What hpc park does

HPC Park is a Russian high-performance computing (HPC) provider founded in 2019 and headquartered in Moscow (Varshavskoe Shosse 118/1). The company operates HPC Park Cloud Service, a self-developed cloud platform that delivers GPU-accelerated virtual Docker containers built on HPE Apollo infrastructure and a multi-generation NVIDIA accelerator fleet (H100, A100, A30, T4, P40, V100). Core platform capabilities include NVIDIA Multi-Instance GPU (MIG) partitioning (1/7 to 7/7 of a physical card), pre-configured PyTorch/TensorFlow/CUDA/JupyterLab environments, S3-compatible object storage, NVIDIA Triton Inference Server integration, and API-based container management. The cluster has been ranked among Russia's Top 50 most powerful supercomputers (214.9 TFlop/s Rmax, 405.47 TFlop/s Rpeak, PCT 2022). All infrastructure resides in Russia, providing 152-FZ compliance as a hard differentiator against foreign hyperscalers.

The company monetizes across five streams: usage-based and monthly-subscription GPU container rentals, dedicated bare-metal GPU server rentals, full-cycle HPC cluster integration (design, procurement, installation, commissioning, post-warranty service), IT equipment resale from HPE, Allied Telesis, Lenovo, Arista, Huawei, and Hitachi, and 24/7 consulting and technical support. Public pricing benchmarks include 290,304 RUB/month for a full NVIDIA H100 and 89,500 RUB/month (promotional) for an NVIDIA A100, with a 30,000 RUB free trial grant available to qualifying legal entities. GTM is hybrid: direct B2B field sales (phone, email, website forms; Sales Director Andrey Selikhov) for cluster integration and large deals, plus a self-serve cloud platform with Personal Account portal for mid-market consumption, supplemented by a reseller channel via RCloud by 3data.

HPC Park exclusively serves Russian legal entities (no B2C) across AI/ML/DL companies, enterprise HPC infrastructure buyers, healthcare/medical AI developers, sports technology firms, and research/academic institutions. Named enterprise customers include JuniStat (sports analytics), Neiroseti (video conference AI), Care Mentor AI (medical imaging), ironov.ai / Art. Lebedev studio (generative branding AI), and partner RCloud by 3data. The company is founder-led (Maxim Dudkin), privately held as a Russian joint-stock company (АО «ЭЙЧ-ПИ-СИ ПАРК»), with no disclosed external venture or growth equity investment, indicating bootstrapped operations.

Name

hpc park

Legal name

АО «ЭЙЧ-ПИ-СИ ПАРК»

Website

https://hpc-park.ru

Company type

Private

Founded year

2019

Operating status

Operating

Short description

HPC Park is a Russian high-performance computing provider, founded in 2019 in Moscow, offering GPU-accelerated cloud containers, dedicated servers, and HPC cluster integration. It serves Russian legal entities in AI/ML, healthcare, sports analytics, and research using NVIDIA H100/A100 hardware with 152-FZ compliance.

Ownership category

akta.pro rank

Product category

GPU Cloud Infrastructure

NAICS

Computer Systems Design and Related Services (5415), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (51821)

SIC

Computer Storage Devices (3572)

akta.pro primary industry

AI/HPC Converged Systems (Hybrid Training/Simulation) (HDACAOAD)

akta.pro secondary industries

GPU-Accelerated HPC Systems (HDACAOAC), HPC-as-a-Product / On-Prem Managed HPC (Integrated Hardware + Ops) (HDACAOAK), HPC System Software & Resource Management (Schedulers, MPI, OS Images) (HDACAOAH)

Keywords

• GPU cloud services
• High-performance computing
• AI infrastructure
• GPU rental
• Cluster integration

Headquarters

HQ city

Moscow

HQ country

Russia

HQ region

Europe

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Infrastructure, Personnel, Operations, Marketing or Sales

Revenue model

1. GPU Container Rental: Rental of virtual containers with GPU accelerators (NVIDIA H100, A100, A30, T4) on a usage-based or subscription model. Containers are billed per minute with minute-by-minute rate or via fixed monthly subscription plans. Storage is charged separately.
2. Dedicated GPU Server Rental: Rental of physical bare-metal servers with GPU accelerators (NVIDIA Tesla A100). Offers customizable configurations with scalable resources. Pricing based on selected hardware configuration.
3. HPC Cluster Integration: Full-cycle system integration services for building high-performance clusters and supercomputers, including design, procurement, installation, commissioning, and post-guarantee service.
4. IT Equipment Sales: Sale of server and network equipment from HPE, Allied Telesis, Lenovo, Arista, Huawei, Hitachi with installation and commissioning services.
5. Consulting Services: Technical consulting, sizing of solutions, architecture design, and 24/7 technical support services for HPC infrastructure.

