There are two technologies available to regulate power consumption in the recently introduced Nehalem servers using the Intel® Xeon® processor 5500 series.  The first is power proportional computing where power consumption varies in proportion to the processor utilization.  The second is Intel® Dynamic Power Node Manager (DPNM) technology which allows the setting of a target power consumption when a CPU is under load.  The power capping range increases with processor workload.

An immediate benefit of the Intel® Dynamic Node Manager (DPNM) technology is the capability to balance and trade off power consumption against performance in deployed Intel Nehalem generation servers.  Nehalem servers have a more aggressive implementation of power proportional computing where idle power consumption can be as small as 50 percent of the power under full load, down from about 70 percent in the prior (Bensley) generation.  Furthermore, the observed power capping range under full load when DPNM is applied can be as large as 100 watts out for a two-socket Nehalem server with the Urbanna baseboard observed in the lab to draw about 300 watts under full load.  The actual numbers you will obtain depend on the server configuration: memory, number of installed hard drives and the number and type of processors.

Does this mean that it will be possible to cut the electricity bills by one third to one half using DPNM?  This is a bit optimistic.  A typical use case for DPNM is as a "guard rail".  It is possible to set a target not to exceed for the power consumption of a server as shown in the figure below.  The red line in the figure represents the guard rail.  The white line represents the actual power demand as function of time; the dotted line represents the power consumption that would have existed without power management.

Enforcing this power cap brings operational flexibility: it is possible to deploy more servers to fit a limited power budget to prevent breakers from tripping or to use less electricity during peak demand periods.

There is a semantic distinction between energy management and power management.  Power management in the context of servers deployed at a data center refers to a capability to regulate the power consumption at a given instant.  Energy management refers to the accumulated power saved over a period of time.

The energy saved through the application of DPNM is represented by the area between the dotted line and the white graph line below; the power consumed by the server is represent by the area under the solid white graph line.  Since power capping is in effect during relatively short periods, and when in effect the area between the dotted line and the guard rail is relatively small, it follows that the energy saved through the application of DPNM is small.

One mechanism for achieving significant energy savings calls for dividing a group of servers running an application into pools or "platoons".  If servers are placed in a sleeping state (ACPI S5 sleep) during periods of low utilization it is possible to bring their power consumption to less than 5 percent of their peak power consumption, basically just the power needed to keep the network interface controller (NIC) listening for a wakeup signal.

As the workload diminishes, additional servers are moved into a sleeping state.  The process is reversible whereby servers are taken from the sleeping pool to an active state as workloads increase.  The number of pools can be adjusted depending on the application being run.  For instance, it is possible to define a third, intermediate pool of power capped servers to run lower priority workloads.  Capped servers will run slightly slower, depending on the type of workload. 

Implementing this scheme can be logistically complex.  Running the application in a virtualized environment can make it considerably easier because workloads in low use machines can be migrated and consolidated in the remaining machines.

We are conducting experiments to ***** the potential for energy savings.  Initial results indicate that these savings can be significant.  If you, dear reader have been working in this space, I'd be more than interested in learning about your experience.

If this topic is of interest to you, please join us at the Intel Development Forum in San Francisco at the Moscone Center on September 22-24.  I will be facilitating course PDCS003, "Cloud Power Management with the Intel(r) Xeon(r) 5500 Series Platform."  You will be the opportunity to talk with some of our fellow travelers in the process of developing power management solutions using Intel technology ingredients and get a feel of their early experience.  Also please make a note to visit booths #515, #710 and #712 to see demonstrations of early end-to-end solutions these folks have put together.

I’m here this week in familiar stomping grounds, the Moscone Center, in San Francisco. Today’s event started off strong with John Chambers keynote address. His speech was very engaging as he wandered through the audience, capturing the attention of nearly 10K attendees. What caught my eye especially was his focus on collaboration and Web2.0. The example he used was the recent launch of the Cisco Unified Computing Solution (UCS) which was launched via online tools such as blogs, telepresence, and flicker, check out this photo:

This shows that the virtual launch reached 10x the audience at 1/10^th the cost! I am really glad to hear that since this is what I do for a living.

