Cisco UCS M-Series servers have been purpose built to fit specific need in the data center. The core design principles are around sizing the compute node to meet the needs of cloud scale applications.
When I was growing up I used to watch a program on PBS called 3-2-1 Contact, most afternoons, when I came home from school (Yes, I’ve pretty much always been a nerd). There was an episode about size and efficiency, that for some reason I have always remembered. This episode included a short film to demonstrate the relationship between size and efficiency.
The plot goes something like this. Kid #1 says that his uncle’s economy car, that gets a whopping 15 miles to the gallon (this was the 1980s), is more efficient than a school bus that gets 6 miles to the gallon. Kid #2 disagrees and challenges Kid #1 to a contest. But here’s the rub, the challenge is to transport 24 children from the bus stop to school, about 3 miles a way, on a single gallon of fuel. Long story short, the school bus completes the task with one trip, but the car has to make 8 trips and runs out of fuel before it completes the task. So kid #2 proves the school bus is more efficient.
The only problem with this logic is that we know that the school bus is not more efficient in all cases.
For transporting 50 people a bus is very efficient, but if you need to transport 2 people 100 miles to a concert the bus would be a bad choice. Efficiency depends on the task at hand. In the compute world, a task equates to the workload. Using a 1RU 2-socket E5 server for the distributed cloud scale workloads that Arnab Basu has been describing would be equivalent to using a school bus to transport a single student. This is not cost effective.
Thanks to hypervisors, we can have multiple workloads on a single server so that we achieve the economies of scale. However there is a penalty to building that type of infrastructure. You add licensing costs, administrative overhead, and performance penalties.
Customers deploying cloud scale applications are looking for ways to increase the compute capacity without increasing the cost and complexity. They need all terrain vehicles, not school buses. Small, cost effective, and easy to maintain resources that serve a specific purpose.
Many vendors entering this space are just making the servers smaller. Per the analogy above smaller helps. But one thing we have learned from server virtualization is that there is real value in the ability to share the infrastructure. With a physical server the challenge becomes how do you share components in compute infrastructure without a hypervisor? Power and cooling are easy, but what about network, storage and management. This is where M-Series expands on the core foundations of unified compute to provide a compute platform that meets the needs of these applications.
There are 2 key design principles in Unified Compute:
Big Data is not just about gathering tons of data, the digital exhaust from the internet, social media, and customer records. The real value is in being able to analyze the data to gain a desired business outcome.
Those of us who follow the Big Data market closely never lack for something new to talk about. There is always a story about how a business is using Big Data in a different way or about some new breakthrough that has been achieved in the expansive big data ecosystem. The good news for all of us is, we have clearly only scratched the surface of the Big Data opportunity!
With the increasing momentum of the Internet of Everything (IoE) market transition, there will be 50 billion devices connected to the Internet by 2020—just five years from now. As billions of new people, processes, and things become connected, each connection will become a source of potentially powerful data to businesses and the public sector. Organizations who can unlock the intelligence in this data can create new sources of competitive advantage, not just from more data but from better access to better data.
What we haven’t heard about – yet—are examples of enterprises that are applying the power of this data pervasively in their organizations: giving them a competitive edge in marketing, supply chain, manufacturing, human resources, customer support, and many more departments. The enterprise that can apply the power of Big Data throughout their organization can create multiple and simultaneous sources of ongoing innovation—each one a constantly renewable or perpetual competitive edge. Looking forward, the companies that can accomplish this will be the ones setting the pace for the competition to follow.
Cisco has been working on making this vision of pervasive use of Big Data within enterprises a reality. We’d like to share this vision with you in an upcoming blog series and executive Webcast entitled, ‘Unlock Your Competitive Edge with Cisco Big Data Solutions’, that will air on October 21st at 9:00 AM PT.
I have the honor of kicking off the multi-part blog series today. Each blog will focus on a specific Cisco solution our customers can utilize to unlock the power of their big data – enterprise-wide-- to deliver a competitive edge to our customers. I’m going to start the discussion by highlighting the infrastructure implications for Big Data in the internet of Everything (IoE) era and focus on Cisco Unified Computing System initially.
