I will set up and manage your ai hpc GPU cluster with slurm
Infrastructure Specialist
About this Gig
Are you scaling AI/ML research workloads and need a rock-solid High-Performance Computing (HPC) setup?
With over 15 years of infrastructure leadership, I provide end-to-end configuration, optimization, and management for enterprise GPU clusters.
What I bring to the table:
- GPU & CUDA Integration: Installation and configuration of NVIDIA CUDA drivers and libraries tailored for AI/ML performance.
- SLURM Workload Manager: Expert deployment of SLURM, job queues, partition strategies, and resource allocation.
- Enterprise Storage Sync: Integration with high-throughput storage (VAST Data, NetApp, Dell EMC Isilon).
- Observability: Setting up Zabbix or Nagios monitoring for GPU health, temperature, and cluster utilization.
Whether you need a quick diagnostic check on a single node or a multi-node enterprise cluster build, I will ensure maximum uptime and compute efficiency.
Contact me before placing an order to discuss your hardware specifications and goals!
Server:
Virtual private server
•
File server
•
Windows server
•
DNS
Operating system:
Windows
•
Linux
•
Unix
•
Vmware
Other Support & IT Services I Offer
FAQ
What information do you need to get started on my GPU cluster / SLURM setup?
I will need temporary administrative SSH access to your head node and compute nodes, details about your server hardware (CPU, NVIDIA GPU models), OS version (e.g., Ubuntu/RHEL), and your storage architecture preferences.
Do you support specialized high-performance storage like VAST Data or NetApp?
Absolutely. I specialize in enterprise NAS and iSCSI storage configuration, including NetApp, VAST Data, and Dell EMC Isilon, ensuring optimal throughput for high-performance workloads.
Can you automate server deployment and configuration for my environment?
Yes, I can create custom Ansible playbooks, Puppet modules, or Bash scripts to automate server provisioning, diskless Linux setups, and application deployments across single or multi-server environments.
Is my infrastructure data and credentials secure during our engagement?
Security is top priority. All credentials shared during our contract are kept strictly confidential, handled via secure channels (such as encrypted password vaults), and should be revoked or rotated immediately upon project completion.
Do you offer post-setup monitoring and ongoing maintenance?
Yes. I can set up full observability for your infrastructure using Zabbix or Nagios to monitor uptime, GPU utilization, and hardware health. Ongoing maintenance can also be provided via custom milestone offers.
What kind of cluster can you help with?
GPU clusters running SLURM for AI/ML training and research, from a few nodes to larger setups. I cover scheduling, GPU drivers and CUDA, and shared storage.
Can you review a cluster I already have?
Yes. A review is a good starting point: I'll look at scheduler configuration, GPU utilization, storage performance and queue behavior, then send written recommendations.
Do you work with cloud GPU instances?
Yes, including AWS. I can set up a SLURM cluster on cloud instances or advise on whether cloud or on-prem makes more sense for your workload.
