Bootcamp, Bikes and the Personal Touch

Here at Extreme Networks we are fortunate to be in Silicon Valley, a part of the world blessed not only by tremendous technology and cultural diversity, but also blessed by nature with stunning mountains and redwood forests ringing the valley. These mountains also make for good times on motorcycles, one of the “extreme” pursuits I enjoy outside of Extreme Networks.

A while back I had a flash of insight, which is rare as I am not normally prone to such things. A friend was relatively new to motorcycles and was spending a great deal of time with books and videos and other such instructional aids. Despite this, he was still having trouble with tight uphill turns. We got together and I watched what he was doing and as it turned out some relatively minor changes in form were all he needed and suddenly he had a night and day type of transformation in his confidence when facing the bugbear of tight uphill turns.

That flash of insight is that regardless of the internet and other resources you might have at your command, there is no substitute for face time with folks who have been there and done that.

Extreme Bootcamp

Extreme Networks cares about our channel partners. Here we are doing a five day Bootcamp session with distributor Catalyst Telecom.

That is why I was excited to hear from our guys in the field about their America’s Channel Bootcamp in Greenville, South Carolina, out on the East Coast. Instead of just emailing PowerPoints and PDFs or having webinars, Rob Karnbach got together with one of our distributors, Catalyst Telecom, and put together a five day TOI session for about 20 different solution providers.

The session broke down to four days of classroom instruction covering the latest and greatest. The industry moves quickly and we have introduced a number of new products and features over the past few years (BlackDiamond X8, Summit X670, Summit X440 & X430, SDN Support, AVB, Enhanced Physec Support and so on), so it was an excellent opportunity to get everyone up to speed on the new stuff as well as review traditional best practices.

Just like the new rider up in the hills, the folks in the class room and in the hands-on lab got more than just watching internet videos and staring at *.pptx decks, they also got to talk to peers from other solution providers as well as some of our more experienced SEs. It is amazing how sometimes just a couple minutes of Q&A can be worth weeks of email and solo study, particularly when you need to clear up some things that just are not clicking.

Anyway, it was great to get together with the solution providers who are the face of Extreme Networks in the field and it was great to see the positive reactions to the latest versions of ExtremeXOS and some of the new switches like the Summit X430.

If you are a solution provider or distributor, check out our Partner News Blog, which has all the latest and greatest for our partners in the channel.

Posted in Uncategorized | Leave a comment

Easier with XOS

My name is George Smith and I am a senior systems engineer with Extreme Networks. In the course of doing my job I have had the opportunity to talk with some amazingly talented individuals from both our partners and our customer base. One recurring theme we hear a lot is that it’s a lot easier to work on Extreme Networks Ethernet switches then it is to work on Cisco IOS switches.

jason_place_network_engineer

Jason Place, Sr. Network Engineer with Envision Technology Advisors.

I was talking to one such engineer, Jason Place, Senior Network Engineer with Envision Technology Advisors, LLC., an Extreme Networks partner and solution provider in Rhode Island (US East Coast). In the interview, Jason states “I would easily rather use an Extreme switch than a Cisco at any time. It’s much much easier to set up. I love the access [control] lists especially. What I can do in ten lines on Extreme is a hundred times more than what I could do in ten lines on Cisco.”

Of course not everything is immediately obvious, with some concepts in ExtremeXOS being expressed differently than they are in IOS. One example is the way VLANs are handled, with XOS applying ports to a VLAN instead of VLANs to a port. But different doesn’t necessarily mean harder, Jason explains “once you see it, it’s elegant.”

Jason Place does recommend “getting your hands dirty,” in other words, get some time on some switches in a lab and experiment. One thing we should probably mention here is that while Jason recommended getting hold of the Concepts Guide and some switches, another approach, particularly for those on a budget or with limited physical space, is to use the XOS VM to build a virtual lab, which you can read more about here.

The bottom line is that we talked to a clueful Network Engineer with years of experience and formal training on other platforms including IOS and he had some very nice things to say about Extreme Networks and the ease of use we bake in to XOS and our switches. Your mileage may vary, but with the low cost (as in free) of building a virtual switch lab, we certainly hope you can check out the power and simplicity of XOS for yourself. In the meantime, I hope you enjoy the video.

