IT/Datacenter & Super Computing - Page 9

Supermicro announced the first server, a 2U 2-Node UltraTwin featuring hot-swappable 1U nodes, to be released in the new Ultra Architecture SuperServer series. The new server is built on dual Intel Xeon E7-2880 v2 processors that now support NMVe SSD's.

With the ability to support a high number of cores, large memory capacity, and high performance storage techonology, Supermicro is targeting this server for virtualization applications, datacenters, cloud computing, and HPC. According to Shannon Poulin, vice president of Enterprise IT for Intel's Data Center Group, these servers are built "to handle increasingly complex data intensive workloads in mission critical, highly virtualized environments."

And according to Charles Liang, President and CEO of Supermicro, there will be several more additions to this Intel based series - "UltraTwin is the first in a line of upcoming servers to take full advantage of this groundbreaking architecture and expands our MP solution range with support for Intel Xeon E7 v2 series processors and peripheral technologies."

Supermicro announced their entry to the software defined network market at Interop Las Vegas 2014 this week by demonstrating its new SDN SuperSwitch and MicroBlade microserver solutions. The new 1U SDN switch platform is based on the Intel Open Network Platform using Intel switches and processors combined with Intel Open Network Software (ONS). The MicroBlade also incorporates Intel switch modules featuring SDN functionality and an Intel Atom C2000 control plane processor.

Leveraging Intel's technology, the new Supermicro SDN solutions offer flexibility, agility, security and dynamic manageability. According to Rose Schooler, vice president of Intel's Data Center Group and general manger of the Communications and Storage Infrastructure Group, "The new Supermicro SuperSwitch, developed from the Intel ONP Switch Reference Design, delivers a high performance, easy to deploy and cost effective network switch that offers management and control functionality."

While there are most likely cost benefits for enterprises that deploy SDN, it is this improved flexibility in management of the network that makes this technology so appealing. According to Charles Liang, President and CEO of Supermicro, the "new SDN enabled SuperSwitch and MicroBlade switches provides Data Center, Cloud and Enterprise environments the greatest flexibility to dynamically allocate networked resources as data demands shift." With datacenter workloads constantly shifting and blending, this ability to dynamically allocate network resources to hungry workloads is paramount.

The so-called "Internet of Things" will disrupt the technology world, with 26 billion units supported by 2020, shaking up the data center, according to the Gartner research group.

IT services will help monetize the market further, with IoT product and service suppliers queuing up to join a quickly growing market.

"The enormous number of devices, coupled with the sheer volume, velocity and structure of IoT data, creates challenges, particularly in the areas of security, data, storage management, servers and the data center network, as real-time business processes are at stake," said Joe Skorupa, VP at Gartner, in a press statement. "Data center managers will need to deploy more forward-looking capacity management in these areas to be able to proactively meet the business priorities associated with IoT."

SolidFire, the Boulder, Colorado based manufacturer of all-flash storage systems has announced Q2 2014 availability of Carbon, Version 6 of its Element OS. Included in this upcoming release is a whole new set of enterprise class features such as real-time replication and integrated backup & restore.

SolidFire is providing this upgrade for free to existing customers including all the new enhancements. The new features included are:

One of the greatest benefits for Toshiba when they purchased OCZ was their access to PCIe technology. Toshiba has a strong stable of enterprise SSDs, but they did lack the PCIe component. The release of the new OCZ Storage Solutions Z-Drive 4500 is a good example of the benefits of the acquisition in action. The access to Toshiba flash will also allow for more competitive pricing and a solid product supply for OCZ.

The Z-Drive 4500 features 8 SF-2582 processors in conjunction with Toshiba 19nm MLC NAND and the VCA (Virtualized Controller Architecture) technology to deliver some impressive specifications. The drive delivers maximum read bandwidth of up to 2,900 MB/s, maximum write bandwidth of up to 2,200 MB/s, random 4K block read throughput of up to 252,000 input/output operations per second (IOPS) and random 4K block write throughput of up to 76,000 IOPS.

Capacities of 800GB, 1.6TB and 3.2TB deliver endurance ratings of 680, 1300, and 2500 TBW, respectively. The drive features power capacitors for data protection and a new full height-half length design. A new casing is designed to cool components, and the addition of thermal throttling protects the SSD from overheating. Support for VXL Virtualization Software is baked in, and this is all wrapped up with a five-year warranty.

