sizing esx for rich graphics user
Transcription
sizing esx for rich graphics user
April 4-7, 2016 | Silicon Valley S6385 - WHAT IS CLOUD AND WHAT CAN IT DO FOR YOUR DESKTOP WORKLOADS Matt Coppinger Director, Technical Marketing, End User Computing, VMware Stephane Asselin Senior EUC Architect, Technical Marketing, End User Computing, VMware Jeff Weiss GRID Solutions Architect Manager, NVIDIA Why GPU in the Cloud? Key technologies for GPU in the Cloud AGENDA Performance & Sizing Best Practices Deployment Quick Start Summary Resources & Next Steps 2 Why GPU in the Cloud? Key technologies for GPU in the Cloud AGENDA Performance & Sizing Best Practices Deployment Quick Start Summary Resources & Next Steps 3 MAJOR MARKETPLACE DEMAND FOR CLOUD-BASED WORKSTATIONS More distributed than ever Large datasets are harder to share Contractors need revocable access Data leakage & corporate security 4 4 TRYGRID DEMO 5 DESKTOP, APP + GPU VIRTUALIZATION ENABLES CLOUD-BASED WORKSTATIONS IT Manages Centralized Virtual Desktops Security & Compliance + Highly Efficient Remote Protocol Anywhere Access Desktop Virtualization with Server-side GPU Business Agility Improves the Delivery of Graphics Mobile & BYOD Access OpEx Savings Users Remotely Access from Various Devices 6 6 WHO BENEFITS THE MOST FROM GPU? DESIGNER POWER USER KNOWLEDGE WORKER TASK WORKER Source: Gartner & NVIDIA 2012 Research 7 7 THE IMPORTANCE OF GPU HIGH-PERFORMANCE GRAPHICS APPLICATIONS Media PLM & Volume Design Engineering & Design 8 8 HIGH-PERFORMANCE GRAPHICS USE CASES Image Quality Interactivity Cost/Seat 2D / 3D Oil & Gas Diagnostics Lightweight CAD Architecture Design Engineering Media & Graphics 3D Viewers GIS WebGL Productivity Apps with Graphics Aero Light Users Fewer Applications Task Worker Heavy Users Many Applications Knowledge Worker Power User Designer 9 9 DELIVERING A COMPLETE PORTFOLIO, FOR EVERY Oil & Gas WORKER, EVERY USE CASE Architecture Design Diagnostics Lightweight CAD Image Quality Interactivity Cost/Seat 2D / 3D Engineering Media & Graphics vDGA 3D Viewers GIS WebGL Productivity Apps with Graphics Aero Dedicated Graphics Acceleration NVIDIA GRID VGPU vSGA Mediated Passthrough Soft 3D Shared Graphics Acceleration Software 3D Light Users Fewer Applications Task Worker Heavy Users Many Applications Knowledge Worker Power User Designer 10 1 VIRTUAL BATTLESPACE IN THE CLOUD 11 12 Why GPU in the Cloud? Key technologies for GPU in the Cloud AGENDA Performance Best Practices and Results Deployment Quick Start Review of Cloud Desktop Components Resources 13 KEY TECHNOLOGY: VMWARE HORIZON Private Cloud infrastructure for delivering graphics rich desktops in the GPU enabled datacenter 14 WHY CHOOSE VMWARE FOR CLOUD-BASED WORKSTATIONS? Market leading hypervisor Proven, industry leading platform, fully integrated endto-end solution Supports huge range of use cases Deliver a superior experience with state of the art protocols Choice of GPU technologies —vGPU, vDGA and others 15 1 VMWARE NVIDIA CLOUD ARCHITECTURE Any Device Any Application Business Mobility: Applications | Devices | Content vGPU Traditional Applications Modern, Cloud Applications Cloud Management HYBRID CLOUD One Cloud PRIVATE MANAGED PUBLIC Your Data Center vCloud Air Network vCloud Air Virtualized Compute, Network, Storage Build-Your-Own Converged Infrastructure Hyper-Converged Infrastructure 16 1 VMWARE HORIZON: CLOUD-BASED WORKSTATION