Hyper-Unity® is the first turn-key, MIL SPEC, SWaP-optimized, hyper-converged infrastructure platform, to deliver all-flash performance for virtualized applications at less than half the cost of traditional storage. It seamlessly integrates two to twelve ruggedized TMS servers, a high-speed, 12-port Mellanox InfiniBand switch, and leading software-defined storage(SDS) / hyper-converged infrastructure (HCI) software.
Ruggedized for Wherever You Need to Go
The Hyper-Unity infrastructure platform features industry-leading performance with enhanced reliability features that ensure survivability in the most demanding environments, and where SWaP is an important consideration. RES rackmount servers provide superior resilience to shock, vibration, and temperature extremes
Rapid, Cost-Effective and Easy Deployment
TMS’ Hyper-Unity platforms deliver all-flash performance at lower cost than traditional storage or competitive hyper-converged solutions.. Instead of integrating costly and complex SAN, NAS or all-flash array storage with servers, Hyper-Unity delivers an all-flash hyper-converged platform, comprising pre-integrated nodes, each with its own compute, enterprise-class SSD storage, SDS, high-performance networking, and virtualization. Organizations can quickly deploy a turnkey solution cluster that provides from 96 up to several hundred CPU cores, from 1 to 6 TB of RAM and 6 to 25 TB of all-flash storage to support several hundreds of server VMs with 12 to 100s of TB of effective storage capacity.
Scale Out in Increments
Hyper-Unity allows you to scale both capacity and performance linearly by providing the flexibility to scale up by adding more nodes or scale up by adding more disks to each server node. Instead of buying large and expensive storage arrays with limited performance, the Hyper-Unity Infrastructure enables enterprises to scale out in small, modular units, allowing them to balance their compute and storage resources while growing. It also supports popular hypervisors, preserving prior investments in tools, time, and training.
Hyper-Unity delivers the performance of an all-flash array in multi-server clusters, enabling IT organizations to easily meet the performance requirements of demanding workloads. Hyper-converged software performs de-duplication and compression before data is written to storage, leading to extremely low latencies, higher IOPS and lower storage and network traffic.
The rapid proliferation of virtualized server guest OS instances and associated VM sprawl for databases, web servers, application servers, and other VMs have outstripped the ability of traditional storage systems to adequately service storage traffic.
In addition to proliferating capacity requirements, virtualized servers require high performance storage that delivers the IOPS, throughput and latency required to make applications perform their best. The purchase of traditional SAN and NAS storage to meet this requirement is costly and, increasingly, exceeds budgets. The Hyper-Unity platform typically cuts data center infrastructure costs by 50 percent or more while delivering all-Flash performance.
Data Reduction & I/O Acceleration
Hyper-Unity lowers costs, using data reduction and I/O acceleration technology to reduce the amount of physical RAM and local flash required per TB of storage capacity. Hyper-Unity eliminates the need for multiple tiers of flash and hard drives. Hybrid hyper-converged systems remain more costly and lack the level of performance delivered by TMS’ Hyper-Unity platforms.
- Superior resilience to shock, vibration, and temperature extremes
- Four-node base solution cluster – for high availability and performance
- Four RES-XR5-1U, eight drive servers with two E5-2600 V3 Series Intel Xeon processors, fourteen cores per socket, and up to 512 GB DDR4 ECC DIMMs
- Base cluster has 6 TB of SSD raw capacity from 12 to over 30 TB effective storage capacity
- Easy scale-out growth, by adding additional nodes to solution cluster (4-12 nodes)
- Base 1.5TB per node SSD capacity is expandable up to 16 TB per node today
- Integral high-speed, low-latency Mellanox InfiniBand data network – 56Gb IB, 56/40GbE
- 1GbE resource management network for node configuration and out-of-band management
- Removable fans
- Power supply options Single or redundant 110/220 VAC (50/60Hz, 400Hz)
- Single or redundant 18-36 VDC, 32 Amp
- Single or redundant 36-72 VDC, 18 Amp
- Hot pluggable solid state drive slots (32 per base cluster)
- Operating temperature range: 0°C to 50°C
- Operating shock: 3 axis, 35g, 25ms
- Operating vibration: 4.76 Grms, 5Hz to 2000Hz (SSD)
- Operating humidity: 8% to 95% non-condensing
- MIL-STD-810G, MIL-S-901D, MIL-STD-167-1*
Hyper-Unity Building Blocks
RES-XR5-1U, 8 Drive Servers
Featuring up to (2) Intel Xeon processors with up to fourteen cores per socket, 1TB DDR4 ECC memory, (8) SAS/SATA drives, and enhanced reliability features, the Themis RES-XR5-1U server provides industry-leading performance and superior resilience to shock, vibration, and temperature extremes. Part of Mercury’s EnterpriseSeries, the Themis RES-XR5 is designed for military, industrial, or rugged commercial applications. RES-XR5-1U servers keep mission-critical applications available in demanding environments where SWaP is an important consideration. The RES-XR5-1U server provides configuration versatility and system expansion to meet your current and future system requirements and can be mounted in standard commercial racks or mobile rugged transit cases.
TMS Resource Manager
The TMS Resource Manager is a cost-effective, out-of-band system appliance that facilitates administration, management, and monitoring of up to 22 systems.
The Resource Manager server attaches to one or more managed systems though an independent, private network and presents a single user interface for easy access to the baseboard management controllers (BMC) of the managed systems, displaying information from these BMCs as a single integrated unit.
Functions include system health; operating temperature, fan speed, critical voltages, etc. In addition, the Resource Manager provides console access to the individual managed systems via the BMC KVM feature. This interface enables configuration and monitoring of critical in-band functions, including BIOS setup, O/S boot-up as well as execution of O/S run-time commands. The KVM function also supports the virtual media function, enabling software loading and updates from a single, network-attached client. The RES Manager engine enables in-band reporting from agents that run in the host being managed, reporting CPU utilization, memory capacity, storage and network usage.
The Resource Manager is extensible by adding software functional adapters that can access additional, third party devices, such as switches and other functions not part of the core appliance.
A 12-port managed 56 Gb/s InfiniBand VPI SDN/40GBe Mellanox InfiniBand switch delivers up to 1.3Tb/s of non-blocking bandwidth and 200ns port-to-port latency. The switch module enables data centers to scale out with fourteen data rate (FDR) InfiniBand. Mellanox switching technology provides industry-leading performance, power, and density. The switch also interoperates with existing networks, supports software defined networks (SDN), and provides fabric management for cluster and converged I/O applications.
|Configuration, Processors, and Memory|
|Server Compute (per node)||Themis RES-XR5-1U, 8-drive server with E5-2600 V3 Series Intel® Xeon® processors (up to fourteen cores per socket)|
|Server Memory (per node)||Up to 512 GB DDR4 ECC RAM|
|Hypervisor (customer licensed per node)||VMware vSphere or Citrix XenServer|
|Data Network (per node)||Dual 56 Gbps/40Gbps Mellanox Infiniband/Ethernet Ports|
|Management Network (per node)||Data Network Fabric
12-Port Mellanox 56Gbps FDR Infiniband Switch
|Data Network Fabric||TMS RES-Manager 1U Resource Manager with 22 Gigabit Ethernet Ports|
|Management Network/Switch||TMS RES-Manager 1U Resource Manager with 22 Gigabit Ethernet Ports|
|On Board Expansion|
|Expansion slots (2)||Two PCIe 3.0 16X full size cards|
|Front Panel and Rear Panel I/O Per Node|
|Removable 2.5 Inch SATA or SAS disk drives||Up to 8||Front panel|
|CD-RW/DVD-RW drive||1||Front panel|
|Status LEDs||7||Front panel|
|Gigabit Ethernet ports (RJ45)||2||Rear panel|
|USB 3.0 ports||2 standard||Rear panel|
|IPMI 2.0 port||1||Rear panel|
|Power connector||1 or 2||Rear panel|
|Power switch||1||Rear panel|
|VGA Port (graphics)||1||Rear panel|
|Modular Maintainability Per Node|
|Power Supply Options||Single or redundant 110/220 VAC (50/60Hz, 400Hz)
Single or redundant 18-36 VDC, 32 Amp
Single or redundant36-72 VDC, 18 Amp
MIL-STD-461 options available
|Hot Pluggable Disk Drives||8 slots per node|
|Temperature Range||-40° C – 70° C||0°C – to 50°C|
|Humidity (non-condensing)||5% to 95%||8% to 95%*|
|Shock||3 axis, 35g, 25 ms||3 axis, 35g, 25 ms|
|Vibration (10-2000Hz)||4.76 Grms, 5Hz to 2000Hz (SSD)||4.76 Grms, 5Hz to 2000Hz (SSD)|
|Dimensions||Height: 1U or 1.75 inches (44.45 mm)
Width: 17 inches (432 mm)
Depth: 20 inches (508 mm)
|Weight||20 pounds (9.1 kg)|
|Chassis Features||Coated aluminum for light weight and corrosion resistance
Stainless steel in selected areas to add strength and stiffness
Modular design for easy upgrade and service
Optional rack-mount slides and dust filter
Airflow direction: front to rear
|MIL-STD-810G||Shock and vibration|
|MIL-STD-901D||High impact shock|