In a series of news releases this week, IBM is sending a message to the market that it is serious about being a major player in cloud services. Building upon its acquisition of Softlayer last year, IBM announced that it has invested over $1B in bringing a suite of development tools to the cloud to deliver a PaaS (platform as a service). In addition to this investment, they also announced the acquisition of Cloudant, a DBaaS (batabase as a service) provider.

These new services are geared towards application developers and providing them on-demand infrastructure that can be scaled up or down as requirements change. As part of the announcement, they have launched a beta version of their DevOps in the cloud solution codenamed BlueMix. BlueMix includes services to store and manage code, a built-in Web integrated development environment, and Eclipse and Visual Studio integrations to let developers use whichever environment they prefer. The platform enables agile planning, tracking, and collaboration across team members during the development cycle. Plus, there is application deployment automation to streamline delivery. Combined, the PaaS and DBaaS enable application DevOps teams to start projects faster and complete them more efficiently.

IBM is being very proactive in changing its model to address the reality facing them. Traditionally, developers would make a change request to have resources allocated to them which in turn would drive the storage or server team to purchase more equipment from IBM. With the proliferation of cloud services, developers can now simply go grab resources on demand from any of the cloud providers; thus, speeding up their time to get the resources and reducing the need for more IBM hardware. Realizing that the model has shifted, IBM is now empowering development teams to not just get resources, but is adding value around those resources to make their solution much more compelling than run of the mill cloud service providers.

The move to 12Gb/s SAS is beginning in the datacenter, and as per usual the bleeding-edge devices lead the way. 12Gb/s SAS3 SSDs were first, with three industry heavyweights leading the 12Gb/s charge. We have evaluated the HGST SSD800MM, the Toshiba PX02SMF080, and the Seagate 1200. Surprisingly, a few of these SSDs actually beat compatible RAID adapters and HBA's to market, but we are finally seeing the right pieces of the architecture all fall into place.

Of course the next logical step is to begin applying the benefits of 12Gb/s technology to the platter realm. The fastest HDDs are the first to get the 12Gb/s treatment, and today HGST beat the competition with the announcement of their new C15K600 family.

This 15K RPM HDD also brings a few new features. HGST is utilizing their media caching technology to provide significantly enhanced write performance over solutions with limited NAND or flash-based non-volatile cache (NVC). The result is two times better random write performance than current generation 2.5-inch 15K drives, and two and a half times better performance than 3.5-inch 15K drives. We are reaching out for details from HGST on this new technology, which apparently leverages a much larger 128MB cache buffer to achieve this spectacular performance.

Nexenta has quietly accumulated a significant customer base of more than 5,000 users and nearly one exabyte of storage installed. Continuing upon that success, Nexenta announced today the upcoming release of NexentaStor 4.0 in Q2 2014 that will bring enhancements to failover time and overall reduced latencies. In addition to improved performance, Nexenta has put forth an effort into improving its out of box experience by simplifying installation and management with wizards and advanced monitoring tools.

While solutions around "software defined" networks, storage, and data centers are relatively nascent, Nexenta is staking its claim early on by partnering with industry leading system integrators. As an existing premier reseller partner for Nexenta, Eva Cherry, CEO and President of Silicon Mechanics says,

San Diego based Astute Networks released an update to their ViSX family of performance storage appliances, the ViSX G4 OS version 5. This latest release is an all solid state flash performance appliance that Astute claims is the industry's lowest cost-per-GB with the highest IOPS per dollar solution in the market. It is available in capacities up to 45.6TB in a 2U platform, with add-on flash modules available in sizes up to 1900GB.

Astute adds to their iSCSI offering the additional support for NFS (v2, v3, and v4). Designed for both virtual and physical infrastructure, ViSX is compatible with 1GbE or 10GbE infrastructure and does not require a forklift install. According to Deni Connor, founder and senior analyst, SSG-NOW. "Astute's Networked Flash appliance removes all storage and network I/O constraints to dramatically accelerate virtual machine and application performance - while co-existing with an organizations existing infrastructure."

The Astute Network appliances aim to speed performance for intense enterprise application workloads. Astute claims 5 to 10 times performance boost for enterprise applications such as VDI, SQL Server, MySQl, Oracle, and Sharepoint. With their latest release of ViSX, they also provide solutions for cloud service providers with support for OpenStack, the open source cloud computing platform for public and private clouds.