SOLUTION FOR ALL USE CASES Virtual desktop and app delivery Desktop and app delivery from public cloud Horizon 7 Horizon Air APP NEW MAIN OFFICE REMOTE OFFICE POINT OF SALE Simplifying App Management NEW DEVELOPERS Horizon FLEX Horizon App Management NEW DESIGNERS/ ENGINEERS Secure containerized desktops NEW MOBILE NEW NON-EMPLOYEE 17 1 PRIVATE CLOUD VIRTUAL DESKTOP INFRASTRUCTURE • Physical server host(s) • Hypervisor Software • Virtual machine management software • Virtual desktop management system • Virtual desktop connection broker • Access and authentication systems • Display protocol • Application Delivery mechanisms • User Persona/Profile management • Clever “stuff” that simplifies it all for you…. 18 1 HORIZON 7: FIRST-TO-MARKET INNOVATIONS TO MAKE THE CLOUD A REALITY Enterprise Scale & Security Deployment Options GPU Virtualization Blast Extreme Experience App Lifecycle Management Smart Policies CONFIDENTIAL 19 19 INTRODUCING BLAST EXTREME - THE MOBILE CLOUD PROTOCOL Broad Client Support: H264 available in most devices GRID Optimized: Deliver more 3D users Network Friendly: TCP or UDP works in more places Built for the Cloud: Adapts better to lossy networks Use Less Bandwidth: Deliver better video to more users Better Battery Life: Optimized for mobile workers 20 2 VMWARE AND NVIDIA REFERENCE ARCHITECTURE iOS Android Horizon Clients HTTPS/PCoIP/Blast Extreme DMZ (HTTPS/PCoIP/Blast) PCoIP or Blast Extreme ESX, vCenter, Horizon, SQL, AD traffic Shared Storage MacOS Thin Client Kiosk Horizon Connection Servers VMware Access Point Identity Manager NVIDIA vCenter Svr Horizon Lic. 3D Virtual Desktops Horizon Virtual Desktops App Volumes File Print Server RADIUS Server SSD PC MS SQL Active Directory CAD/CAx “Server” SSD 3D DESKTOP CLUSTER DESKTOP CLUSTER View Composer vCenter Operations Manager for Horizon MANAGEMENT CLUSTER Shared Storage (Optional) 21 KEY TECHNOLOGY: GPU VIRTUALISATION How NVIDIA GRID is bringing GPU power to cloud based workstations 22 HOW DOES NVIDIA GRID WORK? Virtual Desktop Virtual Desktop Virtual Desktop Virtual Desktop Virtual Desktop Virtualization Layer Virtual Desktop Hardware Hypervisor CPUs Server 23 HOW DOES NVIDIA GRID WORK? Virtual PC Virtual PC Virtual Workstation Virtual Workstation Virtual Workstation NVIDIA Graphics Driver NVIDIA Graphics Driver NVIDIA Graphics Driver NVIDIA Quadro Driver NVIDIA Quadro Driver NVIDIA Quadro Driver vGPU vGPU vGPU vGPU vGPU vGPU Virtualization Layer Virtual PC Hardware Hypervisor CPUs Server NVIDIA GRID vGPU manager NVIDIA GPU NVIDIA GPU H.264 Encode 24 NVIDIA GRID PLATFORM Centralized virtual graphics capabilities for the enterprise Support, Updates & Maintenance NVIDIA GRID Software NVIDIA Tesla (Datacenter GPU) 25 DESKTOPS & VIRTUAL APPS Apps NVIDIA GRID Virtual PC NVIDIA GRID Virtual Workstation NVIDIA GRID Virtual Applications 26 GRID vGPU Architecture 101 A quick recap … Hypervisor GRID vGPU Manager VM Apps NVIDIA driver Tesla GPU 29 GRID vGPU Architecture 101 GPU engines Hypervisor VM GRID vGPU Manager Apps NVIDIA driver Graphics And Compute Copy Engine Video Encode Video Decode Tesla GPU Framebuffer (FB) Framebuffer Graphics Compute Video Encode Video Decode Copy Engine 30 GRID vGPU Architecture 101 GPU access Hypervisor VM GRID vGPU Manager Tesla GPU Apps NVIDIA driver GPU Base Address Register Virtual BAR A “window” into the GPU Used to access control registers and framebuffer All driver