Pricing tiers

Go-to-market motion1 record

Distribution channels4 records

Marketing channels8 records

Core offering

HPC Park provides GPU-accelerated cloud computing services (HPC Park Cloud Service) through virtual Docker containers with NVIDIA H100 and A100 accelerators, available on per-minute or monthly subscription billing. The company also rents dedicated bare-metal GPU servers and delivers full-cycle HPC cluster design, hardware procurement, installation, and integration services for enterprise clients in Russia.

Product overview

HPC Park is a Russian high-performance computing provider offering a cloud-based GPU rental platform (HPC Park Cloud Service) alongside dedicated bare-metal server rental and full-cycle cluster integration services. The core offering is the cloud platform providing virtual Docker containers with NVIDIA H100 and A100 GPUs, with optional fractional allocation via MIG technology, plus S3 object storage and Triton Inference Server as add-on modules. The portfolio also includes dedicated GPU servers for customized physical infrastructure, and professional services for HPC cluster design, integration, and 24/7 support.

Differentiator

Problem solved

Functional benefit

Brands

• HPC Park Cloud Service: Cloud platform for ML/DL, Data Science tasks with GPU rental services including NVIDIA H100 and A100 accelerators.

Products and services

• HPC Park Cloud Service High-performance cloud platform providing GPU-accelerated virtual containers for machine learning, deep learning, and AI workloads, offering NVIDIA H100 and A100 accelerators with flexible per-minute or monthly billing for legal entities.
• Virtual Containers with GPU Docker-based virtual containers with dedicated GPU resources (full or fractional via MIG), pre-configured with PyTorch, TensorFlow, CUDA, and JupyterLab, accessible via web interface or SSH for AI/ML workloads.
• Dedicated GPU Servers Bare-metal dedicated servers with NVIDIA Tesla A100 accelerators for stable, uninterrupted GPU access with customizable configurations for scientific computing, rendering, and industrial applications.
• HPC Cluster Design and Integration Full-cycle system integration of high-performance clusters including design, proof of concept, equipment procurement (HPE, Arista, Allied Telesis), installation, commissioning, and 36-month warranty with post-guarantee service.
• IT Equipment and Integration Services Consulting, sizing, hardware procurement, installation, and commissioning of servers, networking equipment, UPS, storage systems, workstations, and thin clients from HPE, Lenovo, Huawei, Hitachi, and Arista.
• AI/ML Consulting and 24/7 Technical Support 24/7 technical support with dedicated specialists assisting with model integration, architecture design, code analysis, and optimal GPU utilization for HPC infrastructure deployments.

Quantifiable outcome

• Nearly 2x acceleration in model optimization for Deep Learning tasks (specific customer case)
• +2 more outcomes

Named customers5 records

Segments5 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration3 records

AI capability11 records

Feature4 records

Partnerships

Ten partnerships are on record, tiered minor and core.

