1 54 55 56 57 58 70 Previous Next

The Data Stack

1,045 Posts

There are literally thousands of ways a new Intel® Xeon® processor 5500 series based server can benefit your business. I’ll spare you the complete list, but let’s start with the major ones. When you ask yourself the following questions about your business and its needs, you may just find that getting a server right now is the right decision for you.

  • Are your servers getting old? Things change, and with the latest server processors, things have changed a lot. The new Intel Xeon processor 5500 series offers up to nine times the performance of a server purchased just four years ago.1 Even though you’ll be outlaying cash for new equipment, spending wisely on a server now will probably save you money in the long run. From energy usage to maintenance costs to software licensing fees, it adds up. Plus, with a new server, you get a new warranty and compatibility with the latest applications. That means fewer hassles for you. Don’t wait until your server breaks, you don’t want to discover the cost of losing data and business downtime.
  • Are increasing employee and data demands taxing your systems and your staff? If so, you need the processing power, energy efficiency and reliability a Xeon-based server can deliver – 24/7 uptime, industry-leading performance, memory protection, and a server that automatically shuts down to save energy.
  • Do you want to improve productivity? The increased performance of the latest Intel Xeon-based servers enables your IT equipment and your staff to do more with less.
  • Is cost-cutting a high priority? With an Intel Xeon processor 5500 series based server, you can benefit from significant energy savings, the reduced costs of easier maintenance and need for fewer servers. By consolidating servers, you can save up on your utility bill. Check out the estimator tool www.intel.com/go/xeonestimator to see how much you could save.
  • Are you ready to implement new software? Some of the latest software advances demand newer server capabilities. If you’re looking to implement VMWare* or Microsoft Windows Small Business Server 2008*, an Intel Xeon processor 5500 series server delivers the performance your new business applications will need.
  • Is your company still using a desktop as a server? Then, now is the time to step up to a real server. Down markets are when smaller companies can take advantage of their agility. Plus, you definitely can’t afford downtime when customer service is so critical.

Ultimately, the biggest question is: Can you afford NOT to invest in the newest Intel Xeon processor-based servers in this economy?

Learn more about our new server processors:

  • Read this brochure to learn more about the advantages of Intel® processor-based servers for small and medium businesses.
  • And talk to your IT solutions provider.

Also, I’d love to hear your best reasons for buying new servers, so I can add them to the list. If you have already made the transition to the new Intel Xeon processor 5500 series, please share you story.

1Performance increase based on Intel comparison using SPECjbb2005 business operations per second (bops) between four-year-old single-core Intel® Xeon® processor 3.8GHz with 2M cache based servers and new Intel Xeon processor X5570 based server. Intel consolidation based on replacing nine four-year-old single-core Intel Xeon processor based servers with one new Intel Xeon Processor X5570 based server while maintaining SPECjbb2009 performance. Costs and return on investment have been estimated based on internal Intel analysis and are provided for information purposes only. Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate performance of Intel products as measured by those tests.  Any difference in system hardware or software design or configuration may affect actual performance.  Buyers should consult other sources of information to evaluate the performance of systems or components they are considering purchasing. For more information, visit www.intel.com/performance/server.

Ken Lloyd

Nehalem Rocks - now use it

Posted by Ken Lloyd Apr 17, 2009

It has been a couple weeks now and just in case anyone may have forgotten, Nehalem rocks.  In my job I talk to customers every day and even though I have become a bit jaded by the numbers associated with the new Xeon 5500 series processors, customers constantly remind me just how significant this change is.  The leap in performance is unprecedented in the history of the "Xeon" family.  The opportunity that this creates for businesses is tremendous.  Chris has blogged a lot about the economics of refresh and anyone who is not paying attention has a job that is just too cushy.  For the rest of you that actually worry about performance, data center power capacity, data center space, etc - please pay attention.

 

Data center space is for many businesses the single most expensive "office" space they own. Consider this coupled with the reality that demand for computing continues to grow, and 81% of businesses report line of site to data center capacity ( power or space ) overflow.

 

Any data center owner who is facing capacity challenges and not aggressively refreshing and consolidating should be "made redundant".  (opinion)

 

some very very round numbers to consider:

If you have servers that are 4 or 5 years old, the new Xeon 5500 series processor based servers can be as much as 10 times faster.

Those old servers ( if they are typical enterprise servers ) are setting at about 10% utilization.

 

When you refresh and consolidate you are going to virtualize - so now, lets do the simple back of napkin math on the opportunity :

you have 1000 servers that are at 10% utilization.

with virtualization you could boost up to 50% utilization - 5 to 1 consolidation - now you have 200 servers

the new servers are 10 times faster - so with an aggressive refresh - now you have 20 servers

 

Demand is not going away, and eventually you will fill up all this new capacity and of course in the real world this isn't all going to happen day one,

BUT, anyone complaining of capacity issues AND using old hardware, must not be paying attention.  Please wake them up.

In my blog titled “why buy for the little guy, I shared that I was in the market for a new home desktop to replace my existing one.  Well, today I spent my 2009 tax refund check on a new computer – an iMac. 