Posted in XOS | 4 Comments

Extreme Performance Rocks Active-Active Cloud Test

At Extreme Networks we are proud of building the best Ethernet switches in the world. When we first took the BlackDiamond X8 into the lab, Nick Lippis validated our feelings:

“The BlackDiamond X8 shattered all performance records with superior bandwidth, lossless performance and extremely low latency, making it the best-in-class choice for high-performance networks,” said Nick Lippis, principal analyst for The Lippis Report. “I commend Extreme Networks for designing such a remarkable product.” Read more here.

That was then.

Now we’re back, this time in “Lippis Open Industry Active-Active Cloud Network Fabric Test for Two-Tier Ethernet Network Architecture”  with the Top of Row Summit X670 and the BlackDiamond X8.

We rocked. Some highlights include:

  • Best Overall Latency
  • 2.23 microsecond server to server connections (ToR-Core-ToR)
  • Only demonstrated core failover – minimal traffic loss

Here are some of the things Nick Lippis had to say in the report:

“The Extreme X670V performed flawlessly over the six Lippis Cloud Performance iterations. Not a single packet was dropped as the mix of east-west and north-south traffic increased in load from 50% to 100% of link capacity.”

“Extreme’s MLAG implementation performed flawlessly, proving that a two-tier data center network architecture built with its Summit® X670V ToR and BlackDiamond® X8 Core switches will scale with performance.”

“Extreme Networks went above and beyond what was required in the Lippis/Ixia Active-Active Fabric Test to demonstrate and test the reliability of a data center Fabric built with its Summit® X670V and BlackDiamond® X8.”

We invite you to read the report and check out the video below featuring Nick Lippis and our own Renuke Mendis. We also invite you to read the press release. We have also gathered everything together on a single landing page, here.

Posted in Cloud | Tagged | 1 Comment

Software, the “S” in SDN

It seems that when we think of SDN, software defined networking, we often separate the discussion into two primary categories: the software of the control plane and the hardware that forwards network traffic.  But as SDN continues to evolve, it is becoming clear that the software component is actually much bigger than just the control plane and actually has an important role to play beyond the controller at the device level on the network.

There has been quite a bit of discussion in the industry around the potential of networking equipment being commoditized in an SDN deployment.  Let’s dig into that a bit deeper.  In order for networking switches to be able to perform at scale with all the reliability needed for demanding networks and data centers, there is functionality that must exist at the device level.   Functionality such as hitless failover, link aggregation and self-healing automatic process restart must be handled at the equipment level and all these features and attributes are enabled by software.  By leveraging an industry proven Networking Operating System, such as ExtremeXOS, network operators can gain all the benefits of SDN while also ensuring their ‘forwarding plane’ operates at the level that is required of today’s most demanding applications.  These things are nothing new, we have been doing them for years.

When you think about it, many of the core ideas in SDN, like centralized control, are also things that we have been doing for years. ExtremeXOS has long had open APIs allowing central control and integration with security and other appliances. Ridgeline and IDM allow central identity and policy management – not to mention Ridgeline, our network management system.

Regardless of whether or not you will be using SDN right away you will want a network operating system that is resilient, robust and modular with a proven ability to function well in highly virtualized environments. Well documented APIs and advanced scripting and automation capabilities are nice to have as well.

It is also important to consider interoperability and how well a particular vendor plays with others. At Extreme Networks, we are dedicated to open standards and interoperability. We show this every day in all the networks we are deployed in and we have shown this in the world of SDN with our functional partnerships with companies like NEC, with their ProgrammableFlow  Controller and BigSwitch Networks, with their Big Virtual Switch and Big Tap applications.

In fashion, if you stick around long enough you can see trends come and go. Ties go back and forth from skinny to fat. Hemlines go up and down and who knows, the mullet may even come back. Other things, like a good pair of wingtips, are, like a good network operating system, timeless.

Posted in SDN | Leave a comment

Open Networking Summit 2013

Day 1 of ONS started with legendary Vint Cerf being escorted into ONS 2013 (http://www.youtube.com/watch?v=QOOL-GdfkZU) by none other than 007 hero James Bond. Mr. Cerf just received the Queen Elizabeth Prize for Engineering, 2013, so being escorted by 007 made sense.

vint

James Bond, left, and Vint Cerf, right. Two legends, one show.

Day 1 of ONS 2013 ended with “ONS Idol.” Yep, singing, dancing, the crowd going bananas.  Ok, maybe more like lively powerpoints and various companies having fun on stage.  Mixed in with cocktail hours, the ONSers had a good time.