• Neiroseti (AO «Нейросети»)minorStrategic or Co-development Partner · 16 December 2024Neiroseti's Index 5 video conference AI product uses HPC Park GPU infrastructure. Product features automatic report generation from meeting recordings. Tested MIG technology (1/7 and 3/7 A100) for text vectorization and voice models.
• Care Mentor AIminorStrategic or Co-development Partner · 1 December 2023Care Mentor AI developed medical image analysis platform trained on HPC Park Cloud Service. Achieved ROC AUC improvement and expanded pathology detection from 8 to 13 types. Case study valued at 0 rubles as client achieved results during free testing.
• ironov.ai (Art. Lebedev studio)minorStrategic or Co-development Partner · 10 October 2023Joint presentation at TAdviser AI Day 2023 conference. ironov.ai AI designer neural network is trained on HPC Park Cloud Service platform. Collaboration enables AI-powered brand identity creation for business clients.
• National Base of Medical Knowledge (NBMZ)minorStrategic or Co-development Partner · 8 September 2023HPC Park joined the NBMZ Association of AI developers and users in medicine. Partnership aims to support AI development in healthcare by providing HPC resources and grants for testing. NBMZ has 20+ registered medical AI products.
• RCloud by 3datacoreChannel Partner/ Reseller/ Distributor · 5 July 2023Partnership to integrate HPC Park GPU-accelerated services into the RCloud cloud platform. Enables RCloud customers to access NVIDIA Tesla A100 computing resources through a unified management interface, scaling GPU, storage, and network resources on demand.
• Hewlett Packard Enterprise (HPE)coreTechnology or Integration · 6 October 2021HPC Park is a Silver Business Partner of Hewlett Packard Enterprise. The partnership provides access to HPE Apollo 6500 Gen10 platform for building HPC clusters. HPE has recognized HPC Park's projects multiple times, awarding them for best HPC/AI implementations in 2020 and 2021.
• JuniStatminorStrategic or Co-development Partner · 21 February 2020JuniStat developed mobile applications (JuniStat & JuniCoach) for football talent assessment using AI. The apps use HPC Park computing resources for video analysis and player skill evaluation. Joint success story in sports analytics.
• Allied TelesiscoreTechnology or Integration · 10 February 2020HPC Park became Allied Telesis Silver Partner in 2020. Allied Telesis is a world leader in Ethernet/IP network equipment with operations in 23 countries. The partnership enables creation of efficient and economical network connections for corporate and private networks.
• NVIDIAcoreTechnology or IntegrationHPC Park is an authorized NVIDIA partner, providing access to Tesla H100, A100, A30, T4, P40, and V100 accelerators. Platform supports CUDA, TensorRT, and other NVIDIA technologies. MIG technology enables efficient GPU resource utilization.
• Digital Ocean (journal)minorGTM or Marketing PartnerDigital Ocean technology journal published an interview with HPC Park founder about AI training and the HPC Park Cloud Service platform, reaching technology-focused audience.

Scale indicators4 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• CoreWeave: US-based GPU cloud specialist offering NVIDIA H100/A100 rentals on per-hour and committed-use terms with strong focus on AI training and inference workloads — directly comparable business model, customer profile (AI startups and enterprises), and silicon choice to HPC Park's cloud service.
• Lambda Labs: US-based GPU cloud provider renting H100/A100 instances to AI/ML customers and also selling Lambda Hyperplane hardware — same core proposition as HPC Park: deep NVIDIA partnership, GPU-first cloud, and dedicated-server add-ons for AI workloads.
• RunPod: GPU cloud platform offering per-second container rentals with H100/A100 access and pre-configured PyTorch/TensorFlow templates — nearly identical product surface to HPC Park Cloud Service including Docker containers, Jupyter access, and fractional GPU options.
• Paperspace (DigitalOcean): GPU cloud service acquired by DigitalOcean, offering on-demand NVIDIA GPU instances and pre-built ML containers — comparable to HPC Park's hybrid cloud + integration approach with a similar mid-market AI developer target customer.

Regional players

• Yandex Cloud: Russia's largest public cloud offering GPU instances and AI/ML platform services alongside broader IaaS/PaaS portfolio — competes with HPC Park in the same Russian market but at broader portfolio scale rather than HPC specialization.
• Selectel: Major Russian cloud and dedicated server provider with GPU instances and bare-metal rental options — directly competes with HPC Park's dedicated GPU server and cloud container offerings within the same Russian legal-entity customer base.
• SberCloud: State-backed Russian cloud computing arm of Sberbank offering GPU compute, AI Platform, and data residency-compliant cloud services — competes with HPC Park for regulated Russian enterprise AI workloads, with materially deeper capital backing.

Broad incumbents

• OVHcloud: European hyperscaler with dedicated HPC cloud offerings (including GPU-accelerated bare-metal clusters) — comparable in HPC cluster hosting capability and broader portfolio scale rather than Russian geographic focus.

Emerging players

• Vast.ai: GPU rental marketplace enabling per-hour access to consumer and prosumer NVIDIA GPUs globally — comparable in GPU rental economics and developer-focused self-serve model, but operates as a marketplace rather than an integrated HPC platform.
• Genesis Cloud: European GPU cloud provider offering bare-metal NVIDIA GPU instances and HPC-oriented pricing — comparable as a smaller regional specialist in the GPU/HPC cloud niche with similar scale of operation to HPC Park.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• LinkedInlinkedin.com/company/hpc-park

Compliance1 record

Revenue estimate

Valuation estimate

Number of profiles

Profiles2 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 hpc park