News

3Leaf's Dynamic Data Center (DDC) is a rack of servers, storage and networking that does something unique—it allows administrators to treat the entire rack of systems as one computer running a single OS over multiple AMD Opteron servers. The location of underlying hardware no longer matters. Now that is virtualization. 3Leaf is positioning the DDC to sell to OEM vendors and limited fixed configurations to end-users. Systems will be available in December, 2009.

At a high level, 3Leaf's DDC product line virtualizes and aggregates computing hardware across multiple platforms. A fully racked DDC rack operates as a single computer with 192 cores and 1TB of RAM. To provision a server, you select the number of cores, the amount of RAM and the required I/O. For example, an application that requires more memory than is available in a single server can access RAM  on other hardware.

Consider that a DDC-Server configured with 1TB of shared memory, 192 cores of AMD processors at 2.8 GHz and 8TB of storage, all connected via an InfiniBand switch and complete with cables, Linux OS and DDC-Pool software lists for $250,000.  A DDC-Server with 256GB of shared memory, 96 cores of AMD Istanbul processors at 2.4 GHz, 4TB of storage with an InfiniBand switch, cables, Linux OS and DDC-Pool software is listed at $99,000.  

Three technologies come together to make this happen. Modern processors support nested page tables, which are a hardware memory management technology in the CPU that allows multiple virtual hosts to have the same view of RAM. 3Leaf leverages nested page tables and their custom ASIC to map memory across multiple computers. The servers are interconnected with a high throughput, low latency (approximately 100 nano seconds) Infiniband switch. Finally, native OS support for ACPI dynamically manages hardware resources.

Representatives from 3Leaf say that a similarly sized system from the likes of HP or IBM might run into the millions, but 3Leaf pricing starts at $99K because they use off the shelf parts for the hardware compared to specialized HPC equipment. 

At the heart of DDC is a custom designed ASIC and 3Leaf software that is installed on the server hardware called a node. The ASIC runs at Ring -1 that 3Leaf characterizes as below the OS kernel. The firmware resides on a hidden disk partition, and ASIC joins the nodes together and manages remapping the memory address for the OS. When a server is booted up, the ASIC runs after the BiOS initialization but before the OS loads and connects with the other nodes in the rack, and it provides the coherent memory mapping required to share memory among servers. For example, similarly configured servers will have the same memory layout on all the machines, so to avoid conflict, the ASIC maps memory addresses to actual locations either on the local server or on a node. Once running, the OS requires no special drivers and simply treats the resources as if they are local.

3Leaf Vice President Shahin Kahn describes the software as "more like the service processor software that has been typical in large Unix systems. As we go from DDC-Pool to DDC-Share, it becomes more involved and when we get to DDC-Flex it starts resembling a hypervisor. For now, we see it as a service processor kind of software." 3Leaf DDC currently support AMD CPUs, but support for Intel is coming in 2010.

Resources on DDC can be allocated in multiple ways:

• DDC-Pool is the most straightforward. Server resources like CPU cores and RAM are pooled across multiple nodes and a single OS runs on the pool. Any modifications to the pool such as adding or removing CPU cores or RAM requires rebooting the OS.
• DDC-Share allows server resources to be split up across nodes. For example, one node provides the CPU course while one or more nodes supplies RAM. Interestingly, in this mode, if other nodes are simply supplying RAM, they just need to be turned on. The ASIC and firmware provides access to the node's RAM.  Like DDC-Pool, any modifications to the resources requires an OS reboot.
• DDC-Flex is the most flexible and allows dynamic re-provisioning of hardware resources such as changing the number of CPU cores or the amount of RAM that does not require an OS reboot. DDC-Flex uses ACPI to coordinate the changes with the running OS. Currently, 3Leaf is supporting RedHat and SUSE Linux, but Flex won't be available until 2010 and neither will Windows support.

 
DDC is the next generation of virtualization and DDC-Flex, if it works as advertised, removes a huge barrier of scaling hardware, allowing administrators to dynamically manage application server performance. Imagine being able to dynamically add and remove CPU cores or RAM as needed. If the DDC systems prove stable and reliable, 3Leaf will turn application deployment and management on its head.
 