In between Mr. Cerf literally parachuting into ONS 2013 and closing with ONS Idol, Customers, Product Teams and Industry Leaders gathered to discuss various topics ranging from Customer Use Cases to Industry Collaboration efforts. For those interested, Curt Collins over a SDN Central has an excellent summary of the Keynote and Plenary speakers here (http://www.sdncentral.com/products-technologies/ons-2013-day-1-highlights-open-networking-summit/2013/04/).

In between sessions and during the evening, around 40 SDN product makers, research groups or Service Provider showcased their efforts with live demos and/or some handouts to discuss the happenings with SDN and OpenFlow.

A theme that continues to resonate is demonstrating and proving interoperability of the SDN Ecosystem. In that regard, NEC is highlighting their ProgrammableFlow Controller (PFC) strategic partnerships, including Extreme Networks.

Another central theme is Applications. The discussions are rising up the SDN Framework while still including the High Performing Network Infrastructure and of course the OpenFlow controller.  There is a focus on what applications can take advantage of the centralized intelligence of an SDN ecosystem.

In that regard, Big Switch Networks continues to highlight not only their SDN Controller, but also Big Tap and Big Virtual Switch with a live demonstration in their booth. Big Switch also provided a live demonstration during their Plenary session so clearly they are production ready.

Extreme Networks brought in a few key folks including Howard Holgate, Sam Hague and Tim Rozet from the SDN Engineering team. These are the guys that either wrote the code or are doing QA, so we were ready for anyone wanting to do technical deepdives.  In keeping with the themes of interoperability and applications, the Extreme demo showcased a couple of unique and interesting things.

  1. Mobile User launching an SDN application and initiating an OpenFlow-based Video stream from their Nexus 7 tablet over WiFi.
  2. Big Switch Controller
  3. Extreme Network OpenFlow agents provisioning the OpenFlow conditions into the Summit X670 hardware tables for high speed performance.

Our unique differentiation, showcased in this demo, is that we can program OpenFlow Quality of Service into our hardware queues providing that high level of service and performance needed for demanding applications like video.

Some of our other unique aspects, besides having currently shipping products with a single OS for consistency and predictability, highlighted at the show:

Hybrid Mode Support for Both OpenFlow and Classic Ethernet Networks

ExtremeXOS allows both network programmability and flow-based forwarding with OpenFlow, as well as more commonly used and traditional CLI and NMS-based provisioning with classic Ethernet-based forwarding decisions. Extreme Networks switches support hybrid mode on a per VLAN basis. A single port can support both OpenFlow controlled VLANS and VLANS with traditional networking services.

Link Aggregation Group for Resiliency and Redundancy

ExtremeXOS OpenFlow supports Link Aggregation Groups for system redundancy and bandwidth scaling. ExtremeXOS represents an entire LAG group as a single high capacity link to an SDN controller, enabling existing SDN applications to utilize the bandwidth scaling, load balancing, and resiliency characteristics of a LAG group without being required to manage the individual member of the LAG directly.

Hardware Queuing with ExtremeXOS

Using the ExtremeXOS CLI, interface queues are configured based on operator-defined service policies and then assigned to physical ports. When those same physical ports are also configured as OpenFlow ports, the Extreme Networks switch will report configured profile queues to the OpenFlow controller with the Queue_Get_Config_Reply message. This enables the controller to dynamically program the flows that are mapped to those configured queues, providing a rich set of traffic-differentiated services.

Automated Flow Management for Increased Flow Table Size

ExtremeXOS intelligently classifies and maps controller flow-mods to the appropriate platform hardware resources to insure maximum flow scaling. Complex flows requiring combinations of L2 and L3 match conditions are instantiated in platform TCAM ACL hardware. Simple L2-only flows are mapped to the more scalable platform L2 forwarding table. ExtremeXOS OpenFlow also fully supports OpenFlow idle_timeout and hard_timeout flow mods to evict flows from the hardware resources efficiently and effectively, allowing new flow entries as required.

Anyway, it was great to see folks like Vint Cerf and equally exciting to be able to talk to and share ideas with the folks who are building the future of networking. It is also interesting to note that while everyone was excited about the possibilities of SDN, it is clear that some excitement as reflected in functional demos was more real than others. Regardless, it was a good show and I know I am looking forward to ONS 2014.

