Delivering GPU-Accelerated Applications from your Private Cloud John Meza, Performance Engineering Manager, Esri Shawn Kaiser, Solution Architect, Cisco
Esri Virtualization Infrastructure • Internal Esri Virtualization - 6,500 – 8,000 VMs operational daily - Windows Server + Desktop, Linux - Supporting users throughout Esri - Marketing - Sales - Development - Education Services - Professional Services - Critical Infrastructure on-premise and off-premise replication - Additional cloud-based infrastructure
ArcGIS Pro New Product – GA Jan 2015 Combines 2D and 3D visualization and spatial analysis into 1 package State of the art 64-bit, multi-threaded application Modern Rendering Engine supporting DirectX and OpenGL that provides fast, responsive, game-like visualization to the most complex 2D and 3D data ArcGIS Pro GeoProcessing Rendering Engine Spatial Analysis Data Management OpenGL Hypervisor - VMWare Nvidia GRID
Business Drivers & Technical Challenges Existing virtualization infrastructure Did not support GPU backed virtualization. All work had to be performed on physical machines Against Corporate push to move more workloads to virtualization
Business Drivers & Technical Challenges ArcGIS User Community Internal Virtualization Needs Virtualization is heavily used Pro Engineering & Development Pro must be tested in virtualized Testing of Pro in environments Hands-on Holistic Lab Pro class development & Testing Technical Support of Pro
Solution: Cisco UCS C240 M4 • Our solution - 4 Cisco UCS C240s + Fiber Interconnect - 2x Nvidia GRID K2 in each - Administrated through existing Vcenter Each C240 server supports up to 16 VMs using K240Q profile Power users use K260Q profile -Graphics Developers -Class Developer
Cisco Data Center Portfolio and NVIDIA • Virtual workstation for high-end graphics applications • Cisco UCS C240 M4 Rack servers • Support NVIDIA GRID 1.0 and 2.0 with K1/K2 and M60 cards • Support for Magma ExpressBox for higher density • Cisco UCS B200 M4 Blade servers • Recently introduced NVIDIA M6 MXM support on Blade server • Cisco HyperFlex – 2 nd generation HyperConverged platform • Phase 1: K1/K2 support • Phase 2: M6/M60 support • Deep Learning and HPC • Cisco UCS C240 Rack servers with TESLA K80
Recommend
More recommend
