Nutanix announces 5 new models including NX-7110 optimized for graphics performance

Nutanix today announced that it will extend its line of converged storage and compute platforms with a new model optimized for heavy graphical usage like 3D and CAD. Also the current NX-3000 and NX-6000 line of servers is extended with 4 new models.

One of the use cases of the Nutanix Virtual Computing Platform is VDI. Because disks (Nutanix used SSD and HDD) are in the same box as the CPU’s performance of storage is improved compared to the traditional way of presenting storage like SAN and NAS. VDI is notorious for being dependant on storage performance. When storage does not perform, end users will complain about bad performance.

New is the Nutanix NX-7110 platform. This server further enhances the VDI user experience by enhancing video performance. This server has three PCIe slots which can be used by NVIDIA K1, NVIDIA K2 and Teradici APEX cards for  server offload..

The Teradici card offloads the PCoIP image encoding tasks thus reducing the server CPU utilization. This enables a better user experience and also allows a better VM to CPU consolidation. More VM’s per CPU reduces costs. PCoIP is only supported by VMware View.

The NVIDIA cards allow GPU virtualization which means the graphical processor can be virtualized and shared by multiple users. This greatly enhances the user experiences of Server Based Computing (Micrsoft RDS and Citrix XenApp) as well as Virtual Desktop Infrastructure solutions like Citrix XenDesktop and VMware View.

Using the Nutanix NX-7110 usage of AutoCAD, Autodesk, SolidWorks CAD and 3D software is possible on VDI and SBC.

The image below shows the hardware specs of the NX-7110. Mind there is only one node per 19″ appliance. Other Nutanix appliances/enclosures allow up 4 nodes per enclosure. Also NVIDIA K1 and K2 cards cannot be mixed.

Nutanix also announced four new models in the current NX-3000 and NX-6000 series. The NX-3060, NX-3061, NX-6060 and NX-6080 have two Intel Ivy processors with 10 cores each per socket. Also the nodes can have up to 512 GB of internal memory. This allows an up to 30% greater VM density per node


Add a Comment

Your email address will not be published. Required fields are marked *

Current ye@r *