So why the secrecy?  ... The iMac is a surprise for my wife’s birthday.  My wife has wanted a Mac since we left college and I’ve always been a PC guy (grew up with one, always used them at work, etc.) and when I started working at Intel almost 10 years ago, I could not justify a Mac given my corporate loyalties and who was paying the bills.  When Apple adopted Intel architecture a couple years ago, my options were now opened (really my old excuses were not longer valid).  

I’m very excited to install the new iMac as the challenges of my old technology and the limitations and headaches they were giving me will be gone - and I think I’ll finally get some good-guy points with my wife.

Chris

Hello.  My name is Omar Sultan and I am from the Cisco Data Center Solutions team.  I am a fairly regular blogger over on the Cisco Data Center blog and I'll be spending some time over here too.  The development of the Cisco Unified Computing System launch gave us the opportunity work with some seriously smart folks at Intel and I got a chance to capture some of those conversations.

 

Here is a podcast that I recorded with Jake Smith from the Intel Advanced Server Technologies team.  One of the things that makes Cisco UCS work is that Intel and Cisco share a common vision for the value of virtualization in the data center and how it can be practically applied to address customer challenges.

 

Over on my blog, you can also check out a conversation with Ed Groden on why it makes sense to invest in new platforms in these challenging times--he makes a pretty good case.

 

Anyway, hope the podcasts are helpful, be sure to also check out my blog, and definitely let me know what you would like to hear about from a Cisco perspective.

 

Omar

 

Omar Sultan

Cisco Systems

With the introduction of the Intel Xeon processor 5500 series last month, I wrote a blog that discussed that server refresh was an intelligent investment in that it could deliver a rapid payback on investment. For the past few years, I have been working to understand the costs and benefits of server replacement and there are a few conclusions I can draw.

1)      Server Refresh is not new concept.  This approach has existed for decades.  People replace technology as it ages because new software and new technologies enable better business capabilities and as technology ages, the warranty expires and incidence of failure increases. How many of you still have your first mp3 player?

2)      ROI and Refresh Vary. The rate of refresh is a balance of the investment required (purchase, install, removal, validation, etc) and the savings achieved (operational costs, cost avoidance, employee productivity) balanced with the business opportunities available to you (business growth or new business markets, cost of capital, revenue generating investments)

3)      One Size Does not fit all.  Every business looks at financials and opportunities for their business a little differently and calculates their costs and savings differently.

So a few months ago, I embarked with some of my peers, with Intel IT, and industry leading ROI and TCO consultant Alinean, to apply what I have been learning and build an interactive tool to help you model your savings opportunity for server refresh and replacement. 

We identified and were able to model eleven cost and savings categories (both pluses and minuses) in the Server Refresh ROI calculation and make these cost category assumptions able to be included, excluded or modified by you.  You can model and view scenario output real time and print/email reports to share with others.

I invite you to learn more about the tool with this informal how-to-use guide , or better yet, use the tool and estimate how much you could save replacing old servers with new.  Try the new Intel® Xeon® processor-based Server Refresh Savings Estimator today.

You can provide feedback through the tool’s registration process or by responding directly on this blog. I look forward to hearing from you either way.

Thanks, Chris

There has been lots of discussion recently about whether its better to replace or upgrade existing CPUs in your installed base of servers rather than purchase new servers. I wanted to share some thoughts with you that might clarify why a new server purchase is the better option for most IT departments.


Here are some of the challenges that an IT department must face when considering replacing CPUs rather than buying new servers.

  1. Does the existing system support the new CPU – most CPUs require specific BIOS versions, is there a BIOS update available for the server that supports the new CPU ? Also the server motherboard may not have been tested by the OEM with the new CPU.
  2. Has the software stack you are running on the server  been validated on the new CPU.
  3. Replacing a CPU is a non-trivial exercise – it takes time and you run the risk of damaging a working server
    • the server must be shut down and dismantled to access the existing CPU
    • the existing CPU/heatsink combo must be removed. The heatsinks used by OEMs in branded servers are specifically designed for the server in which they are used. These heatsinks typically have significant mass so they are usually very firmly attached to the server chassis to prevent damage due to shock and vibration whilst in transit and in use.
      • The existing heatsink must be removed from the current CPU, which may not be easy if the system has been in use for some time the thermal bonding may have hardened permanently attaching the heatsink to the CPU.
      • The heatsink must be attached to the new CPU – with the appropriate thermal bonding.
      • The CPU/heatsink combo must then be correctly re-installed in the system and the system re-assembled.
    • It is also necessary to take into account that removing/changing a CPU may also void or otherwise affect the system warranty.


    It is conceivable that some IT folk may want to consider this approach but the risks associated with undertaking this operation are very high and many IT departments will take the approach of not touching working systems.  If you are still not convinced its also worth considering -


    • Replacing the CPU in an old server may not significantly improve its energy efficiency. The latest generation server designs not only use latest CPUs but they incorporate many new features that improve the overall energy efficiency of the complete platform – making them a much better proposition when looking to reduce overall data centre utility bills and OpEx costs.
    • Upgrading the CPUs in an old server may expose other limitations of the server in terms of memory and I/O, this could result in having to upgrade many other parts of the server resulting in an overall higher cost than replacing the server with a new purpose designed solution

     

    So, as far as I can see very few IT departments are going to seriously consider replacing CPUs in their existing installed base and will look instead to deploy latest generation high performance energy efficient server designs – i.e. servers based on Xeon 5500 or Xeon 7400


    What’s your opinion – are you prepared to attempt to upgrade your CPUs or will you refresh the complete server system to get the latest technology for all elements of the server platform

    Greg Wagnon

    The buzz around Nehalem

    Posted by Greg Wagnon Apr 8, 2009

    A part of my job these days is to interact with and track online press content for our server products.  The launch of the Intel Xeon Processor 5500 Series (codenamed Nehalem) product was a big one.  Big by any standard of measurement (...except for perhaps Geologic time).  I thought I would share with you a peak at some of the metrics we have looked at as an outcome of last week's product launch.

