Cisco Blogs


Cisco Blog > The Platform

VIDEO: CTO Response to Questions on Unified Computing

February 2, 2009
at 12:00 pm PST

There has been a lot of chatter about Cisco and unified computing and the data center/virtualization space…as well as what markets we are or are not entering. Our CTO, Padmasree Warrior, posted a blog on the topic of “unified computing” a few weeks ago and this blog entry is a follow-up to that post. In this video, she answers some of the common questions she’s been getting about unified computing, including why customers need it, how customers will unify their computing, and partnering in Cisco’s Unified Computing strategy. A full transcript of the video is after the jump. (You will note that this video was made via TelePresence…filmed in San Jose with our CTO in Chicago.)Transcript of Padmasree Warrior video on Unified ComputingHello. I’m Padmasree Warrior. There has been a lot of conversation in the blogosphere and media since my blog post on January 19th on Unified Computing. I asked for your feedback and appreciate the debate and dialogue you have provided on this subject. I do want to address some of the issues and questions that you have raised.What is Unified Computing? Is it true you’ve announced a new product for the data center?Cisco hasn’t announced a new product but we have started to talk about what we see as a major inflection point in the data center and how we see companies benefiting from a common architecture linking all data center resources together — what we are calling Unified Computing.We see a market transition as a result of virtualization. Virtualization architectures today are very much “assembly required” islands/silos where the burden of systems integration is on the customer. This increases costs and deployment times while decreasing efficiency. We see a market opportunity to eliminate the manual integration in favor of an integrated architecture that breaks down the silos between compute, storage, virtualization, and network platforms. Cisco is focused on innovating around an architectural approach to this opportunity and we refer to this approach as “Unified Computing.” Unified Computing unifies the three distinct virtualization arenas -- network virtualization, storage virtualization and server virtualization -- into one common architecture, within industry standard technologies and with the network as the platform.Why do customers need this?Now more than ever, IT organizations struggle to be more responsive to the business at lower cost—with the data center remaining a top cost center. Power, cooling, real estate, ever-increasing demand for compute, storage, application and business process agility remain challenges that the IT sector strives to address.Today, server virtualization has increased agility and utilization for reduced CapEx, but significant challenges remain, such as:• Additional management complexity ‘shifting costs’ to opex in IT operations• There is unclear organizational responsibility for data center management, including a lack of centralized policy• IT organizations are left to cobble together complex hardware, network, virtual machine and IT management software configuration…forcing them to serve as the ‘glue’ holding these configurations together. The data center today can best be described as ‘islands of virtualization’. Technical islands, procedural islands, organizational islands, operational islands….with no innovation in the market enabling these islands to be traversed seamlessly, securely or simply.How will customers unify their computing?Cisco’s approach is partner-centric and evolutionary, designed to provide choice while bringing customers a truly virtualized data center.Phase 1 for instance lays the foundation for data center cost containment through standardization. Core to this foundation is consistently applied network intelligence and virtualization in each area of specialization: local and wide area networking, storage networking and server/application networking. Phase 2, or ‘Unified Fabric’ -- This phase optimizes and extends data center technologies through consolidation of virtualization across the network, storage and servers/applications. Phase 3, or ‘Unified Computing’ -- Unified Computing virtualizes the entire data center through a pre-integrated architecture that brings together network, server and compute virtualization. Moving beyond that…Phase 4, or ‘Private Clouds’—is a phase that extends the advantages of unified computing into the cloud, bringing enterprise-class security, control and interoperability to today’s stand-alone cloud architectures. Phase 5, which is the ultimate vision of ‘inter-cloud’ marks our long-term transition with the market, by enabling portable workloads across the cloud. This will drive a new wave of innovation and investment similar to what we last saw with the Internet explosion of the mid-1990s. Why is Cisco expanding into Unified Computing? The architectural shift in the data center marks an IT market transition that will drive the benefits of virtualization to an entirely new level.Cisco is always evaluating market adjacencies and watching for key trends. As the compute and storage platforms, the network platform, and the virtualization platforms converge, architectures will blend and merge, creating technology disruption, but enable customer business challenges to be solved in new and creative ways. Cisco is uniquely positioned to help our customers navigate the evolution and derive value from “integrated virtualization.”What track record does Cisco have in this space? Cisco has a long heritage of working in the data center and we understand the network is critical to the data center’s ability to rapidly provide information anytime, anywhere, to anyone. For instanceo Traditional LAN switching -- Catalyst lineo MDS & Director-class SANo Data center Switching -- Nexus 7000 / 5000o Virtualization -- Nexus 1000 and VMWare partnershipso Network-based services such as Application Caching (ACE) and security firewall service module (FSM)By connecting any combination of servers, storage, and applications the network orchestrates all forms of communication across the Data Center.How do you reconcile the Unified Computing architecture with your message of partnership?The intersection of innovation across industry has spurred the most significant evolutions in the technology sector to date. Today, the architectural shift in the data center marks an IT market transition that will drive the benefits of virtualization to an entirely new level. That is why—as an industry—it is vital that the leaders of IT cooperate to drive the technology innovation and transformation to help advance our customers, industries and economy.Change brings challenges, and challenges lead to innovation. We’ve decided to embrace the challenge and believe that we can innovate for the good of our customers.Cooperation among competitors is not new. And make no mistake, in a consolidating market you will see the leaders of this industry compete in new ways. That said, our responsibility as leaders of the technology industry is to help ensure that—despite a shifting competitive landscape—we work together for the success of our customers.Cisco’s commitment is to work across industry and our customers to ensure their successful navigation through this changing environment.What comes next?Over the next few months you will see us talk in more detail about the work we are doing with our partners and the greater ecosystem to bring unified computing to the market. We’re working closely with our customers and our partners to develop innovative solutions to the virtualization challenges of today. We believe that now more than ever the network is the platform.Thank you for listening and thank you for all your feedback. Let’s keep the dialogue going and I look forward to more comments from you.

