Yes. Each GPULab user receives a persistent private, isolated volume in a secure, fault-tolerant distributed Ceph storage cluster. A private data volume is only accessible from a user’s personal GPULab environment.
GPULab stores personal data volumes in GPULab’s secure data centers located in California, Nevada, Colorado, Texas, or Virginia. GPULab does not store data on any public cloud or third-party provider. GPULab retains sole custody of any user data until it is destroyed 48 hours after subscription expiration.
No. GPULab does not partner with or resale any public cloud service. GPULab retains its own data center property, hardware, and network infrastructure. GPULab employs pre-screened internal staff and qualified remote hands to manage and maintain network infrastructure and server hardware. GPULab purchases network bandwidth from reputable providers and direct peering relationships through CoreSite’s Any2 Exchange.
No. GPULab environments are immutable containers and cannot be modified with system-level package managers such as apt; however, customers may install user-level packages with package managers such as conda and pip. If you require a specific system-level package, please send us a request, and we will do our best to accommodate.
No. GPULab does not permit crypto mining.
Yes. Contact us if you have more than ten users, require advanced GPUs, need custom environments, corporate SSO support, on-premise integration, white-label platform options, custom development, or partnership opportunities.
Yes! Deasil is the team behind GPULab, with over twenty years of experience in custom platform development for various industries, including manufacturing, retail, energy, logistics, and entertainment.
Deasil is a talented team of system administrators, software developers, and creative engineers, covering many industries looking to collect, analyze, move, buffer, queue, process, and present data in significant ways.
Yes! GPULab platforms the very best in open source and proprietary NVIDIA technologies. Every aspect of the GPULab environment consists of well-documented components available online, including JupyterLab, CUDA APIs, Python 3, Xeus Python, R, Julia, Octave, Bash, TensorFlow, PyTorch, Keras, Scikit-learn, fastai, and many Linux utilities useful for data science.
Cancel any subscription at any time. Once a user cancels a subscription, the GPULab service is available until the paid subscription ends, and is not automatically renewed.
GPULab offers a 99.9% service level agreement. Any outages or disruptions beyond our guarantee are eligible for pro-rated refunds. If you believe your service has been disrupted due to a GPULab outage, please contact us for a pro-rated refund. GPULab strives for high availability and customer satisfaction.
GPULab retains each user’s private data volumes on redundant, fault-tolerant distributed Ceph clusters. In the unlikely event of multiple system failures, GPULab encourages users to maintain backups. GPULab retains private data volumes up to 48 hours after subscription expiration.
Yes! GPULab’s public offering focuses on providing unlimited access to teams and individuals with tight budgets and projects suited to entry-level GPU. Contact GPULab for enterprise pricing and solutions with access to NVIDIA V100 Tensor Core GPUs.