Posts tagged with server
Web 2.0 – A New Server World Order?
Posted by Patrick Moorhead in 4:55 PM
Loading ...You may have seen recent blogs from me on the topic of social networks and their growth. I even subjected my Computex blogs to these media. Why do we spend time with this stuff – we don’t sell software! In the end, it’s because trends in our industry are important for AMD, and personally satisfying to me to stay up to date on these. The creative minds in our industry continue to find new ways to use technology and it’s a challenge just saying current. But what’s equally surprising is the speed that the “under 30” crowd adopts these new media/social networks. This new social model is one of the emerging Internet use scenarios that has come to be known as Web 2.0.
I find a great deal of satisfaction in this fast growing market especially in light of some recent announcements that AMD made regarding the AMD Opteron processor’s web performance. AMD recently announced that Quad-Core AMD Opteron™ processor Model 2356 and Quad-Core AMD Opteron processor Model 8356 now hold the top x86 web performance records for both 2P and 4P servers, as measured by the SPECweb®2005 benchmark. This means that the AMD-based server architecture is able to process and manage more data over a period of time than competing solutions and therefore keep the media rich social networks operating at ever faster speeds. As I was writing this blog, I read a related blog from Nigel Dessau on the subject of workloads and balancing loads. We’ve spent a great deal of time and effort to enable and optimize these new networks but we did not stop there. I blogged recently on 2nd generation notebooks and highlighted the new features that we have enabled through our ATI acquisition including our UVD (universal video decoder) which is optimized to consume rich media such as web video.
AMD continues to focus on solutions. Just to be clear, what this means to us is power and performance optimized silicon, platforms optimized with partners and customers to deliver real benefits. These benefits are provided on the server side as a web backbone and on the notebook and desktop as power/performance optimized clients. So AMD is again leading and enabling major transitions in the industry by leveraging our unique capabilities to create, process, serve up and consume this new-fangled thing called Web 2.0.
Pat Moorhead is Vice President of Advanced Marketing at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
There are Benchmarks, and there are Benchmarks…
Posted by Patrick Moorhead in 10:08 AM
I’m a fan of benchmarks. I think they are very helpful in allowing consumers to make informed purchase decisions about products. But they generally have some flexibility built into them so you can focus on those elements you want. And this means you can use a benchmark to tell a number of stories – which means you can choose to tell the story you want.
For example, take a recent review by AnandTech entitled “Sixteen Cores, Four Sockets” published on June 17, 2008. This article featured Quad-Core AMD Opteron™ processor-based systems. One of the performance evaluations in this article was a SPECjbb2005 benchmark estimate. What is particularly interesting about this article is that the published estimates list the 4 socket server running AMD Opteron processors model 8356 as 25% faster than the competition while running at similar frequencies and 7% faster than the fastest competitive solution. These results vary widely from the official scores posted on the SPEC site. Now you might ask – how can that be? How can you run what is considered to be an industry standard benchmark and get a different set of numbers? That can’t be right!
Taking a closer look at the SPECjbb2005 benchmark helps to unravel this mystery. SPECjbb2005 is a memory-intensive benchmark that is intended to evaluate the performance of servers running typical Java business applications. Its results evaluate the interaction of the CPU, caches, memory hierarchy, JVM (Java Virtual Machine), and JIT (Just-In-Time) compiler. SPECjbb2005 can be configured to run in a variety of ways, resulting in different performance outcomes. Different configuration = different story. For example, you can get different results based on the operating system used, the version of JVM used, the level of optimization of the JVM and JIT, JVM tuning options, and thread allocations.
The SPECjbb2005 scores published by SPEC tend to be achieved using very aggressive software tuning and processor settings. These settings help to achieve a “best possible score” but do not necessarily reflect how a system would be configured in a data center environment to provide the most stable and efficient performance. The scores published in the AnandTech article, according to the author, are more likely to reflect real world configurations with optimizations used consistent over the different processor architectures.