During Tuesday's keynote speech at VoiceCon San Francisco 2009, Avaya's Dr. Alan Baratz said that in order for enterprises to plan for their next generation of business communications infrastructure, businesses must first analyze which types of communication tools are most effective for achieving their goals.

Baratz, Avaya's senior VP of global communications solutions, divided the needs of business communications into two categories: those needing rich, lasting engagements such as conference calls, and short, immediate exchanges that can be executed via Twitter, SMS, or an instant message. A business' communication infrastructure must be able to handle the two types in a well-integrated fashion in order to deal with what Baratz calls Voice 2.0. Part of the challenge will be taking the familiar control one has over e-mail and applying it to other communications vehicles.

"In the near future, communications will easily convert voice to text across platforms and devices, so what works best for the sender also works best for the recipient," Baratz said. "Users will have more control and be able to include attachments to real-time conversations which can subsequently be archived and searched by topic."

Avaya released its Aura infrastructure about five months ago, and the company is planning the second version of its SIP-based platform later this month. The latest version of Aura will come with more features that can deliver a quicker return on investment, Baratz said. Those features may include centralized dialing, trunking lines, and enabling users to register and auto-register devices. Avaya specifically designed the Aura platform to make it easier for an enterprise to connect and manage its telecom equipment centrally, he said.

"It's not about everything, it's about the right set with the right controls," said Baratz.

Looking forward, Baratz said the company is working on reinventing the telephone itself by producing a chameleon-like device that can take on the properties of a desktop PC, mobile phone, video conference equipment, desktop phone, and other communications equipment. Avaya will reveal more details on this hardware in the next six to nine months.

"It's hardware without personality purely under software control," Baratz says. "You can turn it into a variety of endpoint devices."

Baratz said that Google (NSDQ: GOOG), Yahoo (NSDQ: YHOO), and Skype have done a better job integrating voice, video, and IM than many enterprise-facing companies, and Avaya will work to develop an alliance with a well-known brand name IM company to develop an enterprise-grade IM product. He did not elaborate on when this product is expected to be announced. 

Today AMD is introducing a revision of its flagship Phenom II X4 965 processor rated at 125W, replacing the 140W part, as well as a new 3.1 version of its Overdrive overclocking software. We take a quick look at both to see what advantages they offer.

Networking vendor Enterasys has announced an upgrade program for wireless customers of Extreme Networks'  WLAN products. Free until February 2010, the program offers software upgrades of certain models of Extreme's wireless access points and controllers, as well as honoring up to one year of existing Extreme Networks wireless maintenance and support agreements. The software upgrade essentially transforms existing Extreme Summit controllers and Altitude access points into the HiPath equivalent products from Enterasys. 
Competitive trade-in programs are nothing new among networking vendors, but these programs usually entail physically replacing one set of products for discounted price on another. Enterasys's offer is a bit unique in that Extreme WLAN customers will become Enterasys customers while retaining all of their existing hardware. This is possible because of Extreme's prior OEM relationship with Siemens, which delivered HiPath WLAN gear under the Extreme Network's brand. This relationship was strained with the merger of Siemens and Enterasys in October 2008, effectively putting Extreme Networks in the position of selling WLAN for one of its switching competitors.
In recent weeks, there has actually been a flurry of WLAN OEM changes among the switching vendors. The Enterprise Mobility group at Motorola has not only become the OEM WLAN for Extreme Networks, but has also displaced Meru Networks to become Brocade's wireless partner.  While these wins are certainly good for Motorola, these shifts inevitably are a concern for customers who have invested in the prior technology. 
The lesson for enterprise customers is to do their homework when purchasing wireless networking products from their wired networking vendor. The switching market is littered with OEM relationships for WLAN products, many of which are tenuous at best. Combined with the stream of current acquisition activity in the WLAN market, partners can quickly become competitors, and many times customers are left having only token support for older products and endless sales pitches to move to the new product. While there are advantages for single source networking vendors for wired and wireless products, namely a single support team to deal with, customers need to know exactly where their products are coming from and understand the nature of any OEM relationships that may be in play.

