three quarters_new overlay.png

ZX9210

by

ZNYX Networks

http://www.znyx.com/products/hardware/zx9210.jsp
Contact:
Kevin Austin
Description:
Virtualized environments have become a key requirement for next-generation applications, reducing system infrastructure costs and offering improved efficiency and scaling. The ZX9210 allows System Architects the highest performance Intel® Xeon® E5-2600 Series processors with the largest memory footprint available in an AdvancedTCA® (ATCA) compute blade to host up to 32 virtual machines (VMs) in a single compute blade, each with a dedicated hardware thread.

The ZX9210 is a dual socket compute blade that can be configured with a single E5-2620 (6C/12T) for entry-level cost-effective solutions up to dual E5-2648L or E5-2658 (8C/16T) processors for high-end configurations. The ZX9210 includes support for "Ivy Bridge" processors, which allow customers to leverage advances in Intel Architectures (IA) without the need to qualify a new compute blade.
With 16 DDR3 memory sockets, the ZX9210 provides the highest number of sockets on any ATCA blade. This allows for a maximum of 512GB of memory and serves as a cost-effective configuration for applications with lower memory requirements. The additional sockets can be used to optimize memory by utilizing lower-cost, low-density DIMMs.

The ZX9210 includes dual USB and DVI video on the front panel, as well as dual Gigabit Ethernet ports for direct connection to the processor and a single Gigabit Ethernet port for access to the base interface network. Two mSATA sites enable up to 512GB of high-speed local storage with additional storage capabilities via an RTM. Dual 10 Gigabit links are routed to an RTM for applications that require up to 20 Gb/s external connectivity. The ZX9210 supports the Oracle® RTM specification, ensuring compatibility with a wide-range of RTMs available from multiple vendors.
Designed to handle a large number of VMs, the ZX9210 includes several hardware-based virtualization technologies to improve VM performance, including:

Intel Data Direct I/O - provides direct access to processor cache. This improves the LLC hit rate, relieves congestion on the memory sub-system which frees up bandwidth and provides additional power savings

SR-IOV - Shares a single I/O resource between multiple virtual machines.

Flexible Port Partitioning - Efficiently divides physical Ethernet ports into multiple virtual devices, providing QOS
Categories:
AdvancedTCA® AdvancedTCA® > AdvancedMC Form Factors Supported > (ATCA Node Board) Physical Number of Processors on Board > (ATCA Node Board - # CPUs) 4 or more AdvancedTCA® > AdvancedMC Form Factors Supported > (ATCA Node Board) Physical Number of Processors on Board > Two Physical CPUs Installed AdvancedTCA® > AdvancedMC Form Factors Supported > (ATCA Node boards) Fabrics Supported AdvancedTCA® > AdvancedMC Form Factors Supported > (ATCA Node boards) Fabrics Supported > (ATCA Node boards) PICMG 3.1 Option 1 (1G Ethernet) AdvancedTCA® > AdvancedMC Form Factors Supported > (ATCA Node boards) Fabrics Supported > (ATCA Node boards) PICMG 3.1 Option 9 (10G Ethernet) AdvancedTCA® > AdvancedMC Form Factors Supported > (ATCA Processor boards) Processor Architecture AdvancedTCA® > AdvancedMC Form Factors Supported > (ATCA Processor boards) Processor Architecture > (ATCA Processor boards) x86 Based