RTU HPC Center

What RTU HPC Center does

RTU HPC Center is the Latvian national high-performance computing competence center and a structural unit of Riga Technical University (RTU), established in June 2012 as a continuation of FP6/FP7 BalticGrid initiatives (2005–2010). The center operates the 'Rudens' cluster — 54 compute nodes with 2,736 CPU cores (AMD Epyc 7713, Intel Xeon Gold 6154, Intel Xeon E5-2680 v3), 28 NVIDIA GPUs (Tesla K40, V100, A100), 25 TB RAM, and approximately 1.6 PB of storage (BeeGFS, NFS Isilon, NVMe scratch) interconnected via InfiniBand at 240 TFlops total performance. It runs Torque/Moab job scheduling on CentOS 7.9 with a pre-installed scientific software stack spanning ANSYS, COMSOL, MATLAB, OpenFOAM, GROMACS, GATK4, CUDA Toolkit, and Intel Parallel Studio.

The center serves three primary user groups: (i) RTU scientists, doctoral students, and undergraduates; (ii) external Latvian academic and government research institutes (OSI, BMC, BIOR, Silava, LU, RSU); and (iii) Latvian SMEs, startups, and industrial firms seeking CFD simulation, code optimization, and digital prototyping services. Revenue is generated through usage-based pricing on a cost-recovery basis — 0.015–0.019 EUR/core-hour for CPU, 0.18–0.22 EUR/GPU-hour for GPU, 20 EUR/hour for software development, and 20–25 EUR/TB/year for storage — supplemented by RTU's ZAF (Zinātnes atbalsta fonds) covering 80–100% of costs for qualifying academic users. The Atvērtā HPC platform (Waldur-based), launched in 2023, aggregates the center's resources with LU Skaitliskās modelēšanas institūts and NCC Latvia into a unified eID-authenticated web portal exposing Galaxy, JupyterHub, Seqr, and Moodle LTI integrations.

Distribution combines direct remote access (SSH/HTTPS), self-service web provisioning, and managed CFD/simulation services, with active outreach through EuroCC Latvia enterprise consultations, LinkedIn, training courses (CUDA, MATLAB, SolidWorks), and the EuroHPC JU network connecting users to the LUMI supercomputer in Finland. The center is a core participant in EuroCC/EuroCC2/EuroCC3, CERN Tier-2, CoE RAISE, EOSC NORDIC, and the new AIFA-LAT AI Factory Antenna project, and in July 2026 inaugurated a MikroTik/Tallay-co-funded AI data server dedicated to research and medical data processing.

Name

RTU HPC Center

Legal name

Riga Technical University High Performance Computing Center

Website

https://hpc.rtu.lv

Company type

Private

Founded year

2012

Operating status

Operating

Headcount range

1–10 employees

Short description

RTU HPC Center is a Riga Technical University structural unit operating Latvia's national high-performance computing cluster ('Rudens') and the Atvērtā HPC platform, providing CPU/GPU compute, scientific software, CFD simulations, and training to academic researchers, students, and Latvian SMEs on a cost-recovery basis.

Ownership category

akta.pro rank

Product category

High-Performance Computing Services

NAICS

Computer Facilities Management Services (541513), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518)

SIC

Services-Computer Integrated Systems Design (7373), Services-Computer Processing & Data Preparation (7374)

akta.pro primary industry

HPC Supercomputers & Turnkey Systems (HDACAOAA)

akta.pro secondary industries

AI/HPC Converged Systems (Hybrid Training/Simulation) (HDACAOAD), HPC Clusters & Scale-Out Compute (x86/ARM) (HDACAOAB), HPC-as-a-Product / On-Prem Managed HPC (Integrated Hardware + Ops) (HDACAOAK), HPC System Software & Resource Management (Schedulers, MPI, OS Images) (HDACAOAH)

Keywords

• High-performance computing
• HPC cluster services
• GPU computing
• Scientific computing
• Computational research

Headquarters

HQ city

Riga

HQ country

Latvia

HQ region

Europe

Offices2 records

Markets served

GTM type

B2B

Offering type

Services

Cost components

Infrastructure, Technology or R&D, Personnel, Operations

Revenue model

1. HPC Computing Time (CPU): CPU core-hours billed based on actual job execution time. Prices differentiated for RTU internal users (0.015 EUR/core-hour) and external users (0.019 EUR/core-hour). Charge only for execution time, not queue waiting time.
2. HPC Computing Time (GPU): GPU hours billed in addition to CPU core-hours. RTU internal: 0.18 EUR/GPU-hour. External: 0.22 EUR/GPU-hour. CPU core-hours also charged alongside GPU usage.
3. Software Development Services: Custom software development, code optimization, and HPC migration services. Billed at 20 EUR/hour for RTU users. Covers C/C++, Python, Perl, PHP, MATLAB programming; OpenMP and MPI parallelization.
4. CFD Simulation Services: Computational Fluid Dynamics (CFD) modeling and simulation services performed by RTU scientists. Priced per hour based on project scope. Includes 2D/3D model creation, parallel HPC simulation, and results interpretation.
5. Data Storage Services: Long-term data storage. RTU internal: 20 EUR/year per TB. External partners: 25 EUR/year per TB.