Riverbed Technology has announced Steelhead Mobile 3.0, an update to its network optimization product aimed at mobile users of enterprise networks. The new version is intended to more closely work with the recently-released RIOS 6.0 to increase network user productivity while on the road, working from home or connecting wirelessly in the office. Steelhead Mobile 3.0 introduces new optimization of Windows and other Microsoft applications, such as SharePoint, Office, Server and CRM. With the new release, Riverbed provides acceleration benefits to Windows 7 and 64-bit systems. Steelhead Mobile 3.0 will be generally available on December 2, 2009. Pricing was not available as of press time.

With Mobile 3.0, as with the Steelhead appliances, WAN acceleration extends further than in previous versions into the application layer for HTTP and HTTPS users, making Web applications even faster. The common architecture of the Steelhead appliance and Steelhead Mobile gives organizations a single solution for optimizing performance of Web-based applications.

Among the technologies employed by Mobile 3.0 are URL learning, page parsing, embedded object pre-fetching and metadata acceleration modes. In addition, the new release features Branch Warming, which allows mobile and branch office users to share optimized data and experience greater overall acceleration. By sharing the data references between the data stores of the Steelhead Mobile client and the branch office Steelhead appliance, mobile workers not only take advantage of all of the optimization benefits of the Steelhead appliance but are also able to contribute data references from their data store to help improve performance for the entire branch office, enabling "warm" performance regardless of location.

Michael Vassallo is senior network administrator for Dancker Sellew & Douglas, a furniture resource and asset management company with headquarters in Somerville, NJ. He says that his firm has been using Steelhead Mobile since "1-point-something" and jumped at the chance to get into the Mobile 3 alpha and beta programs. He says, "I had moved to a 64-bit workstation, so I was without a client until they introduced the alpha. [The new version] also has a lot of features that coexist with RIOS 6, including things like the branch warming and location awareness."

Vassallo is impressed that Mobile 3 works in tight integration with the Steelhead appliance and manages the licenses based on client performance over a WAN link, rather than simple use of the mobile WAN. "We have people in the Syracuse office; they have a Steelhead [appliance], and it looks at the data they've looked at on their mobile client, with information on the data requested passed back and forth so it doesn't have to be refreshed again and again."

A key advantage to the entire Steelhead implementation, according to Vassallo, is a reduction in what he calls "WAN sprawl." He says, "Within the first year of having [the Riverbed deployment] we were able to get rid of five or six T1s. The VoIP talks between switches in the branch offices, and now we can go over a single pipe instead of needing lots of additional bandwidth."

By expanding the WAN optimization to mobile users, Vassallo says that the IT department has increased the level of user satisfaction with applications. "The users are definitely happier having Mobile 3 in use. They don't necessarily notice it all the time when it's working, but when it's not working they definitely notice it. If, for example, the laptop comes up wrong and it's not optimizing, they certainly notice the drop in performance," he says.

WAN optimization is increasingly important as companies look to push more individuals out into working on customer sites and away from central offices. Network performance plays a key role in essential worker productivity and is strongly coupled, as Vassallo indicated, in employee satisfaction, as well. Products such as Mobile 3.0, that expand the notion of WAN optimization beyond cabled backbone networks to wireless final-step links, will be increasingly important as mobile workers and mobile platforms each proliferate.

Cisco and EMC expanded their partnership Tuesday, tightening their data center and cloud computing links as they move to capture more IT infrastructure. The new offering, called Virtual Computing Environment (VCE), is glued together by EMC's VMware even as it draws on the traditional networking strengths of Cisco and the storage power of EMC.

The ambitious effort is tailored to appeal to customers with as many as 6,000 virtual machines and as few as 800 virtual machines. Cisco and EMC have been working aggressively on expanding their tentacles beyond their traditional markets ever since EMC welcomed Cisco as in investor in its VMware company in the summer of 2007.

