Performance That Stacks Up
siCluster is designed to enable you to scale up your capacity, storage and network performance, without scaling out of your budget. SiCluster configurations start at less than $1 per raw GB, while providing some of the fastest and most reliable storage capability in the market.
Real World, Real Tests
siCluster is put through its paces with the same configurations, software setup, tuning, and file sizes that your business will utilize. This presents you with the most accurate representation of the real-world performance you can expect to achieve running in your facility. We don't believe in reporting theoretical maximum "marketing" numbers you often find in this space—your numbers will match our benchmarks while running your applications.
Speed That Towers Above
During normal operations, as well as tests, siCluster consistently sustains speeds that outdistance competitive systems.
- siCluster-NAS: Up to 16[12] GB/s read[write] per rack
- siCluster-FDS: Up to 40[28.8] GB/s read[write] per rack
- Capacity from 10TB through hundreds of PB
* Bricking Not Included
siCluster features Scalable Informatics' JackRabbit™ and Delta-V™, which are built with long-term utilization in mind. We designed the systems with open source Linux and standard drivers so that you have the ability to support them in house, should you so desire. Gone is the possibility of these systems becoming "bricked"—made obsolete as their manufacturer discontinues support—that has happened in the HPC industry so recently with tier 1 manufacturers.
Scalable Informatics systems will continue to run and be supportable regardless of any circumstances external to your business. We are not out to entrap you with closed systems and keep you beholden to us. Instead, we want you to succeed, see the value of our systems, and come back for more as your business grows.
siCluster helps your business turn up the volume!
siCluster-NAS |
siCluster-FDS |
|
| Network connectivity per storage node | 2-6x 1GbE, 1-4x 10GbE, 0-4x DDR/QDR IB | 2-6x 1GbE, 0-4x 10GbE, 1-4x DDR/QDR IB |
| Bandwidth | Up to 8 GB/s read/write per rack | Up to 28.8 GB/s read/write per rack |
| IOPs |
2,500/storage node w/o SSD > 100k/storage node with SSD |
2,500/ HDD node 240k/ SSD node > 1M/ Flash node |
| RAID levels per storage node | RAID 6 | RAID 1, 5, 6, 10, 50, 60 |
| Storage nodes size and number per 41U rack | 4U: 1 – 10 5U: 1 - 8 |
4U: 1 – 10 5U: 1 - 8 |
| Storage Media |
Hard Disk Drive (HDD) Solid State Drive (SSD) |
Hard Disk Drive (HDD) Solid State Drive (SSD) Flash Disk |
| Maximum Raw Storage per Rack | 4U: up to 720 TB 5U: up to 1.1 PB |
4U: up to 720 TB 5U: up to 1.1 PB |
| Maximum storage capacity | 147.4 PB | 147.4 PB |
| Maximum power consumption per rack | 4U: 9 kW 5U: 10.4 kW |
4U: 9 kW 5U: 10.4 kW |
| File system and block targets supported | File: NFS, GlusterFS, CIFS, Ceph* Block: iSCSI, SRP, FCoE* |
File: Lustre, NFS, GlusterFS, CIFS, Ceph*, pNFS* Block: iSCSI, SRP, FCoE* |
| Client OS support | Linux, Windows, iOS | Linux, Windows, iOS |
| Standard Warranty | •3-year parts exchange | |
| Support Options | • Extended Warranty and next business day parts • Remote Support • NBD on site |
|
* When released
To take the best advantage of siCluster, we recommend the following file systems:
|
Gluster delivers an open storage software platform focused on simplifying the task of storing and managing the explosive growth of unstructured data. Our scale-out storage platform runs on commodity hardware scaling out in building block fashion from a few terabytes to multiple petabytes to meet the most demanding capacity and performance needs at a fraction of the cost of traditional systems. Gluster Storage Platform uses a unified global namespace to virtualize disk and memory resources into a single shared pool that is centrally managed. Our scalable NAS solutions address the needs of traditional data centers, shared storage for virtual server environments, and both public and private storage clouds. Gluster is used by enterprises in a wide range of industries, including digital media delivery, healthcare, internet, energy, and biotech. Gluster is privately-held and headquartered in Milpitas, California. More information can be found at www.gluster.com. |
|
For the world's largest and most complex computing environments, the Lustre™ file system redefines high performance, scaling to tens of thousands of nodes and petabytes of storage with groundbreaking I/O and metadata throughput. |
|
Our partners at Newservers.com and Sabalcore provide high performance computing on-demand, with Newservers.com creating dedicated cluster systems, and Sabalcore providing pre-constructed cluster systems.
Please contact us for more information.
All trademarks are properties of their respective owners.