In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.

12 Comments.


  1. Note to webmasters: You guys need to post videos that do not automatically auto-play, but require a user click. someone not expecting audio in a quiet environment would be very surprised!NOTE: Rakesh, thanks for the feedback…we are in the process of changing the video. Thanks.

       0 likes

  2. It takes courage for a software developer to face the big Cisco.1. I am not a hackerI received an e-mail this morning from “cisco-us@emessage.cisco.com” with following sentence high-lighted:“Carders (hackers) are no different than other criminals – they look for the most vulnerable victim.”Following is my response to this e-mail:“I am not interested in this seminar. First, I am not afraid of hackers because I already have McAfee installed. Second, I am not hacker. If I have different opinions, I always talk to people face to face. I personally dislike people attacking from behind and attacking without a reason.”In fact, in one of my e-mail to my old friends, I said:“I have a couple of dreams. Promoting green IT is one of them, which is also the highest priority on my list. I have put in efforts for years. My goal is to let big IT players be aware of green IT and start doing something. I am not trying to bring down Cisco (in fact I can’t and I don’t want to). In stead, I want to tell them that I can bring them new business opportunities. It’s actually a win-win.”2. What’s my version of next generation ITMy goal is simple: 1) make it green, 2) solve end-user’s problems and save their money, 3) give the vendors new opportunities of making money. With 2) and 3), promoting green IT becomes “telling them” instead of “pushing them”. Following is the summary of my version:2.1High performance, high density and low power consumption hardware: saving energy, saving space while improving robustness and performance.2.2Hardware consolidation: less boxes, less wires since each box or wire can do more things. The goal is to improve each hardware uint efficiency and make the data center simpler.2.3Data center automation: reducing administration cost and reducing human introduced errors. Does “reducing administration cost” means laying off system administrators then? My answer is:”Don’t worry. When this technology is applied, we have already get out of recession. BTW, since there will be more data, there will be more data centers.”I might write my own blog with more details in the future.3. Some opinions about virtualizationServer virtualization: It’s one of the most important elements of next generation. However, if you claim that every server must have a VM, I can’t agree. For low CPU computation, it’s a good idea to consolidate several servers into one with VM. For high CPU computation, I like the tradition straightforward computation. For higher performance computation, we have to go the other direction – clustering.Storage virtualization: At storage virtualization peak time, I said:”The storage virtualization derived applications are more valuable than storage virtualization itself.” I also gave an example how to use storage virtualization to achieve non-disruptive back up and quick disk recovery.Networking virtualization: VLAN??? It’s present, not future.4. Green IT vs. virtualization, who is the next generation?It’s obvious: green IT. Server virtualization is one of the most important sub-set though.5. Why my version of next generation data center is better than Cisco’s?One reason is enough. With my version, Cisco can produce “next generation green Ethernet switch” and “next generation green router”. With their own version, they have to move towards “next generation virtualization Ethernet switch” and “next generation virtualization router”.