accesses to GPU resources go through a BAR Framebuffer Graphics Compute Base Address Register (BAR) Video Encode Video Decode Copy Engine We assign the VM a subset of the GPU’s BAR – a virtual BAR 31 GRID vGPU Architecture 101 Access protection Hypervisor VM Page Table GRID vGPU Manager NVIDIA driver CPU’s MMU Tesla GPU Apps GPU Base Address Register Translation Virtual BAR Access to BAR controlled by CPU’s Memory Management Unit (MMU) Hypervisor creates MMU page tables that allow VM to access its Virtual BAR Blocks access to locations outside the Virtual BAR Framebuffer Graphics Compute Video Encode Video Decode Copy Engine 32 GRID vGPU Architecture 101 Work posting Hypervisor VM GRID vGPU Manager Tesla GPU Apps NVIDIA driver VMs post work to GPU via channels Each channel is a queue of work from an app to the GPU engines GPU BAR Each VM has dedicated channels, exposed inside VM’s Virtual BAR Channels Framebuffer Graphics Compute Video Encode Video Decode Copy Engine 33 vGPU Architecture 101 Scheduling Hypervisor VM GRID vGPU Manager Tesla GPU Apps NVIDIA driver Engines sequentially execute work, signal work completions to NVIDIA driver GPU BAR Channels Scheduler Framebuffer Graphics Compute Video Encode Video Decode GPU’s hardware scheduler places work from active channels onto GPU engines Scheduler immediately loads engine with next work item Copy Engine 34 GRID vGPU Architecture 101 Memory protection Hypervisor VM GRID vGPU Manager Sysmem resource NVIDIA driver All GPU engines access memory Each vGPU is assigned fixed range of FB for its exclusive use Tesla GPU GPU’s MMU manages all engine accesses to FB and system memory Memory Management Unit (MMU) Framebuffer Graphics Compute Video Encode Video Decode Copy Engine Guest VM’s FB Prevents engines accessing outside of vGPU’s FB, or the guest VM’s system memory 35 GRID vGPU Architecture 101 Summary • vGPU dedicated resources: virtual PCIe BAR, framebuffer, channels • Guarantees uncontested access when posting work, accessing FB • vGPU shared resources: Graphics/Compute, Video Encode/Decode, Copy Engines • GPU engines are timeshared, execute in parallel • GPU hardware scheduler places work on engines • GPU MMU protects engine accesses to framebuffer and system memory 36 QUADRO: THE CHOICE OF PROFESSIONALS Enables designers to bring their visions to life Quadro: 80%+ workstation share Suite of algorithms incorporated in major design tools Source: IDC 37 QUADRO IN THE DATACENTER NVIDIA GRID delivers Quadro features anywhere, on any device Realistic Models Performance Enhancements Application Certification Improved Productivity 38 Why GPU in the Cloud? Key technologies for GPU in the Cloud AGENDA Performance & Sizing Best Practices Deployment Quick Start Summary Resources & Next Steps 39 SIZING YOUR VMWARE INFRASTRUCTURE FOR HIGH PERFORMANCE GRAPHICS 40 HIGH PERFORMANCE VIRTUAL DESKTOP IS NOT YOUR AVERAGE VDI Trade-offs CPU Memory IOPS Rich Graphics Density Trade-offs 41 4 SIZING ESX FOR RICH GRAPHICS USERS: COMPUTE VMware HCL & the NVIDIA GRID Certified Server —2 Display Adapters (onboard + GFX card) —Virtual Technology for Directed I/O (VT-d) —PCIe x16 Host CPU —Heavy 3D applications are often monothreaded —CPU frequency is more important than the number of cores • 3.1 Ghz • 10 core • 160W —Workloads will still require multiple vCPUs OS, Protocol, Improved memory performance Multi threaded rendering applications ESXi