Posted in SDN, shows, Uncategorized | Tagged , | Leave a comment

Virtual ExtremeXOS for Real Labs

Over the years I have met with many customers and usually I am asked about our “IOS” and how it works.  In some cases I have been asked about eXtreme’s OS or the customer will just ask what does XOS stand for.  In any case, I always use it as an opportunity to explain that while the marketing guys will tell you that the full name of the OS is ExtremeXOS, that to me XOS stands for eXtensible Operating System.  This is more than just trying to get out of the “IOS” shadow; it explains the power of our operating system.  eXtensible means that we are able to provide a solution not only between our products, but with our open API we can tie into other applications or other manufacturers to provide a true integrated network solution.  A system that is intelligent and can adjust to current network and environmental triggers such as time of day, user identity, device types as well as log events or third party application integration. XOS is more than just an OS; it is one that adapts to make the network smarter.

Since XOS is eXtensible it also means that it can adapt to any hardware that Extreme develops allowing us to have a single OS, edge to core.  But what else can it do? Would it be a benefit to have a switch work inside a virtual environment?  I don’t mean working with virtual servers but actually residing in a VM, a virtual switch that can be accessed just by starting it on your PC.  Being extensible means that XOS can be a full functioning switch running on a VM, either in VMWare or VirtualBox, which can be accessed for training, testing, or creating configurations.  Furthermore we have the ability to connect multiple VM switches together to create a virtual lab.

XOS VM Screenshot

Budgets are being stretched thin already and that means that the IT department cannot always have a lab to test out new features needed for new business initiatives.  Before XOS VM a customer would have to invest in a separate lab setting.  Although that is still a very practical and important step, it is not always reality.  The XOS VM means that a lab is not only available, but it is available to every member of the team, to create a test network that will make the install of a feature more of a success.  The XOS VM also means that new IT talent can be trained and on-boarded to the workings of the network with a more accurate understanding then just reading the Concepts Guide, or spending money or time out of the office in training.  The new employee can bring up the XOS VM and work with the free ENA or ENS content to learn the specifics to Extreme and to implement the lab testing at their pace.

Recently I had a situation where I needed to help a customer with a BGP design.  The design required that we use eBGP to two ISPs and iBGP on the two internal routers.  Since this design required Core licenses on 4 different switches it was hard to accomplish in a traditional lab.  Using XOS VM I was able to configure the switches, test the design, and send them, not only the configs for each switch, but also the virtual lab so that they could continue to test the config, as they developed route maps and other restrictions to the network.  XOS allowed us to work and build a network together in a virtual setting sharing information and experience remotely.

Networking is more than HW sending packets.  XOS allows us to extend the network so it to be an intelligent and integral part of the IT staff.  With the Virtual Switch functionality we are able to allow customers the flexibility to work from their desktops, regardless of where they are, and to test and validate new network features and changes, before they are implemented into production.  XOS VM is not just the next cool toy it is a real tool that allows for better integration of the IT roadmap to the business drivers of the company by allowing the creation of a full network lab on a laptop.

Want to take it for a spin? Use your eSupport userid and password to download from xkit.extremenetworks.com.

Posted in Uncategorized | Tagged , | 13 Comments

Affordable Rocket Science at ONS

Extreme Networks has always been focused on building the best Ethernet switches in the world. Over the years requirements for the network have changed and evolved, but we have always believed that aggressively priced, high performance hardware combined with a robust, modular, proven OS that runs on everything from edge to core is a winning combination that not only represents the best engineering available but also provides tremendous value to our customers.

For years we have felt that the network should be more intelligent and the design of our switches and the network OS that runs on them, ExtremeXOS, reflect this approach. We were pioneers in the real world applications of QoS and Identity Management and experience we gained there makes Software Defined Networking (SDN), OpenStack and OpenFlow all natural next steps for us.

With this in mind, we are excited to be at the Open Networking Summit in Santa Clara. ONS, running from 15-17 April, brings together the best and the brightest in the networking industry for forward looking talks and demos and this year is no exception.

We will be there with our SDN partners Big Switch Networks and NEC. With Big Switch Networks we will be showing interoperability with their SDN applications Big Virtual Switch and Big Tap. With NEC we will be showing interoperability with the NEC ProgrammableFlow Controller.

turtles

Regardless of how many turtles there are, the last one needs something to stand on.

While some like to say that SDN is bringing the commodification of networks, we need to point out that while some abstraction and virtualization is good, that it isn’t turtles all the way down. At some point, underneath all that fancy software, you are going to need some fancy hardware too and the fancier your software is, the fancier your hardware is going to need to be in order to keep pace. Remember, affordable or not, it’s still rocket science!