Enterprises who want to make strategic use of data throughout their organizations will need to take advantage of the power of all types of data. As IoE increasingly takes root, organizations will be able to access data from virtually anywhere in their value chain. No longer restricted to small sets of structured, historical data, they’ll have more comprehensive and even real-time data including video surveillance information, social media output, and sensor data that allow them to monitor behavior, performance, and preferences. These are just a few examples, but they underscore the fact that not all data is created equally. Real-time data coming in from a sensor may only be valuable for minutes, or even seconds – so it is critical to be able to act on that intelligence as quickly as possible. From an infrastructure standpoint, that means enterprises must be able to connect the computing resource as closely as possible to the many sources and users of data. At the same time, historical data will also continue to be critical to Big Data analytics.
Cisco encourages our customers to take a long-term view—and select a Big Data infrastructure that is distributed, and designed for high scalability, management automation, outstanding performance, low TCO, and the comprehensive, security approach needed for the IoE era. And that infrastructure must be open—because there is tremendous innovation going on in this industry, and enterprises will want to be able to take full advantage of it.
One of the foundational elements of our Big Data infrastructure is the Cisco Unified Computing System (UCS). UCS integrated infrastructure uniquely combines server, network and storage access and has recently claimed the #1, x86 blade server market share position in the Americas. It’s this same innovation that propelled us to the leading blade market share position that we are directly applying to Big Data workloads. With its highly efficient infrastructure, UCS lets enterprises manage up to 10,000 UCS servers as if they were a single pool of resources, so they can support the largest data clusters.
Because enterprises will ultimately need to be able to capture intelligence from both data at rest in the data center and data at the edge of the network, Cisco’s broad portfolio of UCS systems gives our customers the flexibility to process data where it makes the most sense. For instance, our UCS 240 rack system has been extremely popular for Hadoop-based Big Data deployments at the data center core. And Cisco’s recently introduced UCS Mini is designed to process data at the edge of the network.
Because the entire UCS portfolio utilizes the same unified architecture, enterprises can choose the right compute configuration for the workload, with the advantage of being able to use the same powerful management and orchestration tools to speed deployment, maximize availability, and significantly lower your operating expenses. Being able to leverage UCS Manager and Service Profiles, Unified Fabric and SingleConnect Technology, our Virtual interface card technology, and industry leading performance really set Cisco apart from our competition.
So, please consider this just an introduction to the first component of Cisco’s “bigger”, big data story. To hear more, please make plans to attend our upcoming webcast entitled, ‘Unlock Your Competitive Edge With Cisco Big Data Solutions’ on October 21st.
Every Tuesday and Thursday from now until October 21st, we’ll post another blog in the series to provide you with additional details of Cisco’s full line of products, solutions and services.
If you follow the news in the world of data center you probably noticed a small announcement from Cisco last week regarding the UCS portfolio… :)
To net it out in a simple way, I’ve been telling people that the trail of innovation that Cisco has been blazing with UCS just got a lot wider. That’s because this rollout is all about three key vectors that our customers have guided us to expand on:
Edge-Scale computing: taking UCS to the growing sources of computing demand beyond the core data center and to smaller scale IT organizations with UCS Mini
Padma Warrior and Joe Inzerillo discuss how technology is transforming the #MLB fan experience.
We had a stellar lineup at the event in New York. Our CTO, Padma Warrior, headlined and did a fantastic job setting the context for this wave of innovation in the frame of IoE and Fast IT. Paul Perez followed, explaining the sea change occurring in the application landscape and the customer imperatives guiding development of the UCS platform. Finally, Satinder Sethi stepped us through all the new technology we’ve added to the portfolio. Frank Palumbo hosted the event for us in New York, and I think it’s no coincidence he was rewarded later in the day by a thrilling walk-off win by the Yankees. Note that my last link there is to MLB.com, whose CTO, John Inzerillo, joined our event to share all the cool fan experience technology they’re developing.
I’d like to thank our #CiscoChampions for joining us at the event and bringing their unique and (trust me) unfiltered perspective to the news. Another highlight for me was the opportunity to tour the MLB Advanced Media Center with Matt Eastwood of IDC who joined us in New York to moderate a panel on scale-out computing. Matt, so sorry about the results of the Yankees/Red Sox game…it’s tough to overcome Palumbo-level karma. Having several of our customers and partners at the event really rounded it out, making a special day for everyone that joined us in New York and in the streaming sessions.
Jim Leach (L) and Tech Field Day panel of Cisco Champions.
To hit on all the details, the team has taken a divide-and-conquer approach here on the blog as well as youtube and our other social media venues. In addition to the links above, here are some of the pieces you can check out to learn more. Scanning the #USCGrandSlam hashtag on Twitter is another good way to take a look at the news and reactions.
On September 8th, 2014 Cisco announced support for the Intel® Xeon® Processor E5 v3 product family on the Cisco Unified Computing System™ . Simultaneously, Cisco unveiled a significant expansion of the Cisco UCS portfolio based on the Intel® Xeon® Processor E5 v3 product family for cloud-scale computing and workloads at the enterprise edge. For additional details on the new Cisco UCS servers launch with the Intel® Xeon® Processor E5 v3 product family check the blog by Cisco CTO Padmasree Warrior “Introducing New Computing Platforms for the Internet of Everything”
On the same day as the Intel announcement, Cisco captured four World Records on industry benchmarks on Cisco UCS to highlight the way in which Cisco UCS can accelerate performance across the data center. As we know, there is no better way to compare performance than by using industry-standard benchmarks, and with FOUR new World Record benchmark performance results Cisco has demonstrated Cisco Unified Computing System’s outstanding performance and IT productivity across key data center workloads.
Cisco UCS World record Benchmark results announced this week includes:
SPEComp®G_base2012- Number-one 2-socket server for SPECompG_base2012 Result: Cisco UCS C220 M4 delivered Java transactions at the rate of 160,283 concurrent Java operations per second (jOPS) and 58,478 concurrent critical jOPS on the SPECjbb®2013 benchmark. Check out the detailed benchmark disclosure report for additional information on benchmark configuration.
Oracle E-Business Suite Extra-Large Model Payroll- Number-one server Result: The Cisco UCS B200 M4 Blade Server delivered world-record results on the Payroll Extra-Large Model Benchmark, exceeding one million employees per hour with a score of 1,125,281 and outperforming the IBM Power System S824, setting the standard for performance on this benchmark. This result is an 11 percent improvement over the previous-generation Intel Xeon processor running on the same Cisco UCS server and a 3 percent improvement over the IBM Power System S824.The Cisco UCS B200 M4 Blade Server result of 1,125,281 employees per hour outperformed all competitive solutions. Check out the Performance Brief and detailed official benchmark disclosure for additional information on the benchmark configuration.
Oracle E-Business Suite Large Model Order-to-Cash- Number-one server Result: The Cisco UCS B200 M4 Blade Server set up a world record on the Order-to-Cash workload with a result of 243,803 order lines per hour, outperforming the same server configured with previous-generation processors by 5 percent on the Order-to-Cash Large Model Benchmark, processing more than 11,000 more order lines per hour. Check out the Performance Brief and official benchmark disclosure report for additional information on the benchmark configuration.
It is interesting to note that although all vendors have access to same Intel processors, only Cisco UCS unleashes their power to deliver high performance to applications through the power of unification. The unique, fabric-centric architecture of Cisco UCS integrates the Intel Xeon processors into a system with a better balance of resources that brings processor power to life. Cisco’s results demonstrate the degree to which Cisco servers deliver the power of the new Intel Xeon processor E5 v3 family. Cisco UCS maximizes Intel innovations and with performance improved up to 145 percent since the last processor generation, you can count on both innovation and versatile performance from Cisco UCS servers.
Cisco UCS delivers versatility with performance leadership across a wide range of workloads, enabling customers to eliminate infrastructure silos historically driven by unique application needs. Todd Brannon sums up in his blog post Cisco UCS: Powering Applications at Every Scale
The architectural advantages of a single cohesive system optimized for virtualized environments coupled with the industry leading benchmark performance results makes the Cisco Unified Computing System an “infrastructure platform of choice” to provide industry-leading performance in your data center. For additional information on Cisco UCS and Cisco UCS Integrated Infrastructure solutions please visit Cisco Unified Computing & Servers web page.
The Java application performance improvement of 145 percent compared the critical-jOPS score of the Cisco UCS C220 M4 Rack Server with the Cisco UCS C240 M3 Rack Server, a SPECjbb2013 MultiJVM result that was available on February 5, 2014.
The parallel processing performance improvement of 39 percent compared the SPECompG_ peak2012 score of the Cisco UCS C220 M4 Rack Server with the Cisco UCS C240 M3 Rack Server, a result of that was available on September 13, 2013.
The batch processing performance improvement of 11 percent compared the Oracle E-Business Suite Benchmark results of the Cisco UCS B200 M4 Blade Server with the Cisco UCS B200 M3 Blade Server, a result that was available on September 10, 2013.
SPEC, SPECjbb, and SPEComp are registered trademarks of Standard Performance Evaluation Corporation. The benchmark results used to establish world-record status are based on those available at http://www.spec.org as of September 8, 2014.
We have a number of exciting additions to the UCS family, which we are announcing today. Application architectures are fundamentally changing, which is driving a change to the core infrastructure. Today, you will hear about new innovations in the area of scale-out computing, which are optimized for the delivery of new applications, data analytics and the scale that is required to support these apps.
As part of this announcement, we are delivering significant innovations around what I’ll call the core data center applications. Put another way, these are the more traditional applications that keep the business running and keep employees productive. Cisco is announcing our fourth generation of our workhorse two-socket blade and rack systems as well as a broad set of enhancements to UCS Management.
Over the past five years, when I ask customers why they purchased UCS, it comes down to two key areas: more power to drive applications faster and simplified management, which accelerates IT operations. Let’s talk first about the core data center platforms and the new innovations there. Once again, Cisco is raising the bar on “enterprise class” stateless computing and performance. Today, we’re announcing the Cisco UCS M4 series, which are optimized for the widest set of enterprise use cases. Delivered through the new UCS B200 M4, C220 M4 and C240 M4, Cisco continues to deliver the highest performance CPUs, maximum memory and highest I/O performance. Besides support for fastest and highest core count and memory, we have added significant flexibility to the platforms via modular LAN-on-motherboard (LOM), Modular HBAs and Flex Storage options.
In addition to the M4 generation of servers we are introducing our 3rd gen VIC technology. This generation brings native 40Gb and advanced networking features such as NVGRE and VxLAN to UCS. The latest VIC will also enable ultra low latency RDMA capabilities.. As really fast flash storage becomes mainstream and clustered applications gain popularity, high bandwidth, low latency interconnect technology will be required for applications to see the full benefit of performance improvements. With this technology, applications requiring low latency can see up to an 85% improvement without the complexity of building Infiniband fabrics.
As much as we like talking about performance (and believe me, we like that a lot), the biggest innovation UCS brought to the datacenter was a dramatically simplified management paradigm with our UCS Management portfolio. Today, we’re announcing the addition of the software programmable ACI fabric through the Nexus 9000 and APIC into UCS Director. UCS Director is also delivering a software development kit and open API to accelerate third party integration with Cisco’s technology partners.
Finally, we are delivering new application container support within UCS Director, which provides both virtual and bare metal isolation of workloads, automated with UCS Director. Users simply request a container for a particular set of application components and UCS Director deploys the necessary infrastructure configuration required. Taken together, UCS Director becomes an even more robust control point for IT infrastructure as IT customers can now request application services and have computing, storage and networking services (such as firewalling or Layer 4-7 switching) delivered across a programmable networking fabric.
Taken along with the M-series and the UCS C3160, this announcement represents a huge step forward in the UCS portfolio to enable customers to move faster into the new application paradigm while still supporting their core data center applications.