     

    Metrics are always tricky, because the source of the data is always something you can question, and frequently find holes.  But, if you take a bunch of data from different sources, stand back a bit, then look at it with your hands cupped together over your eyes to block the shiny distractions (think big picture), you often get some actionable tidbits out of it.

     

    Something as simple as Pageviews is a metric of success.  The idea is that you are measuring the number of people who look at your page... then you look deeper and find that bots and search sites are also looking at the new content to categorize it, log it, and have it ready for people to search upon.  So, Pageviews are a bit of a can of worms.  Good can, good worms, but not necessarily what you were expecting to find when you opened it.  So, we look at it with some measure of caution.  Again, big picture, you do see some trends that tell a little bit of the story.

     

    Note: I only discuss and cover a few items... there are many, many, many more.

     

    First here are just a few of the landing pages that I personally keep an eye on...

    Community sites; 'The Server Room' and its new sub-community the 'Server Solutions Insider' where you are right now.

    On Facebook, we have some fan pages: 'The Server Room'  and  ''Intel Xeon 5500 "Nehalem"'

     

    For the Community sites we can use a simple tool like Google Analytics to see Pageviews.  The following shows the pageview trend of 'The Server Room' over a few months.  Again, numbers are not really important as much as the pattern you can see.

    TSR_Pageviews.JPG

    Results: Weekend traffic is lower than mid-week, a couple of product launches and the 'buzz' around them drive a fair amount of traffic, and finally that last peak was pretty big compared to anything previous.  All good trends to be aware of.  The next big step is to do something with that information... and that is something to share another day.

     

    For the Facebook pages, we get some nice metrics directly from the admin tools.  Here is some nice trend data on the 'Intel Xeon 5500 "Nehalem"' page

    IX5500NHM_Fans.JPG

    Results: The total number of 'Fans' (dark blue) are growing, but on a daily (light blue) basis, we only see some peaks and not a consistent growth.  Actionable item (assuming we find such a metric as growing Facebook fans of vital importance) is that we look into how to promote the site more, and make it worth people's time to join.  But, you have to look at the forum and the point of it all... not be pushy.  Then decide what to do (if anything) from there.

     

    Now we switch gears a bit and look at some external (non-Intel) sources that we like to keep an eye on.  These are journalist websites and specifically we look for certain articles and product reviews.  The ones that actually test hardware, then give their results and analysis are key to watch since their influence is vital to knowing how well a product might be perceived.  Here are some articles in particular that I found especially interesting (no particular order):

     

    The Tech Report - Intel's Xeon W5580 processors

    AnandTech.com - The Best Server CPUs part 2: the Intel "Nehalem" Xeon X5570

    The Inquirer - Nehalem proves its server mettle

    IT168.com (Chinese Language) - Intel Nehalem-EP处理器首发深度评测

    CRN - Review: Intel's New Nehalem Historic, Game Changing

    InfoWorld - Test Center: Intel's Nehalem simply sizzles

    2cpu.com - Nehalem: Xeon Gets Core i7 Upgrade

    Tecchannel (German Language) - Test: Intel Xeon X5570 Nehalem-EP

    RealWorldTech.com - Nehalem Performance Preview

    Hardware.info (Dutch Language): - Intel Xeon X5570 'Nehalem' test

    The Inquirer - Double Nehalem for double power

    The metrics you gather from these (as with anything), depends on what you want to measure.  If you simply want to count the number of reviews... ok, there are a 11 (many more are out there).  If you want to look at the number of benchmarks where our product came out on top... ok, a large majority.  If you want to look at the 'tone' of the article... ok, that is very dependent upon the reader's mood (I'm feeling pretty good actually) and even more on the mood of the writer at the time it was written.  So, what do we get from all this?  We take all of these things and give it the 'take a few steps away' view (big picture again).  Hey, it all looks good.  Actionable items... something else to share with you another day.  ;o)

    All in all, a lot of good stuff to consume around the launch of the Intel Xeon Processor 5500 Series (codenamed Nehalem) products.  It is a joy to follow it, a deeper joy to be a part of it, and this product represents a 'new normal' for those of us that interact with the social media aspect of things.

    Leave me a note, I would be happy to explore this topic with you more.

    - GW

    In my last blog I talked about working on great projects which were “special”. Special in that everyone enjoyed coming to work, they worked well together, and part of the “magic” was we all knew we were working on something revolutionary.  Well that special, revolutionary project is now available for all to see, and it is known as the Intel® Xeon® processor 5500 series and Intel Xeon® 5500 series chipset.