This is why we are proud to have the industry’s first Automated Flow Management and hardware-based QoS for SDN flows. Remember the old adage, do in hardware what you can and do in software that which you must, an adage that is alive and well today even in the world of software defined networking. That last turtle needs to stand somewhere, right?

If you are at ONS, come see us at Booth 304. We will also be at NEC in booth 103 and Big Switch Networks in booth 307. The official PR for Extreme Networks at ONS is here.

Posted in Uncategorized | Leave a comment

Bringing it Together with the Extreme Networks Open Fabric Edge

At Extreme Networks we focus on building the best Ethernet switches in the world. Our Ethernet data center architecture, the Open Fabric, reflects this approach by bringing together a compelling combination of open standards with high speed, low latency, low power, lossless, resilient multipath and advanced automation and scripting. An approach like this helps ensure that those who want to build a best of breed network can do so with confidence.

Open Fabric Edge

The Open Fabric Edge combines the attributes of the Extreme Networks open fabric with IT edge services.

The world is coming together over Ethernet. In the past different services ran on different networks. Just like there used to be a dedicated analog network for voice/PBX, video used to be the same way. Now we see a single, shared Ethernet infrastructure providing the connectivity for voice, video and data. It gets even better than that, though. It isn’t just data moving over Ethernet, increasingly it’s power for the end device as well in the form of Power over Ethernet (PoE, or its burlier cousin, PoE+).

Open Fabric Edge: Knocking Down Silos

The Open Fabric Edge is all about removing silos and 1:N Convergence

With an intelligent network, the edge is an exciting place. The edge is where services are done and the edge is where users and devices come in. For example, you probably want to assign different access rights to different devices.  An IP camera probably doesn’t need to see source code repositories nor is it likely a desk phone need access to payroll backend systems. Neither should be doing port scans and if they did both should probably be blocked from the network.

Speaking of physical security and IP cameras, having the ability to easily manage cameras, including cycling power via PoE/PoE+, is not only convenient but can result in considerable savings as camera resets no longer require truck rolls or climbing ladders as they can be achieved via an easy to use web interface. See screen grab, below:

Extreme Networks Chalet Interface

Extreme Networks Chalet provides an easy to use GUI to manage PoE/PoE+ and other physical security related issues.

But the Extreme Networks Open Fabric Edge is more about enabling things than blocking them. For example, we have partnered with Microsoft to help deliver the best possible user experience for their unified communications platform, Lync. Lync integrates well with Outlook/Exchange as you would expect from Microsoft. Lync 2013 even includes cool video codec upgrades, including H.264 SVC. Long and short of it is that this enhancement alone should allow for better quality video over roughly half the bandwidth.

Despite this more efficient use of bandwidth, the end user quality of experience is still dependent upon having a good network underpinning all the magic happening on various servers and endpoints. That is where we come in. With low latency, low jitter and no lost packets, we provide exactly the quality foundation upon which a good unified communications system can be built.

Same thing for BYOD and Wireless. We have a sophisticated architecture that eliminates controller backhaul bottlenecks. This means better performance and scalability but we still don’t sacrifice our commitment to open standards or cost effective performance. Our wired switches operate just as well with access points from other vendors, not just Extreme and we offer an onboarding portal solution via partner CloudPath.

Another system for locking in the right kind of QoS for sensitive applications is AVB – audio video bridging, a system originally developed to help free the Pro A/V community from the tyranny of the analog cable rat’s nest.  With AVB enabled, houses of worship, concert halls, theaters and other venues are no longer forced to build multiple networks, they can now run everything over a single Ethernet infrastructure. The traffic that needs to be protected and delivered with very strict timing, is, while that which is not critical happens best effort.

In the end, it’s all about breaking down silos and making the most of your network. One way to cope with budgets that are not growing and demands that are is to do more with less and one way of doing that is to run additional services over a single Ethernet network and that is exactly what the Open Fabric Edge is all about.

Want to learn more?

Extreme Networks Open Fabric Edge Press Release

Extreme Networks Open Fabric Edge Landing Page

Extreme Networks Open Fabric Edge White Paper

Posted in Uncategorized | 3 Comments

Operating Room over IP: Containing the Cost of Medicine

or_over_ip

The network is at the center of the Barco “OR over IP” approach.

The world is getting older and with the greying of the world’s population comes increased demand for medical care. This increased demand for medical care comes with a real need to contain rising costs. One example is in the field of radiology where salaries in the range of $300k-$400k per year have been par for the course.