Host 42 4 SIZING ESX FOR RICH GRAPHICS USERS: MEMORY Host Memory —Audit what your workstations use today —Highest frequency memory matters —3D workloads likely require (per user): 8GB (medium workload) 16GB (most common) • 16 x 16GB • 2133Mhz 64GB (heaviest workload) ESXi Host 43 4 SIZING ESX FOR RICH GRAPHICS USER: STORAGE Storage will be the CRITICAL factor in performance CAD users >1TB of storage per desktop SIZE and PERFORMANCE matter now 1,536 Virtual SAN – SSD based local storage —Understand local storage capacity All Flash based SAN / VSAN 44 4 SIZING ESX FOR RICH GRAPHICS USERS: STORAGE Modal Benchmark – Physical vs. Virtual Environment 1444 0.02% Difference 1483 15% Difference 9181 10843 0 2000 4000 6000 8000 10000 Elapsed Time (in sec) Lower is better Test Conditions 12000 Optimized (Non-Virtual) Optimized (Virtual) Non-Optimized (Physical) Non-Optimized (Virtual) Distributed Solver PCG Solver Modal Analysis Number of Nodes 1340453 Number of Contact Elements 21763 Number of Solid Elements 833530 Number of Total Elements 862707 45 4 SIZING ESX FOR RICH GRAPHICS USERS: GPU GPU dictates interactive user experience – rotation/scaling etc GPU dictates performance when handling complex models (no. of polygons) GPU critical for high-performance graphics NVIDIA GRID – most options for vSGA, vDGA and vGPU Maximum 8 GPU cards per ESX host (2-4 typical) GRID M6 GRID M60 Number of GPUs 1 x entry Maxwell GPUs 2 x Maxwell GPUs Total NVIDIA CUDA cores 1536 4096 Total memory size 8 GB DDR3 16 GB GDDR5 Max users or vGPUs 16 32 Card Equivalent Quadro Quadro 46 4 SIZING ESX FOR RICH GRAPHICS USERS: NVIDIA GRID VGPU M60 Framebuffer GRID Virtual Applications 512 MB GRID Virtual PC GRID Virtual Workstation M60-0B M60-1B VMs per GPU VMs per M60 M60-0Q 16 32 M60-1Q 8 16 1 GB M60-1A 2 GB M60-2A M60-2Q 4 8 4 GB M60-4A M60-4Q 2 4 8 GB M60-8A M60-8Q 1 2 1 2 Pass-through 8GB 47 HOMOGENOUS VIRTUAL GPUS Tesla M60 Physical GPU 1 Physical GPU 0 Valid configuration with K240Q vGPUs on GPU 0, K260Q vGPUs on GPU 1: M60-2Q M60-2Q M60-2Q M60-2Q M60-4Q M60-4Q Valid configuration with K200 vGPUs on GPU 0, K240Q vGPUs on GPU 1: M60 -1B M60 -1B M60 -1Q M60 -1Q M60 -1B M60 -1B M60 -1B M60 -1B M60-2Q M60-2Q M60-2Q M60-2Q Invalid configuration with mixed vGPU types on GPU 0: M60-2Q M60-2Q M60-4Q M60 -1Q M60 -1Q M60 -1Q M60 -1Q M60 -1Q M60 -1Q M60 -1Q M60 -1Q 48 SIZING YOUR VIRTUAL DESKTOP FOR HIGH PERFORMANCE GRAPHICS 49 4 SIZING FOR RICH GRAPHICS WORKLOADS – VIRTUAL MACHINE Virtual Machine • Hardware v11 • VMXNET3 NIC • vGPU – Configure Profile • vDGA – Enable Passthrough vCPU vRAM • 2 vCPUs: – 1 vCPU - OS – 1 vCPU - PCoIP • 4GB – Power User/ Entry Level Eng • Recommend at least 4 vCPUs for Power Users • 8GB – Mid Eng./Video • 16GB – Advanced Eng. • 32GB – CAD/CAM OS VMware OS Optimization Tool • vDGA – 64bit Windows • MASSIVE 40% saving! • vGPU – 32/64 bit Windows • Download from labs.vmware.com • Install – VM Tools – View Agent – NVIDIA driver (vGPU) • 64GB – Digital Mock Up 50 5 VCPU IMPACT ON MEMORY PERFORMANCE 4 Cores - % Increase Memory Performance Increase from 2 vCPU to 4 vCPU 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% VM - 2 Cores/16GB vs VM - 4 Cores/16GB 86% 44% 49% 0% Read (in MB/s) Write (in MB/s) Copy (in MB/s) Latency (in ns) 44% 49% 86% 0% 51 5 SIZING FOR RICH GRAPHICS USERS: NVIDIA GRID VGPU RECOMMENDATIONS Entry Level Engineer/Reviewer Mid Level Engineer/Editor Advanced