HeavyHorse - High-Performance Computing Workstations

Humusoft workstation HeavyHorse with AMD processors

HeavyHorse

The HeavyHorse range of powerful computing workstations includes models based on the latest AMD EPYC, Threadripper and Ryzen processors.

The new models with up to 384 computing cores / 768 threads are a unique platform for demanding parallel computing and provide unrivaled performance in relation to energy consumption and price. All models are now equipped with NVIDIA RTX PRO Blackwell, RTX Ada or RTX Ampere (basic workstation models) series graphics / computing cards.

We would like to present to you a series of high-performance computing workstations designed for optimal performance during demanding numerical calculations.

The workstations utilize modern 64-bit AMD processor architecture and professional NVIDIA graphics accelerators and computing cards, making them a stable foundation for running demanding engineering and scientific software.

Modern software tools for numerical calculations allow users to perform calculations with large volumes of data and support parallel algorithms that can significantly reduce calculation time. To fully utilize these capabilities, a computing system with a sufficient number of computing cores and RAM capacity is needed.

Configurations of workstations

Contact sales
MATLAB R2025a Benchmark

MATLAB R2025a Benchmark

GPU vs CPU, Acoustics Module, COMSOL Multiphysics 6.4 (30 793 810 degrees of freedom)

GPU vs CPU, Acoustics Module, COMSOL Multiphysics 6.4 (30 793 810 degrees of freedom)

Impact of CPU core count and memory bus bandwidth on computing speed, COMSOL Multiphysics 6.4, Electromagnetic and Heat Transfer (3 131 979 degrees of freedom)

Impact of CPU core count and memory bus bandwidth on computing speed, COMSOL Multiphysics 6.4, Electromagnetic and Heat Transfer (3 131 979 degrees of freedom)

Configuration of workstations

The hardware configuration is fully modular and can be customized to meet the specific requirements of particular applications. We optimize HeavyHorse workstation configurations based on the expected workload and type of algorithms utilized.

HeavyHorse parallel computing stations are a high-performance and affordable parallel computing system based on 64-bit AMD EPYC, Threadripper and Ryzen processors. They are available in configurations with 6 to 384 computing cores (12 to 768 threads).

HeavyHorse workstations can be supplied as turnkey systems with the operating system and application software MATLAB, MATLAB Parallel Server, Parallel Computing Toolbox or COMSOL Multiphysic, COMSOL Server, COMSOL Compiler already installed.

Application:

  • High-Performance Computing
  • Distributed computing
  • Development of parallel algorithms
  • Finite Element Method
  • Processing of large data

Hardware:

  • CPU AMD Ryzen, Threadripper, EPYC
  • up to 384 cores, 768 threads
  • 32 ~ 768 GB RAM, expandable up to 3072 GB RAM
  • Professional graphics and computing card NVIDIA RTX Ada / RTX PRO Blackwell
  • 1000 GB ~ 8000 GB M.2 NVMe SSD hard drive
  • PC Case: Tower, 4U/5U rack

Software:

Go to the price list

Contact sales
silent workstation HeavyHorse

HeavyHorse price list

All models come standard with a keyboard and mouse. The base configuration does not include a monitor or an operating system license—these components are added based on the customer's specific selection.

The configuration of individual models can be adjusted upon the customer's request by mutual agreement, so that the workstation's performance profile exactly matches the planned deployment.

Download the price list and basic configuration in PDF

Contact sales

Technical support

Do you have a technical question? Technical support is available every working day.

* Required field.

By submitting this form, you give us your consent to process your personal data to the extent necessary to process your inquiry, registration or request.
Our Privacy Policy can be found here.

Contact us

Do you have a question, do you need a price quote, are you hesitant to choose a suitable module or do you simply want to tell us something? Don't keep it to yourself.

Write to us

close