    What amazes me the most about this project/platform is the incredible leap in performance compared to its previous generation platform which was based on the Intel Xeon processor 5400 series. For a new generation platform, 20-30% improvements in performance is typical.  And 50% vs. the previous generation platform is above normal, but the new Xeon 5500 series platform out performs the previous generation platform (Xeon 5400) by 2X or more on many benchmarks.  That’s right…nearly twice the performance!  (Click here for performance details.)

    So how did they achieve this monster leap in performance? 

    -          Did they double the core frequency? 

    -          No…in fact core frequency has gone down slightly.

    -          Did they double the number of cores?

    -          No…same number of cores.

    -          Did they make major changes to the CPU micro-architecture…like issuing and retiring many more instructions per clock? 

    -          No.  Same 4-instruction per clock issue/retire capabilities.

    -          Did they use a new silicon process technology? 

    -          No…both use the same 45nm process.

    -          Did they increase cache size? 

    -          No…total L2 + L3 cache size actually went down (9MB vs 12MB).

     

    So how did they “double’ the performance?  This is what truly amazes me.  This team was able to essentially double the performance of the platform, without changing the most obvious (e.g. # of cores, CPU frequency, major micro-architecture changes, Si process technology or cache size).  Instead, they made many changes and optimizations to the entire “platform” as well as some incremental enhancements to the processor micro-architecture (like deeper queues)…which collectively removed bottle necks in many different places and the results are nothing short of fantastic. 

    The various changes added up to “major” improvements in performance.  Some of these changes are listed below…shown in comparison to the previous generation Xeon 5400 series platform which was/is no slouch.  Even today, more than a year after its introduction, the Xeon 5400 was still was the highest performing 2-scoket platform on many benchmarks.  That is until the Xeon 5500.

    Feature

    Xeon 5400

    Xeon 5500

    # Cores

    4

    4

    Core Frequency

    3.4GHz

    3.2GHz

    Instructions per clock

    4

    4

    Process Technology

    45nm

    45nm

    L2 Cache Size

    2 x 6MB

    4 x 256MB

    L3 Cache Size

    N/A

    8MB

    Intel® Hyper-Threading Technology

    No

    Yes

    Intel® Turbo-Boost Technology

    No

    Yes

    Queues and execution resources

    Baseline

    Deeper queues & more resources

    Bus Connection

    FSB – 1.6GHz

    QPI – 6.4GT/s

    Memory Controller

    Discrete

    Integrated

    Memory Channels         (2 Socket platform)

    4

    6

    Memory Type

    DDR2-FB-DIMM

    DDR3

    Max # DIMMS                (2 Socket Platform)

    16

    18

    Memory Frequency

    533, 667, 800MHz

    800, 1066, 1333MHz

    Virtualization Features

    Intel VT

    Intel VT + Enhancements

    PCI Express

    Gen 1

    Gen 2

    All I can say is wow!  And all this performance comes in a lower platform power envelope than the Xeon 5400.  The performance and power savings are a true testament to this team’s ability to work together and deliver a truly revolutionary product. Congratulations to the entire “Nehalem” team (aka Xeon 5500)!    Click the link below to find out more about “Nehalem”. http://www.intel.com/products/processor/xeon5000/index.htm

    kW of Power.  BTU of Cooling. Square feet of Datacenter Space.  What do they have in common?

    Power, Cooling and Space are resources – more specifically, constrained resources that are available to support the delivery of compute capability. As datacenter managers look at their projected compute capacity in the coming years, it becomes clear that these scarce resources will eventually run out – in fact it’s estimated that 70% will run out of power or cooling capacity in the next two years. Adding power or cooling capacity is expensive and there likely isn’t any budget for that, especially in today’s economic environment. If there isn’t budget for adding resources, there surely isn’t funding to build additional datacenter space. So how does IT get past this impasse?

    By making more efficient use of the constrained resources with Intel® Xeon® Processor 5500.

    In the launch announcements and blogs over the course of the last week, you have heard about the cool features and improvements delivered by the Xeon® 5500.   9X the performance of older single core products.  Significantly reduced power consumption at all points of the load line between idle and max utilization.  Interesting nuggets by themselves, but when taken in the context of the datacenter, they are powerful capabilities that IT can use to address their resource constraint issues.  Let’s look as two scenarios at opposite ends of the spectrum.

    Scenario 1: Same Compute, Less Resources – Assuming an installed base of single core servers, you can replace the legacy servers with approximately 89% fewer Xeon® 5500 servers – fewer servers take up less datacenter space, consume less power and require less cooling – in fact, now you have headroom to add servers to meet growing compute requirements moving forward.

    Scenario 2; More Compute, Same Resources – For those that that crave all of the compute capacity they can get their hands on, deploying Xeon® 5500 servers would increase compute capacity by 9X in the same power, cooling and floor space constraints (again, assuming a single core installed base) and consume approximately 18% less power than the legacy servers.

    The kicker is that although it seems somewhat counterintuitive, when you run the numbers it actually makes financial sense to refresh old servers with Xeon® 5500 servers.  We estimate that the power and OS savings associated with Scenario 1 can pay off the investment in as little as 8 months, and those OPEX savings continue for the life of the server.