Pricing tiers

Go-to-market motion2 records

Distribution channels5 records

Marketing channels6 records

Core offering

RTU HPC Center operates the Latvian national high-performance computing cluster 'Rudens' (54 nodes, 2,736 CPU cores, 28 NVIDIA GPUs, 1.6 PB storage) and the Atvērtā HPC platforma (Waldur-based unified access portal). It provides usage-based CPU/GPU computing time, data storage, custom HPC software development, CFD simulation services, and training courses to scientists, universities, research institutes, students, and enterprises across Latvia.

Product overview

RTU HPC Center operates a unified HPC ecosystem centered on the HPC klasteris Rudens (high-performance computing cluster) and the Atvērtā HPC platform (Waldur-based web portal). The cluster provides 54 compute nodes with CPU and GPU (NVIDIA Tesla V100, A100, K40) resources for parallel scientific computing, accessible via SSH, web interfaces, and virtual desktops. The Atvērtā HPC platform aggregates resources from RTU HPC Center, Latvijas Universitātes Skaitliskās modelēšanas institūts, and NCC Latvia into a single access point. Key access modules include Galaxy (bioinformatics), JupyterHub (interactive notebooks), and Seqr (genomic analysis). The center also maintains a catalog of pre-installed scientific software (ANSYS, COMSOL, MATLAB, GROMACS, GATK, etc.) and offers specialized services including CFD simulations, software development, and professional training (CUDA, MATLAB, SolidWorks). A new AI data server for research and medical data processing was unveiled in July 2026.

Differentiator

Problem solved

Functional benefit

Products and services

• HPC Cluster 'Rudens' Computing Access Usage-based access to the HPC cluster 'Rudens' — 54 compute nodes with 2,736 CPU cores, 28 NVIDIA GPUs (V100, A100, K40), and 1.6 PB storage — delivered via job-queue scheduling (Torque/Moab). Available to RTU staff, students, and external academic/commercial organizations at differentiated per-core-hour and per-GPU-hour rates.
• Atvērtā HPC platforma (Waldur) Self-service unified web portal for HPC resource provisioning and access across multiple Latvian institutions. Features browser-based HPC access, virtual desktops, Galaxy bioinformatics portal, JupyterHub, Seqr genomic analysis, and Moodle LTI integration. Accessible via eID authentication, providing 24/7 access for students, researchers, educators, and enterprises.
• CFD Simulation Service Computational Fluid Dynamics modeling and simulation service delivered by RTU scientists, covering 2D/3D model creation, parallel HPC simulation execution (using OpenFOAM and COMSOL), and results interpretation. Priced per hour on a quote basis; intended for enterprises and research institutes needing expert-led simulation work.
• HPC Software Development Service Custom code porting, parallelization, and optimization services for migrating applications to the HPC cluster. Supports C/C++, Python, Perl, PHP, MATLAB; parallel interfaces OpenMP and MPI. Billed at 20 EUR/hour for RTU users. Delivered by lead researchers.
• HPC Training Courses and Seminars Organized training courses including CUDA (led by Arnis Lektauers), MATLAB (led by Māris Tērauds), and SolidWorks (led by Modris Dobelis), plus HPC consultations and seminars. Practical online training delivered in collaboration with EVITA (EuroHPC Virtual Training Academy) and POP Centre of Excellence. Targeted at students, researchers, and businesses.
• Data Storage Service Long-term data storage on the HPC infrastructure, priced at 20 EUR/year per TB for RTU internal users and 25 EUR/year per TB for external partners. Includes BeeGFS parallel filesystem, NFS Isilon, and NVMe scratch storage tiers.

Quantifiable outcome

• CFD simulation reduced building façade construction costs by enabling accurate wind load coefficients (Cpe) analysis per building region, avoiding conservative overengineering
• +2 more outcomes

Named customers9 records

Segments4 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

Integration5 records

Feature5 records

Partnerships

Twelve partnerships are on record, tiered moderate, core, major, minor and historical.