Engineer • M60-1Q • M60-2Q • M60-4Q • 4 GB RAM • 8 GB RAM • 16 GB RAM • 2 VCPUs (2.6 GHz) • 2-4 VCPUs (3.1 GHz) • 4-8 VCPUs (3.1 GHz) • 1 GB video RAM • 2 GB video RAM • 4 GB video RAM • 2 2560x1600 displays • 4 2560x1600 displays • 4 2560x1600 displays • 4 x 4K displays 52 5 SIZING FOR RICH GRAPHICS USERS: CLIENT DEVICE 3D and multimedia application can drive high FPS at high resolution GPU throughput uncompressed over 2 Gbps Protocol optimization, compression and bandwidth savings —Only send pixels that change etc Plan for 2-2.5 Mbps for CAD users (average) Burst to 5-7 Mbps, can be higher in LAN scenarios (20Mbps-200Mbps depending on hardware) Client device must be capable of decoding PCoIP throughput or H264 offload —Tera2 client or suitable spec Client —Check your device specifications 53 5 VMWARE BLAST EXTREME ACCELERATION WITH NVIDIA GRID 55 CHALLENGES FOR VIRTUAL GRAPHICS Professional graphics workloads require great user experience UX Require “snappy” experience Rely on heavy encoding and decoding User density is limited by CPU bottleneck 56 NVIDIA BLAST EXTREME ACCELERATION Increased scalability and performance for graphics workloads Up to 51ms latency reduction Up to 89% bandwidth reduction Up to 18% increase in density Source: NVIDIA GRID Performance Engineering Lab 57 HOW IT WORKS TODAY: PCoIP SERVER with GRID GPU CPU NIC Encode Render CLIENT IP Network Kybd/Mse Decode Render Capture GRID GPU WORKLOAD NON GPU WORKLOAD 58 NVIDIA BLAST EXTREME ACCELERATION SERVER with GRID GPU CPU NIC Encode Render CLIENT IP Network Kybd/Mse Decode Render Capture GRID GPU WORKLOAD NON GPU WORKLOAD 59 NVIDIA BLAST EXTREME ACCELERATION Apps Apps Apps Remote Client H.264 / H.265 streams Graphics commands GRID GPU HW Encoder 3D Context/Display Capture Render Target Front Buffer • Reduces overall latency • Offloads CPU workload to GPU • Increases scalability • Improves user experience • Lowers N/W bandwidth demand Framebuffer 60 NVIDIA BLAST EXTREME ACCELERATION IMPROVES USER EXPERIENCE AND REDUCES TCO! • Scalability • 18% higher user density for ESRI ArcGIS Pro 1.1 • User Experience • 6-13% higher average FPS • 27-51ms less latency • 48-89% bandwidth reduction with TCP and 19% with UDP You don’t have to chose, you get all at the same time Source: NVIDIA GRID Performance Engineering Lab 61 PERFORMANCE RESULTS 62 M60 OFFERS PERFORMANCE IMPROVEMENTS NVIDIA GRID DASSAULT CATIA V5/V6 SCALABILITY GUIDE 35 Number of Concurrent users on a single server 30 25 20 15 10 5 0 Light Users Fewer Applications Task Worker Aerospace Manufacturing 32 32 Production CATIA workloads Satellite offices, over WAN links GRID K2 GPUs and M60 GPUs Four different user profiles 16 16 16 Frame Rate Limiting (FRL) turned on Results 8 Doubled the number of users per host Observed a 70% productivity improvement Standard 2vCPU Moderate 4vCPU Power User 4vCPU 16GB RAM proximity/transfer 32GB RAMtime savings64GB RAM File 1GB Frame Buffer 1GB Frame Buffer 1GB Frame Buffer / No reported degradation user / M60-1Q) (K240Q M60-1Q) (K240Q / M60-1Q) to end (K240Q K2 M60 Even LAN users reported 30% productivity gain Knowledge Worker Power User ** 2.6 GHz+ Intel Xeon E5 v3 or greater 8 4 Analyst 4vCPU 128GB RAM 1GB Frame Buffer (K240Q / M60-1Q) Heavy Users Many Applications Designer 63 6 VIRTUAL BATTLESPACE RESULTS 64 4/4/2016 65 Why GPU in the Cloud? Key technologies for GPU in the Cloud AGENDA Performance Best Practices