    For both scenarios, Xeon® 5500 also delivers improved energy efficiency with Integrated Power Gates and Automated Lower Power States, which automatically and dynamically adjusts power and performance to the specific needs of the work being done. Throw in system level instrumentation capabilities to report and cap system power, and you can further reduce your operating costs by adjusting the HVAC output to the specific heat output of the servers in real time.

    Power, cooling and space resources aren’t likely to start growing on trees, but Xeon® 5500 is a key to enabling a more efficiency datacenter,  to getting more out of every kilowatt, BTU and square foot that are available and to driving Datacenter Efficiency.

    During a keynote at the recent VMWorld EMEA event, in Cannes, Dr. Wolfgang Krips, VP, SAP Managed Services postulated that the Cloud Computing industry could become like the airline industry - not in terms of its energy consumption as has been speculated by various environmental groups and analysts but in terms of the way IT managers buy Cloud Computing services.

    • Today there are full service airlines ( seat reservation, in-flight meals, luggage handling - the works ) and low cost airlines ( open seating, bring-your-own food & pay extra for hold baggage ) - you pay your money and take your choice as to the type of service you want.
    • Ticket prices vary enormously depending on routing and day/time
    • Over-booking is an accepted practice and having a ticket does not always guarantee a seat
    • Departure/Arrival times are variable - weather, air-traffic delays etc
    • You can but your tickets from the airline directly , via a portal ( www.expedia.com, www.opodo.com etc ), as part of a complete package - flight, hotel, car etc, last minute or discounted from a bucket shop .

    When you think forward as to where the Cloud Computing industry is going it quite easy to imagine that all of these elements could be applied to future cloud offerings

    • Prices will depend on the SLA offered - guaranteed uptime, data integrity or just take lowest cost compute resource available.
    • Portal sites will act as brokers for the various services available and sell capacity - we are already seeing this from companies like Zimory ( www.zimory.com )
    • Underutilised data centres may sell off excess capacity at discounted rates just to fill their facilities or the popular services may raise price to limit demand
    • response time/completion time of a job run in the cloud will be non-deterministic - dependant on network traffic and system loading

    So, definitely food for thought as to what the future of Cloud Computing will bring and how IT might interact with the various providers on the market place.

    Are there other business models being proposed for Cloud services - I would be interested in hearing your opinions.

    The recently introduced Intel® Xeon® 5500 Series Processor, formerly code named Nehalem brings a number of power management features that not only improve on energy efficiency over previous generations, such as a more aggressive implementation of power proportional computing.  Depending on the server design, users of Nehalem-based servers can expect idle power consumption that is about half of the power consumed at full load, down from about two thirds in the  previous generation.

     

    A less heralded capability for this new generation of servers is that users can actually adjust the server power consumption and therefore trade off power consumption against performance.  This capability is known as power capping. The power capping range is not insignificant.  For a dual socket server consuming about 300 watt at full load, the capping range is in the order of 100 watts, that is, for a fully loaded server consuming 300 watts, power consumption can ratcheted down to about 200 watts.  The actual numbers depend on the server implementation.

    The application of this mechanism for servers deployed in a data center leads to some energy savings.  However, perhaps the most valuable aspect of this technology is the operational flexibility it confers to data center operators.

    This value comes from two capabilities:  First, power capping brings predictable power consumption within the specified power capping range, and second, servers implementing power capping offer actual power readouts as a bonus: their power supplies are PMBus(tm) enabled and their historical power consumption can be retrieved through standard APIs.

    With actual historical power data, it is possible to optimize the loading of power limited racks, whereas before the most accurate estimation of power consumption came from derated nameplate data.  The nameplate estimation for power consumption is a static measure that requires a considerable safety margin.  This conservative approach to power sizing leads to overprovisioning of power.  This was OK in those times when energy costs were a second order consideration.  That is not the case anymore.

    This technology allows dialing the power to be consumed by groups of over  a thousand servers, allowing a power control authority of tens of thousands of watts in data centers.  How does power capping work?  The technology implements power control by taking advantage of the CPU voltage and frequency scaling implemented by the Nehalem architecture.  The CPUs are one of the most power consuming components in a server.  If we can regulate the power consumed by the CPUs we can have an effect on the power consumed by the whole server.  Furthermore, if we can control the power consumed by the thousands of servers in a data center, we'll be able to alter the power consumed in that data center.

    Power control for groups of servers is attained by composing power control capabilities of power control of each server.  Likewise, power control for a server is attained by composing CPU power control as illustrated in the figure below.  We will explain each of the constructs in the rest of this article.

    hierarchy.png

    Conceptually, power control for thousands of servers in a data center is implemented through a series of coordinated set of nested mechanisms.

     

    The lowest level is  implemented through frequency and voltage scaling: laws of physics dictate that for a given architecture, power consumption is proportional to the CPU's frequency and to the square of the voltage use to power the CPU.  There are mechanisms built into the CPU architecture that allow a certain number of discrete combinations of voltage and frequency.  Using the ACPI standard nomenclature, these discrete combinations are called P-states, the highest performing state is nominally identified as P0, and the lower power consumption states are identified as P1, P2 and so on.  A Nehalem CPU supports about ten states, the actual number depending on the processor model.  For the sake of an example, a CPU in P0 may have been assigned a voltage of 1.4 volts and 3.6 GHz, at which point it draws about 100 watts.  As the CPU transitions to lower power states, it may have a state P4 using 1.2 volts running at 2.8 GHz and consuming about 70 watts.

