Skip to content
AtheronLABS

You're visiting from the United States. Prices are shown in US dollars. Not right?

labs@atheron:~/industries/research-universities/cloud$ deploy --managed --monitored

Computing costs

Cluster and cloud spend, by group and grant, looked after for you.

Research computing is now a mix of a shared cluster and several cloud accounts, and the bills rarely say who used what. We read the accounting from each source, tag it to groups and grants, and run the dashboards for you, with backups, monitoring and updates.

// what matters here

What is different in this industry.

01

Every source in one place

Your HPC scheduler's accounting and each cloud provider's billing exports.

02

Tagged to groups and grants

Every job and resource tied to a research group and a grant.

03

Idle resources found

Alerts when cloud resources run with nothing using them, or when spend jumps.

04

Run for you

Hosted in Canada, with monitoring, backups and updates under a support plan.

// example projects

Priced examples for this service.

Each one is hypothetical, labelled as such, and priced live from our rate card. Open any of them in the estimator and make it yours.

Example projectWeb app

HPC and cloud cost dashboards

Hypothetical. Not a client, not a result.

// The problem

A research institute pays for cluster time and three cloud accounts out of one budget, and no lab or grant can see what its own work costs until the year-end bill arrives.

// What we would build

Dashboards that read your HPC scheduler's accounting and your cloud providers' billing exports, tag every job and resource to a group and a grant, and show each lab its spend against budget. We host and look after it, with alerts when a group's spend jumps.

// What is in it

  • Job accounting from your HPC scheduler, by group and grant
  • Billing exports from AWS, Azure and Google Cloud
  • Spend against budget for each lab and grant
  • Alerts when a group's spend jumps or idle resources run
  • Monthly chargeback figures for your finance office
  • Managed hosting with monitoring and backups

// Stack

  • Python
  • PostgreSQL
  • Slurm accounting
  • Cloud billing exports
  • Grafana
  • DigitalOcean Toronto

// estimate

Build
≈ US$58,800 to US$90,000, delivered within 16 weeksCAD 83,700 to 128,100
Hosting
≈ US$176 a monthCAD 250 a month
Support
≈ US$5,010 a monthCAD 7,135 a month

Prices in your currency are estimates from today's Bank of Canada rate. All invoicing is in CAD or USD.

Timeline by milestone

Discovery
2.9 to 3.4 weeks
Design approved
2.4 to 4.4 weeks
Core features
1.9 to 3.4 weeks
Full build
1.4 to 2.4 weeks
Testing and fixes
0.9 to 1.4 weeks
Launch
0.9 to 1.4 weeks

Outside our hands, and added to the calendar

Your IT department's security and sign-on review
4 to 8 weeks
Test access to your finance or research system
1 to 4 weeks

How it is paid

Deposit 20%
CAD 16,740 to 25,620
Discovery 7.7%
CAD 6,444.90 to 9,863.70
Design approved 7.7%
CAD 6,444.90 to 9,863.70
Core features 23.3%
CAD 19,502.10 to 29,847.30
Full build 15.5%
CAD 12,973.50 to 19,855.50
Testing and fixes 7.7%
CAD 6,444.90 to 9,863.70
Launch 8.1%
CAD 6,779.70 to 10,376.10
Holdback, 30 days after launch (10%)
CAD 8,370 to 12,810
Example projectAPI

Research data management hub

Hypothetical. Not a client, not a result.

// The problem

A university's research data team asks every project for a data management plan, then has no way to see where the data actually lives, who owns it, or when it may be deleted.

// What we would build

An internal hub that joins your storage, institutional repository, ORCID, grants system and HPC accounts through their published APIs, so every dataset has an owner, a project, a storage location and a retention date, and deposits to the repository start from the same record. We check each system's API before we quote.

// What is in it

  • Connectors for storage, the repository, ORCID, the grants system and HPC accounts
  • One record per dataset, with owner, project, location and retention date
  • Reminders to owners before retention dates and when people leave
  • Repository deposits started from the same record
  • A Python client so research software engineers can use it from their own scripts
  • Documented with OpenAPI, with every call logged

// Stack

  • Node.js
  • PostgreSQL
  • OpenAPI
  • ORCID API
  • Dataverse API
  • DigitalOcean Toronto

// estimate

Build
≈ US$63,700 to US$97,500, delivered within 15 weeksCAD 90,700 to 138,800
Hosting
≈ US$176 a monthCAD 250 a month
Support
≈ US$2,500 a monthCAD 3,565 a month

Prices in your currency are estimates from today's Bank of Canada rate. All invoicing is in CAD or USD.

Timeline by milestone

API contract agreed
2.9 to 3.9 weeks
Core resources built
2.9 to 4.9 weeks
Security, gateway and connections
1.9 to 3.4 weeks
Docs, SDKs and load tests
1.4 to 1.9 weeks
Launch and first callers
0.9 to 1.4 weeks

Outside our hands, and added to the calendar

Your IT department's security and sign-on review
4 to 8 weeks
Test access to your finance or research system
1 to 4 weeks

How it is paid

Deposit 20%
CAD 18,140 to 27,760
API contract agreed 8.7%
CAD 7,890.90 to 12,075.60
Core resources built 26.2%
CAD 23,763.40 to 36,365.60
Security, gateway and connections 17.5%
CAD 15,872.50 to 24,290.00
Docs, SDKs and load tests 8.7%
CAD 7,890.90 to 12,075.60
Launch and first callers 8.9%
CAD 8,072.30 to 12,353.20
Holdback, 30 days after launch (10%)
CAD 9,070 to 13,880

// questions

Questions we get about this.

Can you read accounting from our Slurm cluster?

Yes. Slurm and most schedulers keep job accounting we can read. We check your setup before we quote.

Can it cover cloud accounts that individual labs opened themselves?

Yes, once each account shares its billing export. The dashboard then shows them beside the central ones.

Who looks after the cost dashboards once they are live?

We do, under a support plan, with monitoring, backups and updates. Your team can also run it themselves.

// next

Not quite your project?

Tell us what you have in mind. We will come back to you with a range and the questions that would narrow it. Or book a call and talk it through.