SmartVDI™ Converged Infrastructure Platforms
- VMware Software-defined Components
- Optimized for Defined User Populations
- Designed for NVIDIA GRID adapter compatibility
- Focus on Performance-based Users
SmartVDI™ Converged Infrastructure Platforms
ClearCube’s Smart Virtual Desktop Infrastructure (SmartVDI™) host platform smoothly transitions your workers from using physical Personal Computer desktops to using highly available virtual desktops that can deliver an equal to, or better than, end user experience. SmartVDI desktops are capable of handling complex configurations with the same personalization features as physical PCs.
Incorporating VMware’s core VDI components, SmartVDI uses VMware’s vSAN to cluster storage and deliver high availability and Horizon 6 to broker zero client connections. SmartVDI hosts feature dual Intel® Xeon® processors, up to 512GB DDR4 RAM, 10GB Ethernet and redundant power supplies. A PCoIP Hardware Accelerator Card, and NVIDIA GRID K1 or K2 graphics cards are added to SmartVDI for user environments running simultaneous video intensive applications such as YouTube, data visualization, or video monitoring. Up to two NVIDIA GRID adapters are supported per SmartVDI 1U rack mount system.
SmartVDI’s scalable open stack architecture allows you to integrate the system into existing virtualization environments to target specialized power user groups that need high powered graphics or for general audiences that include task and knowledge workers.
SmartVDI delivers a phenomenal high performance user experience that is easy to support because it utilizes VMware software-defined components with optimized high availability and storage with low cost-per-seat configurations.
SmartVDI’s hardware architecture is very straightforward, consisting of three primary components:
- 1U dual Xeon processor systems featuring SSD drives for caching, spindle drives for storage, dual 10GBe ports for fast network throughput, and 64-512GB RAM to host the virtual desktops
- Hardware acceleration using NVIDIA GRID K1 or K2 adapters and PCoIP Offload adapters
- A range of dual and quad display zero client options
VDI Building Blocks
Incorporating VMware’s core VDI components, SmartVDI uses VMware’s vSAN to cluster storage and deliver high availability and Horizon 6 to broker zero client connections. SmartVDI hosts feature dual Intel® Xeon® processors, up to 512GB DDR3 RAM, 10GB Ethernet and redundant power supplies. A PCoIP Hardware Accelerator Card, and NVIDIA GRID K1 or K2 graphics cards are added to SmartVDI for user environments running simultaneous video intensive applications such as YouTube, data visualization, or video monitoring. Up to two NVIDIA GRID adapters are supported per SmartVDI 1U rack mount system.
SmartVDI™ Host Platform
Full desktop quality computing experience via an IP connection
SmartVDI™ High-Availability Configuration
SmartVDI is configured, tested, and supported as a pre-integrated standards-based VMware VDI solution using VMware core components such as Horizon 6 with View and vSAN for clustered storage..
Designed for mission-critical environments that need 24/7/365 uptime, SmartVDI’s incorporation of VMware’s vSAN software provides clustered storage and high availability across SmartVDI nodes. Using three systems as the foundation, if a failure of any one system occurs, virtual desktops are migrated to the remaining two systems allowing work to continue.
From the storage perspective, the vSAN storage implementation is always up and available if a node fails. From a user perspective, like any normal SAN with VMware HA, the desktops reboot on another host and users simply log back in. That is a huge plus for any business or operation that cannot afford worker downtime.
As a converged compute and storage device, SmartVDI scales well. Once the high availability core foundation is set up with as many SmartVDI Hosts and the vSAN storage is accessible by all hosts, adding more users just means adding more SmartVDI nodes. No external storage is required as each host contains all the needed hardware components and actively shares with other nodes in the environment.
SmartVDI™with K2 GRID adapters
Power users have more demanding CPU, memory and GPU requirements than knowledge users so shared GPU configurations are supported within SmartVDI using NVIDIA GRID Virtual GPU (vGPU).by dedicated GPU per user configurations.
SmartVDI hosts complement ClearCube’s zero clients, Blade PCs, and Sentral single console management in CVDI (Centralized and Virtualized Desktop Infrastructure) and VDI projects to provide you with a single vendor solution with desktop-to-datacenter vision and accountability.
Leveraging ClearCube’s experience as the first manufacturer to embrace PC-over-IP™ (VMware’s native VDI protocol), ClearCube’s in-depth PC-over-IP™ protocol knowledge produces a performance- optimized VMware-centric system like no other.
ClearCube’s objective is to transition you from the physical to the virtual PC world in the least disruptive manner possible. We start with the objective of identifying the requirements of your specific user population, and configuring the appropriate hardware resources to satisfy their needs at the optimal price point. Typically this results in multiple SmartVDI systems that are identically matched to meet the user population application requirements.
Unique to ClearCube customers, “specialized needs” in virtualization projects can be satisfied with centralized Blade PCs and zero clients to reduce implementation risks. Blade PCs and SmartVDI complement each other and utilize the same stateless, highly secure zero clients at the desktops. However, Blade PCs have the capability to run applications that cannot be virtualized, legacy applications, Linux operating system/applications, complex 3D graphics applications and CPU-intensive applications. This Blade PC option offers ClearCube customers a safety net in implementing their virtualization projects, knowing that there is a solution for every targeted user station in the project, and that no one will be dissatisfied. The net result of working with ClearCube is that you will do more, with less, for less.