     

    The P-states by themselves can't control the power consumed by a server.  The CPU itself has no mechanisms to measure the power it consumes.   This mechanism is implemented by firmware running in the Nehalem chipset. This firmware implements the Intel(r) Dynamic Node Power Management technology, or Node manager for short..  If what we want is to measure the power consumed by a server, looking only at CPU consumption does not provide the whole picture.  For this purpose, the power supplies in Node Manager-enabled servers provide actual power readings for the whole server.  It is now possible to establish a classic control feedback loop where we compare a target power against the actual power indicated by the power supplies.  The Node Manager code manipulates the P-states up or down until the desired target power is reached.  If the desired power lies between two P-states, the Node Manager code rapidly switches between the two states until the average power consumption meets the set power.  This is an implementation of another classic control scheme, affectionately called bang-bang control for obvious reasons.

    NM.png

    From a data center perspective, regulating power consumption of just a single server is not an interesting capability.  We need the means to control servers as a group, and just as we were able to obtain power supply readouts for one server, we need to monitor the power for the group of servers to allow meeting a global power target for that group of servers.  This function is provided by a software development kit (SDK), the Intel(r) Data Center Manager or Intel DCM for short. Notice that DCM implements a feedback control mechanism very similar to the mechanism that regulates power consumption for a single server, but at a much larger scale.  Instead of watching one or two power supplies, DCM oversees the power consumption of multiple servers or "nodes", whose number can range up to thousands.

     

    dcm.png

     

    Intel DCM was purposely architected as an SDK as a building block for industry players to build more sophisticated and valuable capabilities for the benefit of data center operators.  One possible application is shown below, where Intel DCM has been integrated into a Building Management System (BMS) application.  Some Node Manager-enabled servers come with inlet temperature sensors.  This allows the BMS application to monitor the inlet temperature of group of servers, and if the temperature rises above a certain threshold, it can take a number of measures, from throttling back the power consumed to reduce the thermal stress on that particular area of the data center to alerting system operators.  The BMS can also coordinate the power consumed by the server equipment, for instance with the  CRAC fan speeds.

     

    DataCenter.png

    With this discussion we have barely begun to scratch the  surface of the capabilities from the family of technologies implementing power management.  In subsequent notes we'll dig deeper into each of the components and explore how they are implemented, how these technologies can be extended and the extensive range of uses for which they can be applied.

     

     

    Two weeks ago, I flew to Mexico City to present on virtualization technologies to both government agencies and private industry.  In both cases their issues were the same.  They are trying to do more with less.  In these times of global economic uncertainty, businesses are being challenged to reduce spending, while still improving infrastructure to keep up with business demand.  This is true, especially in the US, where in one case the construction of a 300 million dollar data center was put on hold and instead IT was task to reduce their server footprint in an at-capacity data center. The new focuses . . . find ways to reduce overall power and cooling costs.  Almost every company is looking at virtualization as one key component of the equation to finding solutions to these data center problems.

    The combination of a managed virtualization solution coupled with an efficient Intel processor based server is one highly effective means to solve the “do more with less” mandate.  Let’s start by talking about the new Xeon 5500 processor that was just unveiled last Monday.  You have a need to reduce power and consolidate servers?  A Xeon 5500 based server can effectively replace eight to nine older single-core servers. 9x performance improvements have been seen using things like Turbo Boost.  The processor idle power drops to only 10 watts, enabling a 50% reduction in system idle power compared to our previous generation chip.  Everything I’m seeing on this is that you can recoup your capital investment in around 8 to 9 months from reduced maintenance, power use, software licensing, and cooling costs. Your energy savings alone can be as high as a 90% reduction!  That’s big! 

    Check out more details on the launch of the new Xeon 5500 processor with Intel’s press release.

    http://www.intel.com/pressroom/archive/releases/20090330corp_sm.htm?iid=pr1_releasepri_20090330smr#story

    Second, let’s talk about Intel’s power management embedded in the chipset.  This component is the key to rapidly recouping power costs and maximizing your server consolidation efforts.  For a good introduction to Intel’s power management system for server power capping in the data center, take a look at Jackson He’s blog “Datacenter Dynamic Power Management – Intelligent Power Management on Intel Xeon® 5500”.

    http://communities.intel.com/community/openportit/server/blog/2009/03/31/datacenter-dynamic-power-management-intelligent-power-management-on-intel-xeon-5500

    Lastly, virtualization management software drives ROI but the challenge in the management of large virtual infrastructures is that there are no clear boundaries in terms of network, storage and datacenter management teams.  This needs to be defined as well as an emphasis on a holistic management approach or a “Service Management” approach.  We have to get beyond just monitoring the uptime or resource usage levels of virtual machines (VM) and physical hosts. Along with Intel’s announcement of our latest Xeon 5500, there have also been a number of new product announcements in the past two months.  From VMworld Europe 2009, we heard about vSphere 4.0 and Citrix Essentials for Hyper-V and at ManageFusion Symantec touting improved virtualization functionality and management with CMS/SMS 7.0 integrating Intel’s vPro functionality.

    Are the current products providing a holistic management approach with virtualization?

    Is it the right strategy to integrate power management with virtualization management?