"This coalition is about more than technology and partnership," said Cisco CEO John Chambers in a statement, as he pointed to new data center opportunities. "It is about an entirely new and unique approach to the data center that improves utilization, power consumption, and security of information, all in a way that lowers the total cost to the customer, not via a box, but with a network-based architectural approach for optimizing virtual resources."

Initial customer trials of VCE's basic building pieces called Vblock Infrastructure Packages have resulted in cost reductions up to 40% for operating and managing virtualized data center infrastructures, Cisco and EMC said. The approach builds IT strategies around private clouds with Vblock Infrastructure Packages that contain a myriad of features that enable a global community of systems integrators, service providers, channel partners, and independent software vendors to participate in the deployment of VCE .

In addition, other major IT players are participating in the rollout of VCE including Intel -- another early investor in VMware -- whose Xeon processors and other Intel data center technology are involved. Intel is a major player in a new joint venture with Cisco, EMC, and VMware called Acadia, in delivering the Vblock architecture to customers. Acadia, designed to quickly transfer Vblock infrastructure to organizations eager to accelerate moves to virtualization and the private cloud, is expected to begin serving customers in early 2010. EMC introduced its Ionix Unified Infrastructure Manager to Vblock. EMC RSA security products have also been developed for use with Vblock packages.

"(Customers) need to be able to shift more of their IT budgets to the development and rapid implementation of new technologies that help their organizations create differentiated business advantages," said Joseph Tucci, EMC chairman and CEO, in a statement. "Many of them understand the vast potential of the private cloud. With shared roadmaps and a long-term commitment, the Virtual Computing Environment coalition will bring true accountability."

The announcement will draw a sharper line between Cisco and EMC on one hand and other IT providers like HP and IBM. Cisco and EMC have been aggressively acquiring companies in recent months with EMC staying closer to its storage roots while Cisco has been ranging far and wide, picking up companies in consumer electronics to video conferencing to networking. EMC has been parlaying its VMware acquisition across wide swathes of computing.

"Customers are increasingly looking to virtualization to dramatically improve the performance and flexibility of their exiting IT systems," said VMware president and CEO Paul Maritz. "Today's announcement provides a compelling vision and set of roadmaps valuable to any company looking to harness cloud computing in a fundamentally more pragmatic and nondisruptive way."

Attend a Webcast on the application grid approach to modern data centers. It happens Tuesday, Nov. 3, 2009. Find out more and register.

Last month, we discovered that many budget P55-based motherboards simply weren't designed to stand up to the rigors of overclocking--bad news for intrepid value-based enthusiasts. Today we revisit the topic with revised boards and a Core i7-870.

CA has announced a broad series of new and updated products and integrations aimed at the enterprise security and risk management markets. CA Access Control 12.5 includes features for privileged user management and host access control as well as improved integration of CA Identity Manager 12.5 and CA Role & Compliance Manager 12.5, including Smart Provisioning capabilities. Also included in the design is CA DLP 12.0 with extended discovery, protection, and control of sensitive data, and new integrations to complementary solutions. CA Records Manager 12.6 has features for governance, content protection, and compliance, and CA Governance, Risk & Compliance Manager 2.5 (CA GRC Manager) offers new risk and compliance features.

According to Gigo Mathew, CA vice president of security product marketing, integration of multiple functions is a central thrust of all the updates and product releases. "Being able to ultimately do more things proactively rather than as a simple reactive process is important. For example, provisioning is a huge part of this," he says, explaining, "Smart Provisioning Integration checks separation of duties and checks permission patterns to make identity lifecycle management a lot smarter. It then brings this together with identity management. No one has linked identity management and compliance before this."

Mathew goes on to say, "We look at three layers of management and governance. The top level is GRC Manager, governance of risk and compliance, and gaining visibility into all the types of compliance issues and risks to the organization, then mapping policies to those requirements. Smart Provisioning is next level. Active Control 12.5 is the third level, giving control around privileged users." Mathew points out that many security functions hinge around this privileged user, typically a security admin or DBA. They tend to have the keys to the kingdom, and the new products are designed to give a level of visibility and control over those users. This can prevent things like the San Francisco incident where a rogue administrator can lock people out of systems.

