Services · Managed HPC and AI
Let our experts run your cluster.
Proactive management and monitoring by experienced HPC/AI engineers, for HPC and AI clusters we built, another vendor built, or you built in-house.
Is this for you?
Managed services are a fit if…
You’re the only admin
and the cluster is one of many systems competing for your time.
Your grant bought the cluster
but didn’t fund anyone to run it day to day.
The person who ran it moved on
and took the know-how with them.
Your IT team knows servers
but not schedulers, GPU drivers, parallel file systems and performance tuning. We can work alongside them.
Your team wants to train models
not run infrastructure. We keep the GPUs, drivers and software stack running.
You’re adding GPUs
to a cluster built for CPUs, and need them working with your scheduler and storage.
Why organizations choose managed HPC and AI services
01
Staffing challenges
Expert HPC and AI administration without hiring a full team.
02
Reduced operational risk
Fewer surprises, and fewer single points of failure.
03
Longer life from your investment
Well-maintained systems stay productive longer.
What’s included
Everything it takes to keep an HPC or AI cluster healthy, from CPUs and GPUs to storage and the scheduler, handled by engineers who do this every day.
Support for end users
Our support reaches the people who use the system every day.
OS updates and security patches
Your cluster stays secure and reliable, without the admin burden.
Fault monitoring
We find and fix problems before they become outages.
Dedicated support
Engineers who know your cluster, for continuity and expertise.
Application installation
Scientific applications and AI frameworks, installed correctly, configured properly and optimized.
Cluster maintenance
Regular maintenance that keeps the cluster running efficiently.
Performance focus
Running at peak performance, for maximum value.
Troubleshooting and tuning
Fixing root causes, so issues don’t come back.
Monitoring and alerts
Alerting and notifications
Problems surface to us first, so we can act before your users notice.
Proactive monitoring
Continuous, automated checks on system health, so issues are caught early.
Capacity planning and lifecycle
Technology needs change. We help you plan for what’s next.
Future growth
Plan capacity ahead of new users, projects and workloads.
Hardware refresh cycles
Know when to upgrade, and what to replace first.
Budget forecasting
Predictable plans for hardware and support spending.
Storage expansion
Grow storage before it becomes a bottleneck.
Managed services in action
What we do while you work.
Fixed before it becomes your problem
ClusterVisor flags failing hardware, and your engineer drains the node and arranges the fix while the rest of the cluster keeps running.
Patched without the downtime
Updates go out a few nodes at a time, so your users keep working while the cluster stays current.
How it works
Two ways to start.
New
A system built by ACT
Your cluster arrives preconfigured with ClusterVisor and ready to go. Managed services can begin the day it’s installed, with no separate onboarding.
Built, tested and configured before it ships
ClusterVisor installed and monitoring from day one
Supported by our team for the life of the system
Existing
A cluster you already have
Built by another vendor or in-house? We bring it under management in three steps.
1
Assessment. We review your hardware, software stack, scheduler and pain points, and agree on what we’ll manage.
2
ClusterVisor deployment. We install ClusterVisor so our team can see your nodes, jobs and alerts in one place.
3
Ongoing management. Our engineers monitor, patch, tune and support your cluster, so your team can focus on research.
Coverage and pricing
Built around how you use your cluster.
No two clusters are run the same way, so no two agreements are either. We set coverage and pricing with you during the assessment, and adjust as your needs change.
Coverage you choose
Coverage during business hours or extended hours
Response times matched to how critical each issue is
Escalation paths and named contacts agreed up front
Maintenance windows scheduled around your users
Pricing that fits
1
Hourly. Pay only for the engineering time you use. Ideal for occasional help, upgrades and one-off projects.
2
Fixed price. A predictable cost for ongoing management, so your budget is settled before the year starts.
Not sure which fits? We’ll recommend one after the assessment.
Get a tailored quoteWhat customers say
Support that answers within the hour.
“
I want to thank the Advanced Clustering crew for being the best support team in the world!
Frank Rash
Emory University
“
“I have nothing but great things to say about the team at Advanced Clustering and the quality-of-service you guys provide.”s
Jeff S.
Toyota Racing Development
“
Jordan and the team at Advanced Clustering are the best. That’s why I keep going back to them for all our computational needs.
Kiran B.
University of Texas at San Antonio
FAQ
Handing over your cluster? Here’s what to know.
Have a question that isn’t here? We’re happy to talk it through.
Ask us →How do you access our cluster?
However works for your security policies. We support multiple remote access options, including VPNs, jump hosts and remote access tools.
We have strict security requirements. Can you still manage our cluster?
Yes. You choose how we connect and you keep root access. Managed clusters can also run ClusterVisor’s enhanced security, with full disk encryption, FIPS mode and SELinux in enforcing mode.
About ClusterVisor security →Do we keep root access and control?
Yes. It’s your cluster. You keep full root access and control throughout.
Can you work alongside our own team?
Yes. We can fully manage your cluster or work alongside your admins, taking on as much or as little as you need.
Is there a minimum contract length?
No. We can work with whatever terms your organization requires.
Can you come on-site?
Yes. Our engineers handle day-to-day management remotely, so help is never more than a call away, and we’ll come on-site whenever you need us.
How do we get started?
With an assessment. We review your cluster, talk through your goals and show you exactly what managed services would look like for your team.
Request an assessment →Interested in managed HPC services?
Engineering the future of HPC and AI
info@advancedclustering.com(866) 802-82223148 Roanoke Road, Kansas City, MO 64111