Dell recently announced availability of the PowerEdge R920 server that utilizes Intel's new XEON processor E7-4800 v2 and it is already making a name for itself. Dell claims it is their highest performing server and that it is built for the most intense enterprise workloads - ERP, large databases, HPC, etc.

Backing those claims up, Dell just released some pretty impressive benchmarks, taking the lead in a few categories seen here. Posted on Dell's tech forum is this snapshot demonstrating nearly double the performance of previous generation Power Edge servers in a SAP benchmark test:

And from Intel's benchmark tracking site, they now hold several world records in the 4 socket category:

Sunnyvale, California based Exablox announced that they have reached an installation base that now exceeds 1PB. Touting themselves as "the company reimagining storage", Exablox entered the market mid last year with their OneBlox Appliance. They attribute their quick market adoption partly due to their go to market strategy of allowing customers to choose their own drives, as opposed to having to buy drives from them. In a related press release, they also just announced the capability of supporting the new HGST 6TB drives so their cost per capacity is looking to be even more attractive in 2014.

Allowing customers to choose their own drives is not unique to Exablox, as other storage vendors in the space have been doing this for years. However, most of the vendors allowing this flexibility do not also provide the same level of enterprise features that the OneBlox Appliance is capable of - features like inline deduplication, continuous data protection, and scaling out on-the-fly. This combination of cost savings and enterprise class software features could certainly make them a force to be reckoned with, especially in the small-medium enterprise space.

Furthermore, looking through their datasheet, it appears they have locked on to another major factor for this space- ease of use. With a supposed 3 minute installation and a cloud based management software, it could certainly appeal to IT shops that need to have a couple hundred TB's, but don't want to babysit it. Not to mention, it definitely looks sleek.

In a report commissioned by Emulex Corporation, the research team at Forrester recently surveyed 158 IT professionals regarding managing business-critical applications and the infrastructure surrounding them. Of the respondents, nearly 90% are in a management role or above and consider themselves infrastructure professionals.

Of these respondents, 56% said that over 25% of their critical performance issues cannot be solved within 24 hours. Further, that the most difficult problems can take well over a month to fix, with the majority of the time being wasted on just figuring out what is going on.

In the report, Forrester extracts 3 key findings from their research:

The Open Compute Project (OCP) is an initiative started by Facebook to build the most efficient computing infrastructures at the lowest possible cost. OCP focuses on total control of the architecture, from custom-designed server components to software and operating systems. This has led to massive disruption in the enterprise space and the upending of the old model, which relied upon massive systems provided by a group of OEMs that were prohibitively expensive. The Open Compute Project relies heavily upon the open source model of information sharing and has drawn in the other titans of the computing world. Apple, Google, and Microsoft are among the 150 members, along with a bevy of hardware suppliers, that contribute to the OCP and their ideals.

The Open Compute Summit is the yearly meeting of the leaders of the movement, and a great forum to showcase the ideas and resulting hardware driving the movement forward.

This radical new way of thinking has saved tremendous amounts of both energy and money, with Facebook disclosing that they saved 1.2 billion dollars over the last three years, and also saved the equivalent of enough energy to power 40,000 homes and the emissions equivalent of 50,000 cars. Facebook's leading challenge revolves around how to store relatively cold data (data that is written once and seldom read) for long periods of time. Facebook intends to keep those pictures you shot five years ago forever, and storing all that data in a cost effective manner is their most pressing issue. Enter the 42U Blu-Ray storage cabinet. This robotic cabinet holds 10,000 triple-layer 100GB discs storing up to a petabyte of information, and future implementations will store an amazing 5PB per cabinet.

We finally got our first look at what HGST's 6TB, helium filled HDD will cost on the open market. If you want one, a seller on Amazon has nine listed but the price is a bit inflated from a consumer perspective, $798. The He6 model has a handful of unique features. It's filled with helium and has 6TB of capacity. Paul wrote an editorial about the drive and went over all of the details last November.

What's enterprise today will be consumer tomorrow...or maybe a year from now. It's nice to see HGST moving forward with this technology.

Nimbus Data, a leading provider of unified all-flash storage systems for enterprise and cloud infrastructure, has announced record-setting financial results for 2013. Nimbus Data specializes in unified all-flash array technology, and their Sustainable Storage systems have surpassed 500 deployments. Their products bring together low-latency flash, data management and protection, and highly-scalable multiprotocol storage features to create a fault-tolerant solid state solution designed for server and desktop virtualization, databases, HPC, and next generation cloud infrastructure.