Joe Ford is VP of professional services and CTO of Patriot Technologies, a system integrator who has been a CA reseller for two years. He says that Patriot recommends CA products but also uses them internally, though many of the CA solutions are intended for organizations considerably larger than Patriot Technologies' 65-employee head count. Ford says that the granularity of the new CA solution is an important part of its appeal to organizations. "CAs DLP product extends the traditional thought of data leak to a more granular level of identifying with identity management to let you create rules based on your data and place it in a matrix based on the user roles hierarchy. Traditional DLP might not let you email social security numbers outside the company, but the CA tool would let you allow someone from HR to email those to the payroll company," he says. Ford points out that the data protection functions in conjunction with user role based controls, amount to capabilities that he hasn't found in other, competing products.

Ford points out that the truly difficult part of deploying any data leakage prevention system is properly classifying data that organization wants to protect. Much as the hardest part of most ERP installations is considering and formalizing business rules, understanding and formally classifying data types is the most involved part of a DLP installation. Companies tend to overlook this rather tedious classification process, unless they have an external reason, such as dealing with classified information, until they are faced with something like a DLP deployment. Then they bundle the cost into the DLP system while complaining about the TCO of security. In fact, understanding the company's data and its importance is a core IT function that should be part of every organization's charter from day one. Waiting until a DLP system is deployed is begging for a significant data breach.

Sony's bold experiment--a high-end luxury system with essentially Netbook components--makes an enviable coffee shop conversation piece, but one that includes some serious sticker shock.

We're looking at two new notebook hard drives today: Fujitsu’s 500GB 2.5" disk and WD’s brand new 640GB 2.5" model (the largest notebook drive you can buy). The capacities on both products are outstanding, but neither drive is completely perfect.

The fight for the data center continues with Juniper's Space, a centralized management framework which consists of a management server and a development program with APIs and an SDK to integrate applications with Juniper's Junos operating system. One the one hand, Space is just a management and orchestration platform with integration hooks much like HP's Insight. What makes Space unique and interesting is the developer program and the potential of opening the API and SDK to developers and enterprises alike.

Space is delivered as a virtual machine appliance or a physical appliance starting at $15,000. Space comes with a sixty day free trial so you can try before you buy. Depending on the size of the Juniper deployment, the more points, the more expensive the product. Installations at less than one thousand points will remain free.

Space is open, meaning it uses standard protocols like SNMP, Netconf, SSH, and SOAP to configure devices. Vice President and General Manager of the Junos Space Business Unit, Mike Harding, said in an interview that Space can be used to manage any vendor's device that uses standard management protocols or an XML interface, though the latter may require custom development.
--PlaceArchetypeHere--
The Space developer program is currently by invitation. Harding said they plan to open the developer program to anyone by mid 2010, which includes the ability to download the SDK and Space virtual machine. Developing for Space is one thing, but without a Juniper device running Junos, building and testing an application would be impossible. Harding wouldn't commit to making a Junos simulation environment available, but we bet there is one in the works. Juniper is looking at a distribution model similar to Apple's iPhone app store where applications submitted to Juniper are tested to make sure they are not malicious and adhere to a coding standard.

Store applications will receive first level support from Juniper, but you will need to rely on the developer to troubleshoot and repair any problems. Applications from partners that have a development partnership with Juniper will receive full support from Juniper. Harding said the split model allows Juniper to provide some control and assurance over the applications they distribute, while allowing customers to develop their own. This model is similar to the one that Cisco, Extreme and 3Com use.

Space will ship with three applications today. Ethernet Activator is a template-based network configuration tool that is simple enough that even a CEO like Juniper's Kevin Johnson can use it, jokes Harding. Network templates are defined so that define network characteristics such as MPLS tags locations, or any other relevant options can be predefined. Building a network between two sites is a simple matter of selecting the objects to be joined and selecting how they should be joined. In their demo, Johnson selected two data center networks and a VPLS network, and committed the configuration out to the hardware devices. In less than 18 seconds, Juniper's target deployment time, the network configuration was complete.