Indeed – if you do a survey around the internet you can find reference to other SPECjbb2005 scores and estimates that reflect a variety of configuration options and the resulting differences in the benchmark scores:
http://blogs.sun.com/bmseer/entry/sun_fire_x4440_best_opteron
Blog featuring SPECjbb2005 results with the 4 socket Sun Fire x4440 running quad-core AMD Opteron processors with Solaris 10 and Sun JVM. Also highlights power consumption of featured systems – reminding us that in today’s economy of escalating energy costs raw performance has less meaning to data centers than performance/watt.
http://techreport.com/articles.x/13176/4
An article by TechReport featuring SPECjbb2005 estimates for 2 socket servers running quad-core processors with Windows Server 2003 x64 edition and the Sun JVM. The author states the goal of this performance evaluation was to test relative performance on equal footing.
Taking a closer look at the official SPECjbb2005 scores and the estimates published in the various articles, you can see how confusing a benchmark can be. This serves as a reminder to us that benchmarks are just an indicator of performance and that a benchmark like SPECjbb2005, which allows for a wide variety of configurations, can produce a wide variety of results. And remember – the story being told is not always the one that best reflects reality….
Pat Moorhead is Vice President of Advanced Marketing at AMD.His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
In Praise of Standardized Virtualization Benchmarks
Posted by Patrick Moorhead in 2:49 AM
I’d like to start off my blog today by extending my congratulations to my AMD colleagues on capturing the VMmark benchmark crown. Our CMO Nigel Dessau covers this achievement today in his blog here.
While the best way in my opinion for enterprises to compare hardware is to load their own data in their own datacenters, it is encouraging to see that “benchmarkship” is already entering the world of virtualization. Because doing real-world testing is difficult with virtualization, this is a good first step. And it’s interesting to see how benchmark developers are begining to develop guidelines around how to properly conduct these virtualization benchmarking activities. These developers are essentially telling the industry that they are going to insist that we standardize our testing methodology, so that we are comparing “apples to apples”. Personally, I think this is a great idea and I hope it will help ensure that meaningfull, truthful data is available to those who need it to make their virtualization purchase decisions.
Of course guidelines are meaningless if they aren’t enforced. That isn’t the case here. We’re seeing more examples of how benchmark developers are willing to enforce these guidelines. Adhering to guidelines is something that AMD learned the importance of firsthand in 2007. And we’re not the only ones. Recently the benchmarks for one of our competitor’s solutions had to be withdrawn from the VMmark results page due to “noncompliance”.
I also want to point out that it is possible to follow the guidelines and still not necessarily tell the whole story. For example, a customer recently pointed out to me that our competitor posted some system benchmarks, but somehow omitted any reference to the AMD-based Dell R905, even though benchmarks for this AMD system had been posted on the VMWare’s site for more than a month. Perhaps it was an oversight?
While real-world datacenter comparisons using real business data is the best way to evaluate systems, I beleive that as vendors we need to take the benchmark high road – we need to use benchmarks as one tool to help customers evaluate technology – so they can make wiser choices. We need to fairly and honestly benchmark systems and report all the pertinent data. Trust and credibility go hand-in-hand, and if we do not keep the process fair and open, then the hard work of many people within my company, at our customers and partners and even within our competitor is at risk. As a demonstration of telling the whole story, let me point you to the “complete” results of the VMWare benchmarks.
Pat Moorhead is Vice President of Advanced Marketing at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
Secret Find of the Family Network Administrator: HP MediaSmart Server
Posted by Patrick Moorhead in 12:14 PM
My name is Patrick Moorhead and I am vice president of advanced marketing at AMD. Most of what I focus on is the non-traditional marketing like developing the new strategies to change the ways we would like people to relate to AMD’s technologies, new ways to leverage our platforms into the marketplace and new ways to communicate them. I think a lot about why people think and behave the way they do and develop how AMD can meet the needs in the consumer and commercial space. On the consumer side, there is no better way to learn about consumer pain points than to “just do it,” and maybe even have a little fun along the way. I guess if that approach is good enough for Steve Jobs and Michael Dell, it has to be good enough for me. :>
I am a gadget guy, particularly on the home side, so I wanted to keep my first blog informative and light. In future blogs, I will be highlighting some of the industry’s misnomers or hype that need some light shed on them.
If you are also the Family Network Administrator (FNA), I want to let you in on a secret find I made over the the last few months: the HP MediaSmart Server. Normally I prefer to build my own rigs, so when I first heard about the MediaSmart server and its capabilities, I did a little bit of head-scratching. How wrong I was…….
I suppose “tech junkie” fully describes a household having 9 PCs, 3 DMAs (digital media adapter), 5 PMPs (portable media player), 4 DVRs (digital video recorder), 5TB storage, 3 switches, 1 router, an Xbox 360, 4 Nintendo DSs, a SlingBox and a 12Mbit internet connection. It is always a challenge to manage all of that technology and content, particularly keeping up with demands of the Moorhead User Group (MUG), already 4 members strong. MUG members include my wife and my three kids. The needs of MUG are far and wide, from crystal clear music to 15 years of pictures to the latest videos to 4 jitter-free game platform experiences to perfect internet connections …… all available 365 days a year, 24×7 in any room. Many of you can relate, I know, painfully, because you also need to manage all of that beautiful mess.
The HP MediaSmart Server simplifies and removes many of the mundane and often never implemented tasks like backups, network monitoring, and password and profile management. All of this stuff is automated and intuitive, unlike most consumer electronics and computer gear. The server also has some cool features like content sharing inside and outside of the house to your friends and family, and remote access to manage the network and PCs when the FNA is on a business trip. The HP MediaSmart server is powered by an AMD processor and is also part the AMD LIVE!™ family of products so you can really experience some AMD goodness, too!
Setup is easy. Attach the server to your wired network, load a CD on any PC you want connected, setup your passwords and you are on easy street. Now every PC and device in your house can share any of the same music, videos, movies, pictures and documents. Sounds easy and it is. It actually “works.”
It is kept safe and secure by password access control and with data replication. In other words, if you don’t want your 5 year old child to have rights to “delete” or “change” mom’s music files, you can restrict it. Complete or selective systems backups are done automagically without any effort but a mouse click. If data corruption is detected on the server, it will alert the FNA and because the data is replicated (written twice), it is safe and sound. The server will also alert the FNA if the backup was interrupted, anti-virus is out of date, malware detection is off, or if the firewall has been disabled.
The biggest benefit we experienced so far is the ability to have all of our music, video, and pictures in one place to “pull” from. Whether it’s one of the 5 iPods, Apple TV, Xbox 360, D-Link DSM-520, desktops, or laptops, they are all getting data from the same place, which means you are never hunting for that “one song that we downloaded on PC #5 that I can’t find anymore.” I am certain that when the first PC blows up and I can restore all of the data immediately, that will then be the next big feature I love. The other benefit is that the server can be left on all the time, and because it is very cool and quiet, it’s non-intrusive. Doing that on your main home PC is possible, but not optimal. The HP server hasn’t crashed once or turned off unexpectedly. It was built on Windows Server 2003 SP2, so you get years of real-world use before they sprung it on the consumer, as well as putting a happy face on it. And, no, I havent encountered any of the data integrity challenges written about.
Remote access was a surprise hit. When I was in Las Vegas recently and the “PC broke and couldn’t play music,” I logged into the PC at home in Austin, reloaded a piece of software, and got MUG happy again. I hope that boosts my customer sat score this month. I doubt it.
In my opinion, you don’t need to worry about losing the investment you may have made in USB or eSATA external storage. They plug right into the server and it automagically becomes part of the server’s storage capacity. Also, don’t worry about running out of storage capacity any time soon. The HP MediaSmart server comes in two configurations, 500GB ($599) and 1TB ($749). It comes with 2 extra SATA hard drive bays so you can add two of the highest capacity SATA drives you can find when you need it. As it comes to storage, I recommend only buying it when you need it because the price per MB decreases so rapidly. When I checked today, I could buy 1TB SATA drive on Newegg for $199, so it seems to me that, when you need more next year, it could be half that price or at least much cheaper.