John also spoke about some emerging technologies and I found out that Cisco has been working very closely with the Dallas Cowboys on increasing the customer experience. I was a little disappointed to hear John is a Niners fan, but had to expect that coming from a man and a company that was named after San Fran’cisco’, so I give him a break on that one.

It was also very interesting to hear a bit about the history of the Cisco logo, looks like times have changed and so has the logo:

After the keynote, I caught up with John and Kirk Skaugen, Executive Vice President with Intel’s Digital Enterprise Group at the Intel booth where Kirk had a surprise. Intel presented to Cisco and John a XEON 5500 processor series wafer (code named Nehalem).

Here’s another shot with a the XEON 5500 wafer:

I’ll being covering more of the event and participating in social media events during the event. Look for future updates here in the Server Room.

Wm. Hank Lea

The computer industry is filled with pundits, speculators, visionaries, salesman, brilliant architects and professors. Each provides invaluable insight into their experience, their intelligence, their alma mater, their ticker symbol, their ego and what’s next. Some win the “what’s next lottery”, others work for years of brilliance in relative obscurity.

Seemingly, a world that has deployed over 1 Billion devices a year for the last 3 years , is incapable of understanding the gravity of a new programming models, a new hardware architecture, a sleek new design that delivers on a vision that Gene Rodenberry thought of in the 1960’s or Da Vinci in the 15^th Century. What is old is new…..and let me tell you why? It will revolutionize the industry (not evolutionize…a term reserved for slower growing industry’s that require government assistant every decade or so…), transform your environment and provide freedoms you had only hoped to enjoy….and we invented it 40 years ago. Does any of this sound familiar?

It should. These are the paraphrased slogans of an industry in transition. Real products matters, product differentiation matters, standards matter, interoperability matters….and shareholders pay for future expectations.

The future of computing…is NOW. The future of the computer industry is NOW. The next generation of computer programming, software architectures and transformational technologies is NOW. As an industry we have finally begun to embrace interface, architectural and software programming standards to usher in a new era of interoperability and scalability. Behind us are the days of “proprietary interfaces” (What does that actually mean other than I am going to sell you some extra accessories that will be worthless in 2 years?), which do not provide a differentiated performance/cost advantage. Gone are the days of developing programming languages that lock-in customers to individual companies, whether vendors innovate or not. These rules of the past are slowly melting away, allowing the entire industry to embrace interoperability and standards at the highest level in history. Industry diversity is healthy and insures that the most innovative and technologically relevant companies will “win” most of the time. Allowing the 1 Billion and the Next Billion customers of the world to enjoy the best interface technology yet developed….each other.  It also provides us with a unique ability to move to the next phase in our dynamic industry’s growth, autonomic instrumentation.

At Intel, we are constantly working to develop the next great performance architecture, filled with new innovative “goodies”, as our Chief Virtualization Architect Rich Uhlig calls them. These “goodies” (a technical term that Rich borrowed from his nephew, I believe) come in the form of virtualization technologies (Intel VT-x, Intel VT-d and Intel VT-c), security technologies (Intel LT-SX), performance technologies (Hyper-Threading, Turbo Boost) and energy efficiency instrumentation (Node Manager and Data Center Manager). Soon they will also include differentiated services in the cloud which facilitate ease of use and growth for a host of vertical industries in need of innovation. The resulting architectures that emerge will be instrument rich, feature capable and as scalable as users are willing to pay for.

Why is this important? Instrumentation matters. As we apply business and personal rules to our growing compute environments it has become increasingly clear that the more tools we make available to users the better informed we are in making decisions. The more disclosure we provide to investors through the use of autonomic programming architectures the more informed they will be of their investing decisions.

How can you day trade $1B in 35 different stocks without clear autonomic controls in your data center, your database, your application and your client devices?

How can you move 450 Million people efficiecntly throughout a country for 2 weeks without autonomic controls on transportation: plains, trains, boats and automobiles, as they do during the Spring Festival in China?

How can you process 1 Billion text messages a day without clear business rules? What happens when these messages are also coming from machines to other machines, modifying databases, applications and clients?

As humans, we must apply guidelines, much like laws,  for our machines to take action when we are asleep, when we are tired, when we are not present, when we are just simply being human….to slow to react to a rapidly changing environment.

