Services
Access to compute resources
We help organisations access and use high-performance compute resources for AI development.
Mimer provides guidance before, during, and after access. From applying to the right compute calls to optimising code and workflows on large-scale systems.
How we support your compute journey
Access to compute can be complex. We support you throughout the process.
What this support includes
- Guidance on selecting the right compute resources
- Support during application and onboarding
- Best practices for running AI workloads on HPC systems
- Hands-on help with optimisation and performance
When is compute access relevant?
Large-scale model training
Need large-scale compute for training or evaluating AI models
Sovereign and secure infrastructure
Benefit from sovereign AI infrastructure
Scaling beyond local limitations
Want to move beyond local or cloud-based resources
AI Factory access calls
Available compute access options
We offer different access paths depending on your compute needs
Playground Access
Fast experimentation access for SMEs and startups.
- For new industry users or those with smaller compute needs
- Access within 2 working days for a 1, 2 or 3-month period
- 5,000 GPU hours
Fast Lane Access
Quick access with medium sized compute allocation.
- For industry users familiar with AI in HPC systems
- Access within 4 working days for a 3-month access period
- 50 000 GPU hours
AI for Science
Research-driven access for scientific AI experimentation.
- For researchers from the public or private sector.
- Evaluation every two months for a 6-month access period
- Up to 360k GPU hours
Large Scale Access
Heavy compute access for industry-scale AI workloads.
- Experienced industry users with large compute needs
- 2-week cut offs for a 3 to 12-month access. Access in 10 days
- Up to 22 million GPU hours
Not sure what you need?
Describe your AI project and compute needs, and we will help you determine the right access pathway.
Services
Access to compute resources
We help organisations access and use high-performance compute resources for AI development.
Mimer provides guidance before, during, and after access. From applying to the right compute calls to optimising code and workflows on large-scale systems.
How we support your compute journey
Access to compute can be complex. We support you throughout the process.
What this support includes
- Guidance on selecting the right compute resources
- Support during application and onboarding
- Best practices for running AI workloads on HPC systems
- Hands-on help with optimisation and performance
When is compute access relevant?
Large-scale model training
Need large-scale compute for training or evaluating AI models
Sovereign and secure infrastructure
Benefit from sovereign AI infrastructure
Scaling beyond local limitations
Want to move beyond local or cloud-based resources
AI Factory access calls
Available compute access options
We offer different access paths depending on your compute needs
Playground Access
Fast experimentation access for SMEs and startups.
- For new industry users or those with smaller compute needs
- Access within 2 working days for a 1, 2 or 3-month period
- 5,000 GPU hours
Fast Lane Access
Quick access with medium sized compute allocation.
- For industry users familiar with AI in HPC systems
- Access within 4 working days for a 3-month access period
- 50 000 GPU hours
AI for Science
Research-driven access for scientific AI experimentation.
- For researchers from the public or private sector.
- Evaluation every two months for a 6-month access period
- 90 000 GPU hours
Large Scale Access
Heavy compute access for industry-scale AI workloads.
- Experienced industry users with large compute needs
- 2-week cut offs for a 3 to 12-month access. Access in 10 days
- Up to 30 million GPU hours
Not sure what you need?
Describe your AI project and compute needs, and we will help you determine the right access pathway.