The next feature I want to test is the ability to invite friends and family to view and even post their pictures to the MUG server. Sort of my own protected Flickr. I am not sure exactly the incremental benefit that provides versus the photo posting services, but we will see. For the daring FNAs out there who want to build their own server, Microsoft sells Windows Home Server, the OS that HP has built their own server upon. Some of my compadres have built a beast of a server on this platform and it takes what HP has developed and extended the functionality to things like encoding video and audio content. HP has kept it simple, focused and didn’t overextend themselves in the spirit of maximizing the feature list. What it does, it does well.
I would love to hear from my fellow FNAs managing their home technology installations and how you deal with the complexity and management.
I’m a fan of benchmarks. I think they are very helpful in allowing consumers to make informed purchase decisions about products. But they generally have some flexibility built into them so you can focus on those elements you want. And this means you can use a benchmark to tell a number of stories – which means you can choose to tell the story you want.
For example, take a recent review by AnandTech entitled “Sixteen Cores, Four Sockets” published on June 17, 2008. This article featured Quad-Core AMD Opteron™ processor-based systems. One of the performance evaluations in this article was a SPECjbb2005 benchmark estimate. What is particularly interesting about this article is that the published estimates list the 4 socket server running AMD Opteron processors model 8356 as 25% faster than the competition while running at similar frequencies and 7% faster than the fastest competitive solution. These results vary widely from the official scores posted on the SPEC site. Now you might ask – how can that be? How can you run what is considered to be an industry standard benchmark and get a different set of numbers? That can’t be right!
Taking a closer look at the SPECjbb2005 benchmark helps to unravel this mystery. SPECjbb2005 is a memory-intensive benchmark that is intended to evaluate the performance of servers running typical Java business applications. Its results evaluate the interaction of the CPU, caches, memory hierarchy, JVM (Java Virtual Machine), and JIT (Just-In-Time) compiler. SPECjbb2005 can be configured to run in a variety of ways, resulting in different performance outcomes. Different configuration = different story. For example, you can get different results based on the operating system used, the version of JVM used, the level of optimization of the JVM and JIT, JVM tuning options, and thread allocations.
The SPECjbb2005 scores published by SPEC tend to be achieved using very aggressive software tuning and processor settings. These settings help to achieve a “best possible score” but do not necessarily reflect how a system would be configured in a data center environment to provide the most stable and efficient performance. The scores published in the AnandTech article, according to the author, are more likely to reflect real world configurations with optimizations used consistent over the different processor architectures.
Indeed – if you do a survey around the internet you can find reference to other SPECjbb2005 scores and estimates that reflect a variety of configuration options and the resulting differences in the benchmark scores:
http://blogs.sun.com/bmseer/entry/sun_fire_x4440_best_opteron
Blog featuring SPECjbb2005 results with the 4 socket Sun Fire x4440 running quad-core AMD Opteron processors with Solaris 10 and Sun JVM. Also highlights power consumption of featured systems – reminding us that in today’s economy of escalating energy costs raw performance has less meaning to data centers than performance/watt.
http://techreport.com/articles.x/13176/4
An article by TechReport featuring SPECjbb2005 estimates for 2 socket servers running quad-core processors with Windows Server 2003 x64 edition and the Sun JVM. The author states the goal of this performance evaluation was to test relative performance on equal footing.
Taking a closer look at the official SPECjbb2005 scores and the estimates published in the various articles, you can see how confusing a benchmark can be. This serves as a reminder to us that benchmarks are just an indicator of performance and that a benchmark like SPECjbb2005, which allows for a wide variety of configurations, can produce a wide variety of results. And remember – the story being told is not always the one that best reflects reality….
Pat Moorhead is Vice President of Advanced Marketing at AMD.His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
(4.43 out of 5)