The innovators of the computer industry today understand this NOW. We do not need to discuss a vision of 40 years ago without a plan to act NOW. Claiming ideas without action is dishonorable at best, criminal at worst. The innovators of today must build products and services that help solve the problems of today. We do not need to look to 2050 without a plan to act NOW. The visionaries of tomorrow are…..not born. The visionaries of today…can call me in 10 years.

Autonomic controls are in place today, machine to machine computer architectures are here today, scalable compute engines are here today. Are they perfect, no. Are they effective, yes. The design architects, product engineers and systems designers of today need to address these concerns. Autonomic Instrumentation delivers control to the administrator, the user and the developer. Rules engines can be modified to maximize efficiency, minimize consumption and increase productivity. All of these will lead to increase shareholder (read: No just people who buy shares of stock) value across your enterprise, your school, our hospitals, our governments, and your home.

When executed properly, Autonomic controls should be able to deliver 20-25% performance and efficiency increases with each new generation of Moore’s law. In some cases, as in the Intel Xeon® 5500 Series these increases have been over 150% in virtualization performance, these increases will be a combination of software architecture enhancement and silicon optimization. In other cases, it will be through the dedicated hard work of increase instrumentation capability of a processor platform at the same price of the previous generation through energy efficiency and memory controls.

Autonomic controls will also allow end users to avert disasters in our data centers, our homes and in our hands. Autonomic instrumentation design frameworks, allow users to set parameters on data migrations, data backup, security, memory access, power consumption and virtual machine architectures.

For Intel and our new Xeon® 5500 Series processor family, and our recently announced

Intel® Nehalem-EX platform provide the new generation of platform instrumentation. As product developers, designers and architects we should all find a way to increase the tools available to our customers to take advantage of these instrumentation capabilities. I look forward to being able to announce more of these new features as we announce them and help to provide development frameworks for developers, engineers and architects to build new products and services, ushering in the future of autonomic computing innovation…today.

Running multiple Unix environments across a range of locations adds increased complexity and cost to the IT environment. I came across an interesting case study and wanted to highlight some of the key findings

YPF SAis the largest company in Argentina operating in the Oil and Gas industry. The company has 29 gas plants around Argentina running different Unix environments such as HP-UX, AIX and Solaris.

YPF SA consolidated their SAP ERP and Oracle DB environment from multiple Unix environments to Red Hat Enterprise Linux 5 with integrated virtualization running on Intel Xeon based platforms from IBM System X

Some of the key findings to highlight

• Key requirement from Unix Administration Team that "migrating from old RISC/Unix and proprietary servers to open and flexible platforms would pose no risk to the reliability, availability and performance of the systems"
• Positive impact on cost and performance; Lowered costs, simplified management and increased compatibility
• Reduction in costs especially when compared to license costs of RISC based platforms
• Increased performance and availability drove decision to scale with RHEL and Xeon
• Ability to leverage Redhat integrated virtualization. Free up internal hardware and technical resources for other projects

I guess the combination of Redhat and Intel deliver the business results that customers are seeking. What do you think?

The debate on how to best increase system capacity to accommodate growing applications has raged on for years; “scale up” with more CPU, memory, and I/O, or “scale out” with loosely connected systems.    Scaling out by adding networked systems to increase capacity has been a good economical solution for many IT managers because it allows them to grow by using less expensive, industry standard building blocks.  However, there are some notable exceptions to this line of thought.  One is that the class of applications that require shared memory and large database support are much better suited to run on a single, expandable system that scales up.  These are typically transaction processing, business intelligence and ERP solutions.   Until now, IT managers running applications that require scale-up systems larger than 4 or 8 CPUs have had limited platform choices and most were proprietary and expensive RISC-based servers.

The other problem with the scale out approach is the people, facilities, software and overhead costs and complexity of managing very large numbers of servers, which can grow to a point where the costs outweigh the performance and system cost benefits.  The industry solution to achieving better ROI has been to consolidate multiple scale-out servers onto single industry standard scale-up servers with virtualization solutions.  This is a good solution, but is limited by the number of application loads the IT manager feels comfortable placing on a single server, given the need to maintain peak performance and availability for each application.