The costs of surgery don’t just come from the high salaries of certain specialists, though. The hospital operating theater itself helps contribute. When you look at the equipment in a surgical theater, the various endoscopy cameras, patient monitoring devices, C-arms and other equipment and devices, it’s doesn’t take much imagination to see that the setup, teardown, maintenance and configuration of all these various pieces of equipment is going to be a time consuming and expensive task. You also have issues similar to those faced by pro AV in terms of large and growing rats nests of analog and proprietary digital cables.

In the pro AV world, we helped solve these problems with the AVnu Alliance and a technology called AVB. In the medical world we are solving these similar problems with Ethernet and a company called Barco.

The folks at Barco have come up with a solution they call the digital operating room or “OR over IP,” an approach that leverages the Extreme Networks Open Fabric approach to high performance, highly reliable 10 Gigabit Ethernet and combines it with their Nexxis OR Management software, digital adapters that convert video and other data to IP, HD surgical displays and other components.

With standardized Ethernet connections, build-out and tear down and maintenance are no longer as long and hard as they were, nor as expensive. With Ethernet, it’s all plug’n’play.

Modern medicine generates  a lot of imagery and video, much of it uncompressed. In order to share imagery and enable collaboration, the network must not only move data very quickly, it must do so with little latency and no dropped packets or artifacted images. Of course, with patient privacy concerns being a very real consideration in light of some of the teeth HIPAA has been shown to have grown, security, identity awareness and policy based enforcement are also nice things to have and Extreme Networks delivers on all counts.

For more on Barco OR-over-IP and Extreme Networks, check out our press release or take a moment and watch the Barco video below:

Posted in Uncategorized | Leave a comment

The Purple Heart of NCSA Blue Waters

What happens when you have 235 Cray XE6 cabinets and 30 Cray XK6 cabinets and you want to connect them together with datacenter class Ethernet? Well, if you are like NCSA, you go shopping and come home with an Extreme Networks BlackDiamond X8!

ncsa_blue_waters_interconnect_blackdiamond_x8

The Purple Heart of NCSA Blue Waters – the BlackDiamond X8

The raw numbers of the NCSA Blue Waters effort are astounding. 48,576 AMD CPUs, 3,072 NVIDIA GPUs, 1.5 Petabytes of high speed RAM with 25 PB of useable storage with 300 PB of RAIT-protected near-line storage all supporting aggregate IO performance of 1.1TB/s. One number you will not be seeing is anything related to the TOP500 list. The Blue Waters folks are smart and realize that optimizing a system to run specific benchmarks and optimizing a system to do real work are different things. While it is fairly simple to bolt together an arbitrarily large collection of high power GPUs and set up the system to run Linpack really fast, it is also very hard to program such systems to do useful work.  To put things in automotive terms, some of these benchmarks are like doing 0-60 times on a family sedan, but only after removing the radio, the air conditioner, the back seat, gutting the interior, fitting racing tires and doing other things that would make for good numbers but a relatively poor car.

type_1a_supernova_simulation

Type 1a Supernova Simulation using VisIT on Blue Waters

Beyond the numbers and of even greater interest would be the applications Blue Waters is being used for. Some of these include simulations of the development of the universe in the first billion years after the big bang and what happens in a Type 1a supernova when the dying star starts to fuse Carbon 12. Other efforts include modeling high temperature plasmas to better understand the impact of solar wind and solar flares, studying the disassembly of the protein capsid, a critical step in HIV infection, climate/weather simulations, simulations of eddy flows in hurricanes, studies of the helium convection shells in giant stars, turbulent fluid and combustion flows and a variety of other efforts requiring vast amounts of computational horsepower. The smallest of the small and the largest of the large, vanishingly brief events as well as massive processes spanning billions of years – these are all within the domain of Blue Waters and all of which have the potential to expand our scientific understanding of the universe.

While we are certainly proud to be a part of any effort with tens of thousands of CPUs and petabytes of storage, we also share the Blue Waters perspective that there is more to life than pure specs and that the real value of a system is what you do with it. With this in mind, considering the various scientific boundaries being pushed by Blue Waters, we are even more proud to be part of this project.

See the Extreme Networks Press Release for more on our involvement. Interested in learning more about the Cray hardware inside Blue Waters? You may find this video interesting.

Posted in hpcc, supercomputing | Tagged | Leave a comment