and Results Deployment Quick Start Summary Resources & Next Steps 66 GETTING STARTED Key concepts to get you going Deployment Requirements Install host driver Attach a vGPU profile to a VM Product Activation Keys (PAK) The GPU Mode Switch Utility Install the NVIDIA driver into a VM Download GRID SW (host driver, guest driver) License server creation, installation and management Configure a VM for a NVIDIA GRID vGPU license 67 DEPLOYMENT REQUIREMENTS NVIDIA GRID enterprise software is designed to run on select NVIDIA Tesla GPUs. Ensure you have the following: NVIDIA GRID Certified Servers: http://www.nvidia.com/object/grid-certifiedservers.html Required GPU hardware: Tesla M6 or M60 A supported virtualization software stack Read the release notes! Correct build numbers for the virtualization software that have been tested for the specific NVIDIA GRID Software release Other important version-specific information Note: You will need to register first to get access to the release notes. 68 QUICK START GUIDE Overview of concepts to get you going Deployment Requirements Install host driver Attach a vGPU profile to a VM Product Activation Keys (PAK) The GPU Mode Switch Utility Install the NVIDIA driver into a VM Download GRID SW (host driver, guest driver) License server creation, installation and management Configure a VM for a NVIDIA GRID vGPU license 69 PRODUCT ACTIVATION KEYS (PAK) Order email template Email content for GRID SW Order: • Product Activation Key (PAK) is dynamically added to order fulfillment email template • Blue underlined content are links to those destinations for registering, logging in or support • Enterprise support portal and email address is provided for help • Email generated from nvidia@flexnetoperations.com (advise on checking spam folders as needed) Sample email 70 NVIDIA LICENSING PORTAL Registration Page • Customers will click on the Register or Log in link within the order fulfillment email to navigate to the NVIDIA Licensing Portal • First time customers will be required to register their information • A link to the Terms and Conditions is provided (End User License Agreement) • Customers must accept the terms and conditions to download licenses and software • Returning customers will not have to reregister but can login to their account 71 QUICK START GUIDE Key concepts to get you going Deployment Requirements Install host driver Attach a vGPU profile to a VM Product Activation Keys (PAK) The GPU Mode Switch Utility Install the NVIDIA driver into a VM Download GRID SW (host driver, guest driver) License server creation, installation and management Configure a VM for a NVIDIA GRID vGPU license 72 DOWNLOAD GRID SOFTWARE The Product Download Page contains… NVIDIA GRID Software: Select based on the hypervisor brand and version you will run. NVIDIA GRID vGPU supports versions of VMware vSphere ESXi and Citrix XenServer GPU mode change utility: The Tesla M60/M6 boards are shipped in Compute Mode. NVIDIA GRID vGPU requires Graphics Mode. The GPU mode change utility is used to switch modes. NVIDIA License Manager Software: Enables Guest OS to utilize the NVIDIA GRID vGPU profiles by checking out a license at boot time All downloads are in .zip format and contain release notes and user guides along with the software. 