       0 likes

  3. It takes courage for a software developer to face the big Cisco.1. I am not a hackerI received an e-mail this morning from “cisco-us@emessage.cisco.com” with following sentence high-lighted:“Carders (hackers) are no different than other criminals – they look for the most vulnerable victim.”Following is my response to this e-mail:“I am not interested in this seminar. First, I am not afraid of hackers because I already have McAfee installed. Second, I am not hacker. If I have different opinions, I always talk to people face to face. I personally dislike people attacking from behind and attacking without a reason.”In fact, in one of my e-mail to my old friends, I said:“I have a couple of dreams. Promoting green IT is one of them, which is also the highest priority on my list. I have put in efforts for years. My goal is to let big IT players be aware of green IT and start doing something. I am not trying to bring down Cisco (in fact I can’t and I don’t want to). In stead, I want to tell them that I can bring them new business opportunities. It’s actually a win-win.”2. What’s my version of next generation ITMy goal is simple: 1) make it green, 2) solve end-user’s problems and save their money, 3) give the vendors new opportunities of making money. With 2) and 3), promoting green IT becomes “telling them” instead of “pushing them”. Following is the summary of my version:2.1 High performance, high density and low power consumption hardware: saving energy, saving space while improving robustness and performance.2.2 Hardware consolidation: less boxes, less wires since each box or wire can do more things. The goal is to improve each hardware uint efficiency and make the data center simpler.2.3 Data center automation: reducing administration cost and reducing human introduced errors. Does “reducing administration cost” means laying off system administrators then? My answer is:”Don’t worry. When this technology is applied, we have already get out of recession. BTW, since there will be more data, there will be more data centers.”I might write my own blog with more details in the future.3. Some opinions about virtualizationServer virtualization: It’s one of the most important elements of next generation. However, if you claim that every server must have a VM, I can’t agree. For low CPU computation, it’s a good idea to consolidate several servers into one with VM. For high CPU computation, I like the tradition straightforward computation. For higher performance computation, we have to go the other direction – clustering.Storage virtualization: At storage virtualization peak time, I said:”The storage virtualization derived applications are more valuable than storage virtualization itself.” I also gave an example how to use storage virtualization to achieve non-disruptive back up and quick disk recovery.Networking virtualization: VLAN??? It’s present, not future.4. Green IT vs. virtualization, who is the next generation?It’s obvious: green IT. Server virtualization is one of the most important sub-set though.5. Why my version of next generation data center is better than Cisco’s?One reason is enough. With my version, Cisco can produce “next generation green Ethernet switch” and “next generation green router”. With their own version, they have to move towards “next generation virtualization Ethernet switch” and “next generation virtualization router”.

       0 likes

  4. This was very informative and I would agree, a very good focus and path to walk towards in the future. Especially with enhanced cost and power saving initiatives well underway amongst most companies today.

       0 likes

  5. Jason: Thrilled to see the focus shifting from virtualization to infrastructure (BTW, remember me from Sun?).Unified computing”" or “”infrastructure orchestration”" (as we at Egenera like to call it) is the icing-on-the-cake for flexible, dynamic data centers. Abstract-away the I/O, network and storage connectivity, and servers become incredibly flexible resources. Plus, operating/administrative costs drop by practically an order of magnitude.I fully believe that infrastructure management is the perfect complement to virtual or physical environments.”

       0 likes

  6. Ken: Could you list all the benefits of infrastructure orchestration”"? I will list all the benefit of green IT. We can then compare whose list is longer.I have another suggestion. Since this issue is between SUN and me, could we move this discussion to SUN space?”

       0 likes

  7. In 2003 Giga Research acknowledged management of Unified Computing of virtualized platforms as important and selected Platespin’s Operation Centre as most promising new product in the category of infrastructure management. Platespin Inc. was birthed in 1999 and founded in 2000 in Toronto, Canada. The company first released Platespin Operation Centre POC”" in 2001 on 911, the company and IP barely survived the down turn, shrunk from 40+ people to 5 people, then split the POC IP into 3 pieces as island products and came back selling through small innovative resellers, to the point where Novell recognized Platespin’s true value and purchased Platespin onMarch 2008 for 205M US.Cisco has done a nice job of raising everyone’s awareness that Unified Computing is long over due, its not new in concept, mainframes offered this 30 years ago but in more or less as a centralized form, Unified Computing is new for distributed computing in the sense of linking the distirbuted islands of virtualization together in a main stream way over what is now truly the world wide web.The software management technology to achieve Unified Computing across distributed virtualized islands of storage, network and computing has been invented by small companies like Platespin,Jareva etc in the 1999 to 2001 time frame.We can thank 911 for the initial delay of Unified Computing for distributed IT to the main stream , and we can also thank the current credit crisis for really waking everyone up on this very important topic.Unified Computing over distributed virtualized infrastructure is long over due, and its great to see th big powers like Cisco in the industry get behind it.Robert Reive, Inventor of Platespin Operation Centre, former first Chairman and CTO of Platespin Inc. 1999-2002″

       0 likes

  8. Padmasree,I find your comments here on this video and on Twitter to be inspiring.In particular, I find it intriguing to see you mentioning the integration and agility of business processes within your video on Unified Computing, we in the BPM and SOA domain are following this trend very closely.By creating a unified fabric that joins multiple layers of the stack into a single interconnected domain, you help to realize a vision for a truly network connected enterprise computing paradigm. Thanks for your leadership.my 2 cents,MikoDeputy CTO, Software AG

       0 likes

  9. I think Cisco is on the right track, and demonstrating how it is a visionary rather than a follower. By rights, this is something HP should find easier to do, as they have many of the components of the IT and SOA Stack with them.

       0 likes

  10. I always expect more offerings from Cisco in cloud computing because
    etwork”" is the “”cloud”" itself.Here is my wish list of these offering … Forgive my ignorance if any of these has already been provided, and let me know where I can find more info.1) On-demand bandwidth provisioning===========================I want to temporarily upgrade my bandwidth so I can upload my data faster. And then downgrade my bandwidth after I use. And I just pay for the bandwidth that I have used.2) Load balancer that can cope with dynamic auto-scaling========================================I know some load balancer can deal with component crashes by periodically PING each endpoint and take them out of the pool if it doesn’t respond. However, in cloud computing, auto-scaling may bring up a lot of new machine instances and shrink them according to work-load fluctuation which will happen more frequently than crash cases. So I am not sure load balancer is design for dynamic scaling envrionemnt where the changes happen more frequent. For example, I am not sure how a newly started component join the pool. Is it just sending a HTTP command to the load balancer or do I need to modify the config file and reboot the load balancer ?3) Firewall that can cope with secure talkback from the cloud into the data center=======================================================Not sure if this is the firewall or the secure gateway ‘s responsibility. Since machine instance in the cloud are very transient and has a volatile network configuration, how to recognize that they come from the same organization may be a challenge. One way is to push some “”credentials”" to the machine instance so the data center can recognize them when they talk back. I guess Networking company may have a better solution.4) Latency sensitive data/compute placement and routing=======================================In a hybrid cloud environment where components are residing in different domains that has different security requirements, as well as latency between domain is higher, we need a layer 7 routing mechanism that is aware of these issue.a) We need a latency sensitive scheduler such that computation is moved close to the data it works onb) We need a activity sensitive data replicator such that data replicas will migrate close to their usersRgds,Ricky”

       0 likes

  11. Good one. I think Cisco is going to invest a huge amount of money in unified computing. Is there any white paper or official document available on the same?Gaurav SomaniGraduate student, DA-IICTIndia.

       0 likes

  12. hi !i want create data and calcul center

       0 likes