sen University

Team Make SYSU-SCC Great Again

China

The Make SYSU-SCC Great Again team consists of six HPC enthusiasts from Sun Yat-sen University, united by a shared passion for tackling complex problems and exploring the frontiers of supercomputing. Most of the team competed together at ISC’26, where they built a solid framework for collaboration and learned what it takes to perform under pressure. That experience motivated them to take on SCC Connect. The team holds weekly meetings for strategy planning and workshops to cover cluster configuration, benchmarking, application profiling, and performance tuning. Alumni and senior club members play a key role in mentoring, sharing insights drawn from their own competition experiences. This year, the team is participating online, and they look forward to connecting with fellow competitors from around the world, exchanging ideas with experts, and bringing fresh perspectives back to strengthen the growing HPC community at Sun Yat-sen University.