    I’ve got my opinion on this, what’s yours?

    Mark

    Sometimes the next step up is a big one. The Intel® Xeon® processor 5500 series (formerly codenamed “Nehalem”) is one of those kinds of steps.

    Over the last few years 10 Gigabit has started to take off, but there have always been some negative mutterings: “Why do I need 10 Gigabit?”, “Why do we need this much bandwidth?” or “My server can’t support 10 Gigabit per second bidirectional traffic anyway.” Despite the volume of 10 Gigabit products shipped, there is still the reality that if you intend to use the entire 20 Gbps (10G both directions) or heaven forbid you try to use 40 Gbps with a dual port product; you will likely be sorely disappointed with the results.

     

    The reason for this is simple. Most current mainstream servers and 10 Gigabit products don’t support the intense usage models needed to drive that much network I/O and they also don’t have the memory architecture to unleash the full potential of dual 10 Gigabit links.

     

    Luckily, that all just changed with Intel® Xeon® processor 5500 series.

     

    In addition to the great processing improvements that the Intel® Xeon® processor 5500 series brings to the table, Intel has also introduced our third generation 10 Gigabit product, the Intel® 82599 10 Gigabit Ethernet Controller which provides two ports, and new capabilities and enhancement to the 10 Gigabit product landscape that help unshackle the new processor from its predecessor’s network I/O handcuffs and unleashes blazing performance in a variety of usage models. These improvements, coupled with the new architecture of the Xeon 5500 provide a symbiotic processor-networking combination that makes new usages possible and expands server and datacenter computing by a big leap… not just a baby step.

     

    One of the key changes with Intel® Xeon® processor 5500 series architecture is a step function improvement in the internal system I/O. The new local memory controller design, faster cache architecture, and support for DDR3 help push Xeon 5500 to be able to support peak memory bandwidth of ~32 Gigabytes, per socket. In a dual socket system this provides for ~64 Gigabytes of bandwidth which is dramatically more than the previous generation server configuration. In addition, the new Intel® QuickPath Interconnect (Intel® QPI) improves the speed both for inter-Processor communication as well as a faster path to the I/O hub. Finally, PCI Express* 2.0 I/O Bus support has been added to improve the entire data path from Processor to the 10 Gigabit Ethernet link.

     

    Taken together, the above improvements are a performance game changer for 10 Gigabit Ethernet.

     

    The chart below** shows the previous generation Intel® Xeon® paired with the previous generation Intel 10 Gigabit Ethernet Controller compared to the latest platform using the newest Intel silicon for both processor and networking. Not only is the performance better in 1-4 port configurations, but the performance scales dramatically better to above 50 Gigabits per second of total LAN throughput in a four port configuration vs. *just* 17 Gigabits on the previous generation! A complete platform architecture solution makes this huge improvement possible.

    82599 + Xeon 5500.jpg

    Now, it’s great that Intel® Xeon® processor 5500 series coupled with the Intel® 82599 10 Gigabit Ethernet can deliver such raw performance, but there is the forever nagging question of usage model. Luckily, the new headroom breathes new life into both Virtualization and storage over Ethernet usages (both of which I’ve talked about here and here) and provides new opportunities to more efficiently utilize your network link.

     

    Intel® Xeon® processor 5500 series allows the vision of consolidation in the datacenter to scale new heights, increasing the number of Virtual Machines (VM) that can effectively live inside a single system enclosure. Each incremental VM will add additional network I/O that is already starting to exhaust a 4 or 8 port single gigabit interface configuration with today’s server capabilities. As more VMs are added in the Xeon 5500 generation, 10 Gigabit will no longer be seen as optional; it will be required. For its part, the Intel® 82599 10 Gigabit Ethernet Controller supports Intel® Virtualization Technology for Connectivity (Intel® VT-c) to improve overall system performance in virtualized server environments. Intel VT- c includes hardware optimizations that help reduce I/O bottlenecks, boost throughput and reduce latency. Components of Intel VT-c include VMDq, and VMDc. VMDc consists of SR-IOV which I’ve covered before, and the ability to support VM mobility; a critical usage model for modern a IT deployment. All together, server systems can support more VMs, more throughput, more flexibility and better performance in a datacenter environment.

     

    Finally, the additional capabilities of the Intel® 82599 10 Gigabit Ethernet Controller product surrounding support for FCoE offloads and full support for the new Data Center Bridging (DCB) standards provide an opportunity for storage convergence over Ethernet in either a datacenter using a Fiber Channel SAN environment or an IT environment more focused on iSCSI. On the performance side of things, iSCSI acceleration along with FCoE data path offloads are supported in the Ethernet controller, and on the processor there is support for the CRC instruction set which insures iSCSI data integrity while minimizing processor overhead.

     

    The ability to converge at least part of the additional storage infrastructure onto Ethernet is just another factor driving massively increased data rates over Ethernet… luckily, the Intel® Xeon® processor 5500 series and the Intel® 82599 10 Gigabit Ethernet Controller solutions are up to the task.