• POP Centre of ExcellencemoderateStrategic or Co-development Partner · 1 May 2026POP (Performance Optimization and Productivity) Centre of Excellence collaborates with RTU HPC Center to deliver practical online training on HPC performance analysis and optimization with European leading experts.
• EVITA (EuroHPC Virtual Training Academy)moderateStrategic or Co-development Partner · 1 May 2026EuroHPC initiative building a unified HPC competency, training, and certification ecosystem across Europe. RTU HPC Center invites Latvian HPC experts to participate in EVITA.
• EuroHPC Joint Undertaking (EuroHPC JU)coreStrategic or Co-development Partner · 1 January 2022RTU HPC Center is the Latvian national competence center within the EuroHPC JU ecosystem, participating in EuroCC, EuroCC2, EuroCC3, and OrthoHPC projects. The center coordinates national HPC adoption, provides training, and connects Latvian researchers to European pre-exascale HPC resources including LUMI in Finland.
• Latvijas Universitātes Skaitliskās modelēšanas institūtscoreTechnology or IntegrationLU Skaitliskās modelēšanas institūta contributes HPC resources to the Atvērtā HPC platforma (Waldur), creating a unified access point for HPC services across Latvia. Also shares the HPC cluster for computational research.
• NCC Latvia (Latvijas HPC Kompetences Centrs 'SuperS')coreStrategic or Co-development PartnerNational HPC Competence Center Latvia provides HPC consultation and drives enterprise adoption of HPC among Latvian SMEs and startups. Collaborates with RTU HPC Center through the Atvērtā HPC platforma. Also known as 'SuperS'. Hosts free enterprise consultations and demo-cases.
• CERN (Tier-2)majorStrategic or Co-development PartnerRTU HPC Center serves as a CERN Tier-2 computing center, providing computational resources for CERN's LHC experiments and data processing. Aleksandrs Gutcaits is the Tier-2 expert. Part of the worldwide LHC computing grid.
• CoE RAISE (Centre of Excellence in AI)majorStrategic or Co-development PartnerParticipation in the European Centre of Excellence in AI (RAISE), bringing HPC and AI expertise together for research applications in Latvia.
• EOSC NORDICmoderateStrategic or Co-development PartnerCollaboration within EOSC NORDIC project to integrate RTU HPC resources into the European Open Science Cloud ecosystem.
• InnoHPC (VPP Covid-19)minorStrategic or Co-development PartnerRTU HPC contributed computational resources to the VPP Covid-19 project, using HPC to model and combat the Covid-19 pandemic in Latvia.
• EuroCC / EuroCC2 / EuroCC3coreStrategic or Co-development PartnerNational HPC Competence Centers project funded by EuroHPC JU. Līna Marta Sarma serves as EuroCC 2 Project Leader. EuroCC coordinates national HPC adoption, training, and access across European countries. EuroCC3 started April 2026.
• BalticGrid / BalticGrid-IIhistoricalStrategic or Co-development PartnerFP6/FP7 funded projects (2005-2010) that initiated computing activities at RTU, forming the foundation for the HPC Center established in June 2012. Legacy partnership that led to the creation of the center.
• OrthoHPCmoderateStrategic or Co-development PartnerEuropean HPC project involving RTU HPC Center. Part of the EuroHPC JU portfolio of national competence center collaborations.

Scale indicators5 records

Recent moves7 records

Expansion highlights6 records

Direct peers

• IT4Innovations National Supercomputing Center: Czech national HPC centre operating Barbora and Karolina clusters, part of EuroHPC JU. Directly comparable as a Central European national HPC centre serving academic and industry users with similar training and competence-center mandates.
• Barcelona Supercomputing Center (BSC): Spanish national supercomputing centre hosting MareNostrum 4/5 and a PRACE/EuroHPC tier-0 system. Comparable on the academic-government ownership model, HPC user-support services and pan-European project participation.
• CINECA: Italian non-profit consortium supercomputing centre operating Leonardo (EuroHPC pre-exascale) for Italian universities, research institutes and industry. Closest peer on consortium/non-profit governance and EuroHPC JU user-allocation model.
• HLRS (High-Performance Computing Center Stuttgart): One of Germany's three national HPC centres, operating Hazel Hen/Hawk systems and serving university and industry users across engineering and scientific domains. Directly comparable on infrastructure model, training programmes and EuroHPC alignment.
• CSC – IT Center for Science: Finnish national HPC and computing infrastructure provider operating the Mahti, Puhti and hosting LUMI supercomputers for academia and industry. Most direct analogue as a EuroHPC JU national competence center serving university researchers, commercial users and European projects.
• Poznan Supercomputing and Networking Center (PSNC): Polish national HPC and e-infrastructure centre affiliated with the Institute of Bioorganic Chemistry, Polish Academy of Sciences. Comparable in providing HPC, storage and networking for research communities and connecting to GEANT/EOSC.

Emerging players

• Gcore: Cloud and edge infrastructure provider offering GPU-accelerated compute and referenced in RTU HPC's own case studies as a partner for genome-sequencing workloads. Comparable as an emerging commercial alternative for Latvian enterprises that might otherwise use RTU HPC.

Broad incumbents

• Forschungszentrum Jülich (JSC): German research centre operating the Jülich Supercomputing Centre (JSC) with JUWELS and JURECA systems. Comparable on national supercomputing scale, scientific software stack and participation in European HPC networks, though embedded in a broader science organisation.
• SURF: Dutch collaborative organisation for ICT in education and research operating national HPC, cloud and storage services. Comparable as a national-scale academic HPC provider but with a much broader mandate (networking, cloud, data) than RTU HPC.

Others

• EuroCC / EuroHPC Joint Undertaking: European joint undertaking coordinating EuroHPC supercomputers and the EuroCC national competence centre network of which RTU HPC is the Latvian node. Direct peer relationship as funder, coordinator and ecosystem counterpart rather than competitor.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks6 records

Key highlights7 records

Customer concentration

• GitHubgitlab.com/eurocc-latvia
• LinkedInlinkedin.com/company/rtu-hpc-center

Revenue estimate

Valuation estimate

Number of profiles

Profiles11 records

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about RTU HPC Center