Nimbus all-flash arrays are available for a acquisition cost that is comparable to disk-based storage arrays with tremendously lower operating costs and total TCO.

Things are moving along quickly for Nimbus Data, with net income growth 4x that of 2012 and doubled revenue as well. They have also expanded their facilities and staff with a new sales and engineering office in Austin, TX, and headcount growth of 125%. These solid financials and continued controlled growth may foretell a possible IPO in 12-18 months.

Diablo and Sandisk received Storage Visions 2014 Visionary Product Award, in the Enabling Professional Storage Technology category, for the ULLtraDIMM. The ULLtraDIMM adds NAND onto a DIMM-type module, leveraging Diablo IP for translation and Sandisk's Guardian Technology for NAND management.

The Diablo MCS (Memory Channel System) technology creates the pathway that allows NAND to be used as either in-memory compute space or block-level storage. This allows for two different usage models that present unprecedented speed and density from flash storage. We were lucky enough to speak with the CEO and co-founder of Diablo, Riccardo Badalone, at this year's Storage Visions conference.

This new technology is very disruptive, and Sandisk and Diablo aren't the only companies eyeing this type of product. They currently are in the lead, with working silicon and hardware already in the validation stages. We expect this technology to be a game-changer of enormous magnitude, and will help to enable the emergence of the all-flash datacenter. We spoke with Jon Scaramuzzo, the SVP and GM of the Enterprise Storage Solutions group at Sandisk, in this interview where he outlines the penetration of flash into every storage tier of the datacenter.

K, one of the world's fastest supercomputers based in Japan, is capable of 8.162 petaflops of performance, thanks to its insane 82,944 processors. The supercomputer is capable of driving 1016 billion operations per second, but even then, it is still hard pressed to compete with the brain in your head reading this article.

It took K around 40 minutes to simulate just 1 single second of human brain activity, even with all of its performance prowess. The experiment on simulated human brain activity involved 1.73 billion virtual nerve cells that were connected to 10.4 trillion virtual synapses, with every virtual synapse containing 24 bytes of memory.

NEST was used on the software side of things, which is a simulator for spiking neural network models that focuses on dynamics, size and structures of neural systems, versus exact morphology of individual neurons.

Just months ago the US government was shut down, with hundreds of thousands of jobs in the air, millions of US citizens affected, but that's nothing when it comes to the blank cheques it signs to the National Security Agency for "research".

The US spy agency is reportedly working on a quantum computer that would break through any encryption thanks to its pure, insane amount of processing power. Edward Snowden is behind the leaks - come on, you're not surprised now, are you - revealing a program that is worth some $79.7 million, dubbed "Penetrating Hard Targets".

The Washington Post is reporting the news, stating that the majority of the research is being done at the University of Maryland's Laboratory for Physical Sciences.

DataON Storage, the leading provider of OS-agnostic Storage enclosures, has announced that their cluster-in-a-box and JBOD enclosures have recvieved certification for Microsoft Windows Server 2012 R2 Storage Spaces. The enhanced functionality of Storage Spaces, in concert with the correct hardware, provides auto-tiering and hybrid cloud support.

One of the major attractions of Storage Spaces is the easy-to-use failover capabilities. Providing a high level of redundancy and failover support for entire storage systems was an incredibly taxing and complex endeavor from an IT perspective even a few short years ago. This adds to the TCO of the solution due to prohibitive expenditures for initial setup and long-term management.

The emergence of Cluster-in-a-Box(CiB)solutions has brought this type of functionality withing striking distance of smaller organizations, and also delivers a lower TCO for larger organizations. While the long-term penetration of the Storage Spaces platform into the datacenter market will spawn a whole ecosystem of products, for now there is only one certified CiB solution available for Widows Server 2012 R2.

Avago, a leading designer and developer of compound semiconductors, has announced their intention to purchase the smaller LSI for $6.6 Billion. Avago, with its market cap of 11.3 Billion, is much larger than LSI, with a cap of 4.32 Billion.

Stocks of both companies are skyrocketing in early trading, with LSI up 38% in the first 6 minutes of trading, and Avago is up 11.7% and climbing.

Avago turned to Silver Lake for partial funding for the venture. Silver Lake, a technology investor, is a common player in the storage space. Silver Lake actually owned, and sold, Avago last year. The Avago deal was the largest payoff for Silver Lake, they received a 5X return on their investment, and have a seat on Avago's board.