    Over the past few days, there has been a lot of noise around Intel® Xeon® processor 5500 series and the many other platform components that help it shine its brightest. Improved processing power, memory controller bandwidth, faster and redesigned FSB, and improved 10 Gigabit networking all converge together to provide a fantastic performance, convergence, scalability and power story. Intel’s strong history in the server and processor markets, coupled with over 25 years in Ethernet makes this latest release a natural evolution of technology. Together these capabilities, along with the improved 10 Gigabit features and performance, are helping to transform the datacenter. It will be denser, more power efficient, more performant, and more consolidated with capabilities like FCoE and iSCSI.

     

    As for “Why do I need 10 Gigabit?” We have the answer, and it’s the new Xeon®.

     

    Ben Hacker

    -----

    ** Source. Intel. Mar 2009. Up to 2.5x performance compared to Intel® Xeon® processor 5300 series. Performance result of a bandwidth intensive network benchmark (IxChariot). Network throughput was measured on 64KB I/O size transfers between the test system and multiple network targets. Intel pre-production system with two Quad-Core Intel® Xeon® processor 5500 series CPUs (2.93 GHz), 12 GB memory (6 x2GB DDR3 - 1066MHz) vs. Intel Production system with two Intel® Xeon® processors X5365 (3.0GHz, 1333MHz FSB), 8 GB memory (8 x 1 GB DDR 2 - 667). Windows Server 2008, stock unmodified installation.

    Intel® has just launched their latest server processor, the Intel® Xeon® processor 5500 series. It really is a breakthrough processor for Intel and a clearly phenomenal solution for HPC. I was watching a keynote presentation this week and our Vice President was downright giddy about it. What makes this processor such a phenomenal solution for HPC? The answer is really easy; it expands capabilities and shortens users’ time to results. The real question is how does this processor perform so much better than other solutions out there? This answer is a bit more complicated but really fun to answer. Here we go…

    Intel® QuickPath Interconnect (QPI) – This is the technology that has replaced the front side bus used in previous generation Xeon® processors. Our previous generation architecture had a bandwidth of 21 GB/s vs. the QPI bandwidth of 46.1 GB/s. This is a speedup of 2.2X, very impressive. For applications that require lots of I/O this is huge. It’s like going from a country back road to an expressway!

    Integrated memory controller – Intel has moved the memory controller from the MCH (memory controller Hub) into the processor.  In addition to integrating the memory controller, Intel is now using native DDR3 with speeds up to 1333MHz and three memory channelsper processor; this is a total of 6 memory channels and 64 GB/s of total memory bandwidth for a 2S HPC node.  This is a 3x jump in memory bandwidth from theprevious generation memory controller which only supported speeds up to 1066MHz and 4 memory channels. By integrating the memory controller you are now in closer contact with the processor for lower latency reads and writes.  Intel added two additional memory channel (one per socket) to increase memory capacity and increase the speed to faster reads and writes. 

    Energy efficient design – The new Intel® Xeon® processor 5500 series has the dynamic capability of turning off cores when not required. There are more power states and has the ability to transition between power states faster than ever before. Net, net this means less power consumption. By consuming less power and providing world class performance Intel has created a solution that cries out HPC!

    By taking advantage of the power saving, Intel has introduced another feature called Intel® Turbo Boost Technology. Intel® Turbo Boost Technology automatically increases processor frequency to boost application performance if thermal headroom is available. Depending on the environment Turbo Boost can increase the processor frequency by as much as 400 MHz!

    Another technology supported in the Intel® Xeon processor 5500 series is Hyper-Threading. Intel® Hyper-Threading Technology enables users to run multiple threads on each processing core to increase total application performance while requiring only a fraction of the power that would be necessary to support additional cores. For highly threaded HPC applications this is showing performance gains over 25%.

    The Intel® Xeon® processor 5500 series is considered a general purpose processor. However, a closer look at the features and capabilities show that this is one heck of an HPC solution. You can’t help but think Intel knew HPC was an important market segment for servers and they had this in mind as they created the architecture and developed the features.

    Well, is Intel pounding their chest…again! They should be. The introduction of the Intel® Xeon® processor 5500 series is breakthrough architecture for HPC users. The industry hasn’t seen generation to generation performance gains like this since the Pentium® Pro was introduced back in the mid 90’s. Congratulations Intel and go ahead and pound that chest, you deserve it!

    Everyday people are facing exploding volumes of data that they need to manage. As the model size continues to grow, they need to figure out how to maximize the efficiency of data movement and where possible to move processing to data, rather than data to processing. I see this as a constant issue for most of my customers and therefore part of my job is to provide system benchmarks to help people understand how to choose the most efficient platforms for their data-intensive computations. I recently co-authored a technical white paper on Data Intensive Computing that will provide you a bit more insight on this topic. Feel free to download at: http://www.sgi.com/pdfs/4154.pdf.

    This week Silicon Graphics posted a large number of standard and application benchmarks for the SGI® Altix® ICE platform and the Intel® Xeon® Processor 5500 Series (codenamed Nehalem). You can find both the standard and application benchmarks on http://www.sgi.com/products/servers/altix/ice.

    As we ran the benchmarks we were able to achieve superior application performance through a combination of different factors, most important being the innovative memory system of the new Intel processor family and the improved ICE platform network topology and I/O. On the website you will find a variety of application benchmarks including MD.Nastran, LS-Dyna, Abaqus, Radioss, Fluent, Gaussian and NAMD just to name a few.

    Filter Blog

    By author:
    By date:
    By tag: