Fluid Numerics' Galapagos Allocation (Single User - Monthly)
Our heterogeneous “Galapagos” cluster provides you with access to a variety of GPU platforms not typically found in public cloud options. Fluid Numerics’ Galapagos gives you the ability to easily test your software’s performance across hardware and the opportunity to improve application portability. In addition, you get access to expert support from Fluid Numerics to help you with porting and optimization activities. Before bringing a new platform into Galapagos, our team spends time learning how to operate bleeding edge hardware and gains experience porting our own research applications so that we can help you avoid pitfalls and quickly get up and running on brand new hardware
The Galapagos cluster currently hosts AMD MI210, AMD MI300A, & Nvidia V100 GPUs alongside a range of software packages that are pre-installed to promote the development of portable HPC/AI/ML applications.
Each Galapagos Allocation provides you with
A fluidnumerics.com cloud identity premium account
100 GB /home directory space
Pre-installed compilers (GNU, NVHPC, Intel, LLVM, AMDClang/Clang++,Flang), MPI (OpenMPI), and common software dependencies (HDF5, NetCDF, etc.) . Additional software can be installed upon request.
A maximum of 1 simultaneously running job per user (up to 10 jobs in the queue)
A maximum of 4 GPUs per job
Wall-clock limit of 8 hours per job
Limit of 8 hours of wall time per day (Up to 976 GPU-hours per month)
Longer running jobs can be requested and allocated in cooperation with Fluid Numerics
Email, Chat, and scheduled video support calls with our HPC System Administration and Research Software Engineer support team.
Once you sign up, someone from our team will provision your account and schedule an onboarding call to ensure you can access and make use of the cluster.
Learn more
Our heterogeneous “Galapagos” cluster provides you with access to a variety of GPU platforms not typically found in public cloud options. Fluid Numerics’ Galapagos gives you the ability to easily test your software’s performance across hardware and the opportunity to improve application portability. In addition, you get access to expert support from Fluid Numerics to help you with porting and optimization activities. Before bringing a new platform into Galapagos, our team spends time learning how to operate bleeding edge hardware and gains experience porting our own research applications so that we can help you avoid pitfalls and quickly get up and running on brand new hardware
The Galapagos cluster currently hosts AMD MI210, AMD MI300A, & Nvidia V100 GPUs alongside a range of software packages that are pre-installed to promote the development of portable HPC/AI/ML applications.
Each Galapagos Allocation provides you with
A fluidnumerics.com cloud identity premium account
100 GB /home directory space
Pre-installed compilers (GNU, NVHPC, Intel, LLVM, AMDClang/Clang++,Flang), MPI (OpenMPI), and common software dependencies (HDF5, NetCDF, etc.) . Additional software can be installed upon request.
A maximum of 1 simultaneously running job per user (up to 10 jobs in the queue)
A maximum of 4 GPUs per job
Wall-clock limit of 8 hours per job
Limit of 8 hours of wall time per day (Up to 976 GPU-hours per month)
Longer running jobs can be requested and allocated in cooperation with Fluid Numerics
Email, Chat, and scheduled video support calls with our HPC System Administration and Research Software Engineer support team.
Once you sign up, someone from our team will provision your account and schedule an onboarding call to ensure you can access and make use of the cluster.
Learn more
Our heterogeneous “Galapagos” cluster provides you with access to a variety of GPU platforms not typically found in public cloud options. Fluid Numerics’ Galapagos gives you the ability to easily test your software’s performance across hardware and the opportunity to improve application portability. In addition, you get access to expert support from Fluid Numerics to help you with porting and optimization activities. Before bringing a new platform into Galapagos, our team spends time learning how to operate bleeding edge hardware and gains experience porting our own research applications so that we can help you avoid pitfalls and quickly get up and running on brand new hardware
The Galapagos cluster currently hosts AMD MI210, AMD MI300A, & Nvidia V100 GPUs alongside a range of software packages that are pre-installed to promote the development of portable HPC/AI/ML applications.
Each Galapagos Allocation provides you with
A fluidnumerics.com cloud identity premium account
100 GB /home directory space
Pre-installed compilers (GNU, NVHPC, Intel, LLVM, AMDClang/Clang++,Flang), MPI (OpenMPI), and common software dependencies (HDF5, NetCDF, etc.) . Additional software can be installed upon request.
A maximum of 1 simultaneously running job per user (up to 10 jobs in the queue)
A maximum of 4 GPUs per job
Wall-clock limit of 8 hours per job
Limit of 8 hours of wall time per day (Up to 976 GPU-hours per month)
Longer running jobs can be requested and allocated in cooperation with Fluid Numerics
Email, Chat, and scheduled video support calls with our HPC System Administration and Research Software Engineer support team.
Once you sign up, someone from our team will provision your account and schedule an onboarding call to ensure you can access and make use of the cluster.
Learn more
What’s included
100 GB /home directory space
Pre-installed software available via environment modules
6 hour wall-clock limit per job
Maximum of 2 jobs running simultaneously
System Administration Support
HPC/AI/ML Applications User Support
Monthly webinars and office hours