73 QUICK START GUIDE Key concepts to get you going Deployment Requirements Install host driver Attach a vGPU profile to a VM Product Activation Keys (PAK) The GPU Mode Switch Utility Install the NVIDIA driver into a VM Download GRID SW (host driver, guest driver) License server creation, installation and management Configure a VM for a NVIDIA GRID vGPU license 74 THE NVIDIA GRID LICENSE SERVER UX NVIDIA GRID vGPU profiles are licensed features on NVIDIA Tesla GPUs • Licenses generated and retrieved from NVIDIA’s Licensing Portal • Installed onsite, enables NVIDIA GRID licensed products • Manages a pool of floating licenses • License acquired during boot time of the Guest OS • The license returned when the Guest OS shuts down Use for license management Enable more use cases Don’t impact work 75 QUICK START GUIDE Key concepts to get you going Deployment Requirements Install host driver Attach a vGPU profile to a VM Product Activation Keys (PAK) The GPU Mode Switch Utility Install the NVIDIA driver into a VM Download GRID SW (host driver, guest driver) License server creation, installation and management Configure a VM for a NVIDIA GRID vGPU license 77 MODE SWITCHING Tesla M60 Tesla M6 Compute Mode Graphics Mode Shipped in this mode Required for NVIDIA GRID gpumodeswitch Bootable Linux or Windows Temporarily boot Linux on the server to change mode; the gpumodeswitch release package includes bootable Linux images for this purpose 78 GPUMODESWITCH EXAMPLE List the current mode of all GPUs use -listgpumodes Switch a GPU’s mode use --gpumode graphics --gpumode compute After a GPU mode switch, the server is rebooted to ensure that the modified resources are correctly accounted for by the OS or hypervisor 79 QUICK START GUIDE Key concepts to get you going Deployment Requirements Install host driver Attach a vGPU profile to a VM Product Activation Keys (PAK) The GPU Mode Switch Utility Install the NVIDIA driver into a VM Download GRID SW (host driver, guest driver) License server creation, installation and management Configure a VM for a NVIDIA GRID vGPU license 80 NVIDIA GRID WITH CONFIGURATION PROCESS • Install NVIDIA GRID software into ESXi Hypervisor • Build a Gold Master VM Image with NVIDIA GRID vGPU Profile • Install NVIDIA Windows Driver into Gold Master Image • Configure vGPU enabled VM pools in Horizon Environment For more details consult the GRID Deployment Guide and additional online training 81 INSTALLING AND TESTING VIB $ esxcli system maintenanceMode set –e true $ esxcli software vib install -v /store/NVIDIA.vib $ reboot Installation Result: Message: Operation finished successfully. Reboot Required: false VIBs Installed: NVIDIA_bootbank_NVIDIA-vgxVMware_vSphere_6_Host_Driver_346.27-1OEM.600.0.0.2159203 VIBs Removed: VIBs Skipped: $ esxcli system maintenanceMode set –e false $ esxcli software vib list | grep -i nvidia $ nvidia-smi NVIDIA CONFIDENTIAL. DO NOT DISTRIBUTE. 82 WINDOWS VM: VGPU ASSIGNMENT In vCenter, edit VM settings, add “Shared PCI Device”, then select “Add” Select GPU profile, then select “OK” Start VM 83 NVIDIA DRIVER INSTALL 84 HORIZON POOL CREATION 85 QUICK START GUIDE Key concepts to get you going Deployment Requirements Install host driver Attach a vGPU profile to a VM Product Activation Keys (PAK) The GPU Mode Switch Utility Install the NVIDIA driver into a VM Download GRID SW (host driver, guest driver) License server creation, installation and management Configure a VM for a NVIDIA GRID vGPU license 86 CONFIGURE NVIDIA GRID VGPU LICENSE Enables a VM access to an NVIDIA