SmartVDI™with K1 GRID adapters
SmartVDI systems are configured with K1 GRID adapters for knowledge workers using applications such as Google Earth, internet video streams, etc.
SmartVDI™ Converged Infrastructure Platform Example Configurations
Three SmartVDI systems are needed to establish a high availability configuration using VMware’s vSAN software layer. Because no administrator wants to risk the loss of services in production level deployments, almost all SmartVDI configurations start with three systems, and scale upward.
SmartVDI is configured, tested, and supported as a pre-integrated standards-based VMware VDI solution using VMware core components such as Horizon 6 with View and vSAN for clustered storage.
Click for larger image
SmartVDI with K1 GRID adapters/vSGA
Virtual Shared Graphics Acceleration (vSGA) is designed so that multiple virtual machines running in SmartVDI memory leverage physical GPUs installed locally in the system to provide hardware-accelerated 3D graphics.
SmartVDI systems are configured with K1 GRID adapters for knowledge workers using applications such as Google Earth, internet video streams, etc. (Open GL 2.1, DirectX 9). GPUs on the K1 adapters are shared (vSGA) by the users hosted on each SmartVDI system. As many as 64 users are supported on one SmartVDI system with 256GB RAM and two K1 GRID adapters. However, because three units are required for high availability, the VDI load is spread across all the systems with sizing to accommodate the loss of one system and the migration of those users to the remaining systems. The ClearCube sales team can help you calculate the optimum number of users supported by SmartVDI based on your specific variables.
Click for larger image
SmartVDI with K1 GRID adapters/vDGA
Virtual Dedicated Graphics Acceleration (vDGA) maps one virtual machine running in SmartVDI memory to a single physical GPU on the K1 GRID adapter to provide hardware-accelerated workstation graphics where a discrete GPU is needed.
Power users have more demanding CPU, memory and GPU requirements than knowledge users so shared GPU configurations are replaced within SmartVDI by dedicated GPU per user configurations. The GPUs each have 4GB of RAM and support the latest graphics libraries including NVIDIA CUDA for those higher end applications. Each K1 adapter has four GPUs and supports four users users with 192 processing cores per user. Therefore, up to eight users are supported per SmartVDI system.SmartVDI with K2 GRID adapters/vDGA
Designers and developers of 3D graphics content with complicated datasets may be best served using SmartVDI with K2 GRID adapters. Each K2 card has two GPUs so a total of four users can be virtualized per SmartVDI system. This gives each user dedicated graphics with 4GB of RAM and 1536 processing cores for the virtualized power user.
Click for larger image
SmartVDI with K1/K2 GRID adapters/vGPU – True Shared Pass Through
NVIDIA GRID vGPU enables multiple virtual machines (VMs) to have simultaneous, direct access to a single physical GPU, using the same NVIDIA graphics drivers that are deployed on non‐virtualized operating systems. This gives VMs unparalleled graphics performance and application compatibility, together with cost‑effectiveness and scalability brought about by sharing a GPU among multiple
ClearCube SmartVDI w/ GRID has evolved with each new GPU capability’s realization. SmartVDI with GRID delivers the capability for multiple virtual desktops sharing a single physical GPU, and multiple GPUs on a single physical PCIe card, providing the long awaited 100% application compatibility of pass-through graphics with lower cost of multiple desktops sharing a single graphics card to provide a rich more cost effective user experience.
With SmartVDI, VMware, NVIDIA GRID with vGPU technology, the graphics commands of each virtual machine are passed directly to the GPU, without translation by the hypervisor. This allows the GPU hardware to to deliver the ultimate in shared virtualized graphics performance. The diagram below represents the vGPU profile K260Q from the table below.
Click for larger image
SmartVDI takes full advantage of vGPU profiles
vGPU profiles allow customized, dedicated graphics memory to be managed by the GRID vGPU manager for simple management of user profiles with a single-pane-of-glass management console. SmartVDI w/GRID vGPU enables IT managers to easily assign the optimal amount of graphics memory and deliver a customized graphics profile to meet the specific needs of each user.
GRID vGPU enables up to eight users to share each physical GPU, assigning the graphics resources of the available GPUs to virtual machines in a balanced approach. Each NVIDIA GRID K1 card has four GPUs, allowing 32 users to share a single card and up to 64 users on a system. Each NVIDIA GRID K2 card has two GPUs, allowing 16 users to share a single card and up to 32 users on a system.
SmartVDI plus ClearCube's Family of Blade PCs & Engineering Workstations Assure Satisfied Power Users
The advantage to working with ClearCube is that our SmartVDI solution is complemented by our portfolio of PCoIP-enabled Blade PCs with dedicated NVIDIA Quadro series GPUs (K2000, K4000, K5000, K6000) for those graphics oriented use cases that don’t fit well in a pure VDI environment . Like SmartVDI, Blade PCs use zero client end points at the desktops and in most cases can be managed by the same VDI tools. Blade PCs may more cost-effectively address certain user profiles, a smaller number user community, specialized applications written for Linux, and a variety of other instances. A blend of Blade PCs and SmartVDI systems may provide you with the very best options for GPU-oriented users. Please consult with ClearCube sales for the best configuration options based on your specific requirements.