TSL Ultra Compute

TSL Ultra Compute

Scalable, Manged HPC solution in the Cloud

Home>>

Ai

>>Ultra Compute

Massive data volumes, AI, and computer simulation in fields from life-sciences to finance, climate analysis to engineering are driving demand for parallel and specialised supercomputing at scale. Historically, these workloads have been managed on dedicated, on-premise high-performance compute clusters – but with significant capital outlay, fixed capacity and frozen architecture.

TSL helps you to process your HPC in the cloud – removing capacity constraints, giving faster results and ensuring you only pay for what you use.

Fast, Scalable, Limitless - this is the future of HPC.

Many organisations have been reluctant to adopt cloud-based parallel computing:

  • Initial setup requires scarce expertise
  • Parallel computing specialists have limited motivation to move to the cloud
  • Perceived financial uncertainty of pay-as-you-go model
  • Security is considered a risk if data is moved to an external platform

TSL offers a configurable SaaS product to rapidly setup a reliable, robust, scalable HPC within your own environment, giving you parallel computing services with high-security and data protection by design.

Finger touching a glass screen with an image of a cloud

Ultra compute environment

Deployed within hours

We deploy your ultra compute environment within hours, based on your specified configuration. We use Cloud HPC best practice to give you high-performance with security, cost, and reliability.

The service is preconfigured for security, with no data leaving your environments. All compute runs within your accounts and all user access is managed through your identity and access management solutions.

Your HPC environment can be modified, scaled, upgraded, patched and kept secure with ease, massively simplifying operational overheads. All environment changes are automated with no manual intervention, so you can focus on results, not complexity. You get evergreen access to the latest compute options, to stay ahead of your competition. With no need to focus on infrastructure, you can give all your attention to managing research. 

The Ultra Compute environment is provided on a licensed basis, allowing you to scale up and down as required with no lock-in and no long depreciation cycles.

FinOps step 1

Management console

Review and manage your HPC environment in real time through our TSL provisioned console.

Users and activity: Control and revoke project access in real time through your existing Identity and Access Management solution. 

Budgets and projects: Assign budget controls to teams and individuals, change limits and dynamically redirect funding, to ensure that projects all meet their cost profiles. Send spending reports to your finance systems for operational visibility.

Costs and financial optimisation: See spend by project, individual and workload, and gain actionable insights for reducing costs.

Security audits: Access immutable records showing which users have run which jobs with which data for IG and wider security audits.

Updates: Access new base images with security patches, new instances and new tooling for an evergreen environment with no technical intervention.

Configuration changes: Review fitness-for-purpose based on continuous analysis of your existing environment and job performance. Rapidly make changes through configuration parameter updates at any time.

FinOps step 2

Additional services

For many organisations the HPC solutions will provide all they need. But if you need additional support – we’ve got you covered. 
Choose from any or all of our five supporting services:

Analyze icon

1. Needs assessment

If you have an existing HPC, (likely on premise) and are unclear on optimal parameters for your cloud HPC design, we’ll analyse your current environment and recommend the optimal setup – including scale (min/max), CPU/GPU ratios, IOPs, networking, security, access controls and costs. We’ll review your workloads, define a range of use cases and review for benchmark speed and costs. We use the information to inform the configuration settings we apply when building your ultra compute service.

f

2. Security

We’ll advise you on industry best practice for security and implement tooling for access in a zero-trust paradigm for starters, movers and leavers. Where you have existing appropriate solutions we’ll use them, or we’ll recommend alternatives if needed, to ensure your sensitive data is protected with rigorous security frameworks and monitoring. We can extend the security posture beyond the core service to other environments that interact with your HPC service.

f

3. Data & Storage Optimisation

We can put in place solutions for high-throughput performance while managing access, retrieval speeds and cost. Our Ultra Compute environment can incorporate ultra-fast data tiers (Weka or Lustre FSx). We’ll ensure lightning-fast data transfer with high-speed networks optimised for HPC and ensure strong data gravity alignment to accelerate compute tasks.

People icon

4. Training and Enablement

If your teams need additional skills to master cloud HPC quickly and confidently, we’ll provide workshops and training for your HPC operations professionals (HPC maintenance, security, optimisation, financial management, policies, user management, and data management) and your HPC users (containerisation, coding for massively parallel computing and optimising scripts).

magnifiying glass on data

5. Managed HPC Services

Want someone else to manage your HPC environment? Leave it in our capable hands. We’ll keep your systems running smoothly with 24/7 expert support, proactive monitoring and optimisation, leaving you to focus on your research mission.

Partner with The Server Labs for fast, scalable, limitless HPC in the cloud

Email us

Download more information

You can download our HPC Solution brochure here

Download PDF