Of course, someone has to define the network objects and customize the templates, but that can be left to senior network engineers while lower lever IT staff can connect the dots.  Building network templates is not new--HP and Linesider, for example, provide similar provisioning capabilities, but Ethernet Activator demonstrates the types of applications that can be built.

Juniper Partner Packet Design developed Route Analyzer which is used to discover, model and analyze routing topology so that network engineers can troubleshoot routing problems. Engineers can also use Route Analyzer to simulate changes to routing to see the effect on traffic paths. Aimed at service providers and large enterprises, Route Analyzer demonstrates useful integration between a route analysis tool and the ability to deploy changes.

Finally, Service Now builds on Juniper's technical support, allowing space to gather up configurations and other diagnostic data and send it to Juniper engineers so that when you call into support, they have automatically have the data they need to begin troubleshooting. You can control what gets sent, so you use Service Now within the parameters of your own security policy, but anyone who has worked with technical support knows the time it takes to gather the data technical support engineers often ask for.

The Averatec D1130's wimpy processor and paltry features give consumers little incentive to choose this Nettop over a faster budget PC or the mobility of a Netbook. If you're dead-set on a Nettop, the Lenovo C300 is a less terrible decision, but you can stretch your dollar much further with a budget slim tower system like the Gateway SX2800-01.

DVD burners are extremely cheap. Obviously, it makes good sense to buy a $30 drive if you don't have one yet, but is it worth replacing an existing drive with a newer one? We compared three product generations from Sony/NEC between 16x and 24x speeds.

WLAN vendor Aruba Networks has announced AirWave on Demand (AoD), a Software as a Service (SaaS) edition of their multi-vendor WLAN management tool.  AoD features the same WLAN management, identification, monitoring and security functionality of the server-based AirWave Wireless Management Suite, but hosted at Aruba's data center. Subscriptions will be available on a monthly or annual basis, and there will be opportunities for resellers to remarket the service. The service is in early customer deployment phase and will be generally available starting in December 2009.
Aruba's AirWave on Demand joins a number of WLAN vendors, including Meraki and most recently Aerohive, in offering a hosted management solution.  Since its acquisition by Aruba in January 2008, the Airwave Management Suite has not only become the de facto WLAN management solution for Aruba's own wireless gear, but has maintained its legacy of vendor neutral support, supporting 15 different vendors including Cisco, HP and Motorola. This history of support is carried over to the online solution, making it a true enterprise WLAN solution.

This new SaaS offering from Aruba opens the AirWave product to a number of new markets that the in-house server offering couldn't address.  Like its competitors, AirWave on Demand might be an attractive option for smaller and mid-sized enterprises that are not interested in the additional capital expense of servers to house the management software.  Additionally, geographically disparate enterprises could also favor an online solution, centralizing management at Aruba's data center instead of funneling all of the management traffic back to their own.

There is another opportunity for AirWave on Demand that its competitors cannot even begin to address. Many enterprises are dealing with heterogeneous wireless networks, and the management of multiple wireless networks built from the gear of various vendors can be a nightmare.  For these type of customers, AirWave's multi-vendor support has always been a compelling option, and an online option could serve as a stopgap measure until the enterprise can move to a common WLAN platform.  Ultimately, it is up to the enterprise customer to decide whether its WLAN management is best served in-house or in the cloud.

Juniper is making a number of announcements today at an analyst's day in New York. Most important is Junos Space, a new software integration program that provides integration capabilities with Junos operating system. Then there is Junos Pulse, an integrated software client, a new chipset for the MX routers for service providers and large enterprises, as well as a cloud computing initiative to deliver a host of services such as scaling, application visibility, and threat protection.

Juniper's announcements are an important stake in the  ground. There are a number of important partnerships, such as IBM's and Dells OEM deals, and a new software licensing partnership with Blade Network Technologies to build Junos blade switches for chassis, driving right at the heart of rivals like Cisco and HP. Of course, missing form Juniper's product plans is storage networking, but it seems that Juniper is focused on delivering networking services and integrating with computing equipment vendors. The more important announcement today is Space.

Junos Space is an integration SDK and API that allows third parties to tightly integrate with Juniper's products running the Junos operating system. This allows companies to create software and services that respond quickly to changes in the network and opens up opportunities for Juniper partners to differentiate their product sets from competitors. Unlike other programs, like Cisco's AXP and 3Com's OSN (both requiring daughter cards inserted into their respective multi purpose routers), Space applications can integrate across the Junos product line.

Junos Pulse seems like a rebranding of their Oddessy Access client. Back in June, Juniper announced plans to unify their client software for SSL VPN, Universal Access Client (UAC), and WAN optimization and acceleration client, as well as integrating anti-malware software from WebRoot.

For service providers Juniper launched the MX series of routers with what they call 3D scaling. The Junos Trio chipset, an 80 million dollar, five year development cycle, offers a massive jump in router scaling over previous chipsets, so representatives from Juniper claim. Targeted for service providers and large enterprises, the MX series of routers are the types of products that Juniper built its business on: robust, high throughput, multi-service routers that are stable and reliable.

Finally, Juniper is announcing a cloud-ready infrastructure based on their SRX router line, increasing capacity with 10Gb connections, as well as support for 10 million concurrent sessions and coordinated threat control. The cloud initiative, through integration with Altor Networks, sets access policies for virtual and physical machines. 

Cisco has introduced a new network architecture: the  Borderless Network architecture, along with a five-phase plan designed to help deliver services and applications to anyone, anywhere, on any device, at anytime. As the first proof point of its Borderless Networks architecture, Cisco announced the second generation of its Integrated Services Router, the ISR G2.

The ISR G2 is designed around several key points identified as critical for future network deployment. The points include business video (which Cisco research says will increase to 77.6 percent of network traffic by 2012), service virtualization (as part of the broader network virtualization movement), operational savings and a borderless network experience, defined as a network in which locations, devices and applications are all delivered without regard to a user's or client's physical location.

According to Sashi Kiran, product manager for Cisco, the number of enterprise branches is anticipated to grow by 17% over the next two years, so branch-office capabilities are important directions for products. Branch and remote offices are critical to Cisco's Borderless Network Architecture, as the company hears from their customers that they need to see corporate services delivered to employees regardless of where the employee is sitting. Kiran says that the general networking capabilities will be several notches higher than the previous generation, and that Cisco is trying to get away from basic "feeds and speeds" comparisons, preferring to talk about the nature of the services delivered by the new-generation routers.

Kiran says, "The emphasis is on delivering richer collaboration and video experiences, with lots of emphasis on video and [quality of service] issues." As an example, he says that the new routers are video-ready, featuring Medianet -- hardware and software to support video conferencing. He also points to video surveillance and Webex (Cisco's web-based remote meeting and collaboration service) as explicit targets for the services delivered through the ISR G2.

In order to deliver on the services, Kiran says that the ISR G2 moves from 160 GB local video storage to 1TB storage per module and is built around improved CPU performance, so multiple protocols can run with desired quality simultaneously.

In talking with Kiran, the most prominent impression one comes away with is the emphasis Cisco is placing on video services in the enterprise network. Once seen as a novelty, video is in the process of becoming the tail that wags the enterprise network dog. To an even greater extent than VoIP, we're now hearing router and network optimization hardware vendors talk about building network infrastructures focused almost exclusively on the needs of IP video.

In the same way that Interstate-grade highways are built for the needs of heavy 18-wheeled trucks, with passenger cars basically riding along on the coattails of the big rigs, video is now driving the needs assessments and quality concerns of corporate networks that, until recently, were all about the needs of basic file and database serving. The requirements of video are such that it may sound quite reasonable to build network around those needs. It's best to keep in mind the complete lessons of the Interstate, though: while designing the pavement around the needs of the big rigs works well, when it comes to safety equipment and access speeds, the equipment that is perfectly suitable for tractor-trailers can be deadly for family sedans. It would be foolish to become so enamored of Web 2.0 services that we completely lose sight of more prosaic requirements. Cisco may well have kept all traffic types in mind while designing the ISR G2 (and I have no reason to doubt that they have), but the broader industry trend should be carefully watched.

The Cisco ISR G2 portfolio is available now, with pricing as follows:  the Cisco ISR G2 1900 Series starts at $1,595; the Cisco ISR G2 2900 Series starts at $1,995; the ISR G2 3900 Series starts at $9,500, Ala-Carte Services Ready Engine Modules starts at $1,000, Ala Carte Video Ready DSP Modules (PVDM3) starts at $800 and the Etherswitch Modules start at $1295.
 
The base price for the ASR1002-F chassis (ASR1002 System, Fixed ESP,Crypto 4 built-in GE, 4GB DRAM ) is US $20,000.
 

Intel and Numonyx B.V. has achieved a research milestone with non-volatile cross point phase change memory (PCM) technology. The breakthrough potentially paves the way for highly scalable, high density memory products that may one day offer low cost, high performance alternatives to traditional NAND Flash and NOR memory employed in solid state drives (SSD).

The new non-volatile PCM technology combines many of the benefits of today's memory types, but for the first time researchers have demonstrated a 64Mb test chip that enables the ability to stack, or place, multiple layers of PCM arrays within a single die. These are findings that pave the way for building memory devices with greater capacity, lower power consumption and optimal space savings for random access non-volatile memory and storage applications. 

"The results are extremely promising," said Greg Atwood, senior technology fellow at Numonyx. "They show the potential for higher density, scalable arrays and NAND-like usage models for PCM products in the future. This is important as traditional flash memory technologies face certain physical limits and reliability issues, yet demand for memory continues to rise in everything from mobile phones to data centers."
 
Intel and Numonyx have been researching new memory technologies and resistive materials over the past ten years as part of an effort to identify next generation solid state (SSD) memory alternatives. "What we wanted to do was to be able to stack multiple cross point memory on top of CMOS at the backend of the process, but this was extremely hard to do," said Al Fazio, Intel fellow and director, memory technology development. "To do this, we employed a thin film, two-terminal OTS (Ovonic Threshold Switch) selector that matched the physical and electrical properties for phase change memory scaling." With the compatibility of thin-film PCMS (phase change memory and switch), multiple layers of cross point memory arrays are now possible. Once integrated together and embedded in a true cross point array, layered arrays are then combined with CMOS circuits for decoding, sensing and logic functions.

The multi-stacked PCMS combines elements of RAM and NAND/NOR solid state memory that could make it a candidate to one day blend these different memory classes into a world-class storage solution. "NAND and NOR memory operate on blocks of data, but the PCM memory has the ability to alter a single bit of data, as RAM does," said Atwood. PCM also employs low voltage technology, in contrast to NAND, which uses an electrical charge to distinguish a "0" from a "1" as it reads memory. Because PCM is low voltage and reads the "0s" and "1s" of memory by detecting the difference between crystalline and amorphous material states of storage, it does not require the high voltages that NAND consumes. This makes PCM theoretically more scalable, and also less power-consumptive. To illustrate, NAND voltages can reach over 20 volts. This makes it hard to scale NAND below 20 nanometers.

PCM's comparatively lower voltage allows it to scale down to five nanometers or lower and still deliver a stable state. The scalability potential of PCM, coupled with PCM's ability to multi-stack as many as four layers of memory in practice, with a theoretical boundary of unlimited layers, holds promise that the technology will exceed both NAND and NOR memory in economy of  cost and in scalability. Cost and scalability have been barriers to enterprise adoption of NAND and NOR solid state memory.

"We're not talking about commercial and production milestones today, but only an initial key research milestone," said Fazio. "That milestone is that we have been able to demonstrate both yield and scalability in a PCM multi-layer megabit array. The technology must still compete against other established memory technologies with large economies of scale. This takes time to overcome." 
 

Fast, 3.5” hard drives have now reached the 2TB mark, combining capacity and performance. The new storage density provides benefits for budget buyers, as well as the mainstream, including less heat, less noise, more capacity, and of course, more speed.