GRID vGPU Configure the Manage License Feature in the NVIDIA Control Panel • Enter the license server address and port • These settings can be preloaded through register keys. • The license settings will persist across reboots. • A license is obtained prior to the user logging into the VM. 87 April 4-7, 2016 | Silicon Valley L6117 - NVIDIA GRID DEPLOYMENT HANDS ON LAB Room 210C 4/7/2016 13:00 - 16:00 Why GPU in the Cloud? Key technologies for GPU in the Cloud AGENDA Performance Best Practices and Results Deployment Quick Start Summary Resources & Next Steps 89 VMWARE HORIZON: DESKTOP TRANSFORMATION SOLUTION FOR EVERY USER Virtual desktop and app delivery Desktop and app delivery from public cloud Horizon 7 Horizon Air APP NEW MAIN OFFICE REMOTE OFFICE POINT OF SALE Simplifying App Management NEW DEVELOPERS Horizon FLEX Horizon App Management NEW DESIGNERS/ ENGINEERS Secure containerized desktops NEW MOBILE NEW NON-EMPLOYEE 90 9 NVIDIA GRID PLATFORM Centralized virtual graphics capabilities for the enterprise Support, Updates & Maintenance NVIDIA GRID Software NVIDIA Tesla (Datacenter GPU) 91 “ “ Streamlined end user access and supplier connectivity When I use SolidWorks on virtual desktops, it’s seamless… from my phone, my iPad, and my computer. ” – Mechanical Engineering Student North Carolina State University Real-time collaboration for engineers in 29 offices across 7 countries We are confident that it is now finally economically viable for CH2MHILL to provide a VDI farm for graphically demanding workloads across the globe. This could truly transform the way we do business. ” – Craig Fletcher CTO CH2MHILL 33% productivity gain, access to radiology images on a variety of endpoints US Based Healthcare Provider 92 9 Why GPU in the Cloud? Key technologies for GPU in the Cloud AGENDA Performance Best Practices and Results Deployment Quick Start Summary Resources & Next Steps 93 TEST DRIVE NVIDIA GRID TODAY http://www.nvidia.com/trygrid Instant access to NVIDIA GRID on VMware vSphere and Horizon Pre-installed apps such as AutoCAD, SolidWorks, ArcGIS, NX, etc. HTML demos and HD videos 94 KEYS TO A SUCCESSFUL POC Important things to remember during your POC: Define “acceptable” user experience Render times, smoothly panning, menu to appear Compare real world workloads With “think time” included, opposed to peak workload benchmarks Test with real users 3D designer in the application, ultimate test of success 95 DETAILED INSTRUCTIONS Deployment guides www.nvidia.com/gridresources 96 OTHER NVIDIA GRID SESSIONS TO ATTEND S6504 - A Data-Driven Methodology for NVIDIA GRID™ vGPU™ Sizing S6392 - AEC Project Execution Using GRID vGPU Enhanced Virtualization S6859 - Unveiling the Impact of Time Slicing with NVIDIA GRID™ vGPU for Realistic ROI/TCO Analysis S6210 - NVIDIA GRID™ and Dassault Catia from Proof of Concept to Production S6194 - Delivering Graphics-Intensive Applications to Computing Labs and BYOD in Education S6833 - Delivering GPU-Accelerated Applications from your Private Cloud S6209 - A Look at Real World Performance Capabilities of NVIDIA GRID™ 2.0 S6198 - The Latest in High Performance Desktops with VMware Horizon and NVIDIA GRID™ vGPU S6200 - Delivering 3D Workstations with VMware Horizon and NVIDIA GRID L6117 - NVIDIA GRID Hands-on Lab S6332 - Training and Simulation in a Secure Cloud Environment 97 April 4-7, 2016 | Silicon Valley THANKS – Q&A PHOTO CAPTION 99