Well, it looks like the scale-up, scale-out debate is about to take another turn.  In the server product update Intel gave on May 26th, they talked about new levels of system scalability and choice supported by the upcoming Nehalem-EX processor.  This processor will support systems that scale up to 8 sockets natively (shared memory, without any additional silicon), and up to 16 sockets and higher with node controllers from system manufactures that allow single systems to share memory beyond 8 sockets.   So far there are over 15 different designs from 8 OEMs that offer 8 socket or higher scalability.  But of course, for the class of application where scaling is important, socket count doesn’t tell the whole story of what’s needed for scalable performance.  Thread support, key for transaction processing and virtualization, scales at the rate of 16 threads per socket with 8 cores and Hyper Threading (2 threads per core).  That would be 128 threads for an 8-socket system, and 256 threads for 16 sockets.   And in order to keep those threads fed with data close to the CPU, each processor supports up to 24 MB of shared cache (1.5X current generation Xeon), and an impressive 16 memory slots per socket or 128 DIMMs on an 8-socket system.  In addition, the Scalable Memory Interconnect gives these systems 9 times the memory bandwidth of today’s top Xeon processor.  Finally, four QuickPath interconnect links per socket allow for high-bandwidth sharing of data across the system.

So the net of it is that the industry is going to see a broad selection of highly scalable, next-generation servers that significantly extend the economic advantage of industry standard scale-up solutions for business-critical, large database, and high-end virtualization/consolidation deployments.     I would expect these systems to give IT managers a very cost-effective alternative to the much more expensive and proprietary RISC-based servers they use today.

What are your thoughts?  Mike

Related Topics:

• NHM-EX Press Fact Sheet
• NHM-EX May 26th Press Briefing Video – condensed version
• IBM 8Socket Demo Video
• NHM-EX--A New Standard

Intel and Emulex will be hosting a webinar on June 3 ant 9am PDT to discuss how Emulex adapters and Intel Xeon 5500 processor based servers can help manage server sprawl, lower capital & operating costs and enable deployment of larger virtual servers & increase the number of VMs per server.  During the webcast the speakers will discuss new technologies, share benchmark results and provide tips and tricks on how to supercharge your virtual server.

Event Synopsis:

Challenging economic conditions are driving requirements to optimize performance and reduce costs in the data center. Since a majority of IT costs are related to the number of servers deployed, it’s imperative that servers are selected which provide scalable performance, automated energy efficiency and superior virtualization ratios. The time is right to leverage new technologies from Emulex® and Intel® to drive critical IT initiatives.

The webcast registration link can be found at http://www.emulex.com/company/events/webcasts.html and selecting “Next-Generation Server Technologies from Intel and Emulex”.

Fast Food. Fast Servers. Fast Savings.

A recent customer who worked for the US Department of Defense expressed an interest in using the Intel Xeon Server Refresh Savings Estimator off-line (no internet connection) due to security concerns of using their own internal business data over the internet.

For those of you who may have similar concerns, here is a procedure that will give you access to the ROI estimator on the safety of your own laptop or desktop computer.

http://communities.intel.com/docs/DOC-3204

Developing a server refresh strategy requires coordination .. among IT, business units, facilities, finance and possibly others

For many organizations, who buys the servers, maintains them and sees the power bill are all different silo'd organizations.  The issue in developing a strategy is that if each of these independent organizations don't get together refresh may never happen - why?  Because each organization only sees a portion of the overall costs and savings, what is right for one group may show a negative impact or cost.  However, because the new benefits of server refresh (doing more with less) touches so many pieces of the collective organization that the end result is usually a positive.  Kind of like how athletes need to rely on each other to achieve a common goal - winning the game.

So how do you get everyone on the same page?  In sports, this is the role of the coach or in some cases the on-field leader (quarterback, captain ...).  Last week i sat in on a data center summit hosted by Intel IT.  Inside intel, the quarterback is corporate finance who can see all the pluses and minuses that impact the corporate P&L and help optimize a decision that is best for the company and shareholders.

Last year Intel IT saved $45M in operational savings and cost avoidance while supporting growing compute demands.  Read the 2008 Annual Perf Report

As they would say in Disney's High School Musical - Get your Head in the Game

Chris