Scalabiliti blog

4K Resolution Video to Push Computing Boundaries

Mon, 14 Feb 2011 06:33:20 -0700

Youtube has provided us with a glimpse of the future on their [ demo page] with 4k resolution video files. In order to see them select “Original” above the 1080p mode in the settings menu. You wont be able to really enjoy the 4K resolution since there is hardly any monitors available yet that will render 4K resolution.

6 SSD Drives in Parallel - LSI Warp Drive spotted in the wild

Fri, 17 Dec 2010 06:16:46 -0700

The LSI WarpDrive which consists of a pci express board made of six SSD drives that are run together in parallel (RAID 0). All the six solid state disks are based on the SandForce SF-1564 SSD processor and feature 32nm SLC NAND flash memory chips. As a result, LSI's PCIe SSD offers 300GB of storage space as well as some impressive performance numbers, the company stating the WarpDrive has a latency of 50 microseconds, MTBF of 2,000,000 hours, maximum continuous read and write speeds of 1400MB/…

800 Core Server Arriving mid 2011

Wed, 29 Sep 2010 02:42:26 -0700

Tilera to launch 800 processor core server Tile-GX100 RISC chips run at 1.5GHz A server with 800 processor cores could ship as early as next year when Tilera lauches its next generation 100-core chips inside. The new Tile-GX100 are made for 64-bit computing. Each core runs at up to 1.5GHz and include 32MB of aggregate cache. The Tile-GX100 will be able to run a Linux OS, and will support applications written in Java, C or C++ languages.

AMD's Llano A8-3800 gets bench-marked

Wed, 15 Jun 2011 07:30:17 -0700

AMD's Llano A8-3800 gets bench-marked AMD's Llano technically A8-3800, sports 4 cores at 2.4 GHz where each core has 1 MB L2 cache 400 Radeon cores at 600 MHz. This chip is continues AMD's fusion of graphics(GPU) and CPU with nearly 50% of the space on this processor taken up with the GPU. This doesn't make for a data intensive capable processor but gives good graphics performance for a budget price. For compute performance is remains to be seen if the GPU cores can be used effectively for…

British Company Dezineforce offers High Performance in the cloud or in a box

Fri, 03 Dec 2010 11:04:52 -0700

Dezineforce, which has its roots in the University of Southampton, aims to bring high performance computing to the UK and Europe through two offerings. It is offering a cloud based option as well as a 'black box' appliance to be hosted within the business.

IBM to help China build a city for Cloud Computing

Mon, 14 Feb 2011 01:27:47 -0700

IBM will be one of the companies help China to build a city-sized cloud computing and office complex that will include a mega data center, The entire complex will cover some 6.2 million square feet, with the initial data center space accounting for approximately 646,000 square feet. This allows China to have the economics of scale as well as centralized control while still offering could resources.

China leads the Supercomputer Race

Thu, 28 Oct 2010 02:11:03 -0700

China has at last surpassed the US with the world's fastest supercomputer. The Tianhe-1A achieved 2.507 Petaflop, surpassing Cray's 2.3 petaflops Jaguar, built at the Oak Ridge National Lab in Tennessee. An important part of this news was the design of the system. As with many recent Supercomputer efforts, designers moved away from either custom CPU's or 10's of thousands of generic CPUs. Instead the designers concentrate on melding traditional CPUs and NVIDIA's Tesla M2050 compute GPUs. This …

China's Latest Supercomputer Online and Probing the Secrets of the Universe

Mon, 13 Jun 2011 09:57:50 -0700

China's Latest Supercomputer Online and Probing the Secrets of the Universe The Chinese Academy of Sciences (CAS) announced that the Tianhe-1A GPU supercomputer is up and running and investigating complex molecular dynamics (MD) simulation. The 7,168 NVIDIA Tesla GPUs are being used to achieve 1.87 petaflops per second. The scientists simulating structure of crystalline silicon which is used in solar panels and other electronics.

China's Next Generation Super Computer will use a Chinese Processor

Mon, 14 Mar 2011 06:18:40 -0700

Not resting on already having the worlds fasted computer. China is pushing ahead with their next generation supercomputer the Dawning 6000. Expected to arrive this summer and this system will use thousands indigenous Loongson processors for a total capacity of 1 petaflops.

China's $10Bn strategic move to create their own x86 processors

Sun, 27 Feb 2011 04:02:12 -0700

With Intel's eyes mostly focused on the mobile arena. China is moving strategically to seize some of the server processorhigh ground. Latest IEEE presentations shows China is developing an x86-emulating Godson line of 64-bit processors? Godson chip effort is one of 16 different projects, in fact, that are each funded with between $5bn and $10bn.

Chrome 7 is to be 60 times faster

Thu, 16 Sep 2010 02:42:48 -0700

Google has claimed that its forthcoming Chrome 7 browser will be 60 times faster than the previous version of the software. The Google Chromium blog is showing off its benchmark results with the browser's new 2D graphics acceleration.

LMU Munich - Cluster Experimentation with Apple Tv

Tue, 12 Apr 2011 13:56:38 -0700

LMU Munich - Cluster Experimentation with Apple Tv High Performance, multi core multi cluster experimentation can have high hardware costs and prove a serious barrier to entry. Enter the Munich University Apple TV Cluster project that aims to share and discuss software related to building collaborative, distributed computing using low cost low power Arm based processors.

4 Cool Super Computing Technologies

Tue, 29 Jun 2010 11:37:59 -0700

Perhaps not the sexiest side of super computing but increasingly power and heat is it's biggest problem. Here are 4 new takes on reducing power and heat in the server room. IBM individually cooled server doors 1. IBM Rear Door Heat eXchanger A Rear-door heat exchanger is fitted to the back of this rack of servers. The coils circulate cold water that is cooled by the building's existing chillers. The liquid-cooled panel on the back of the unit can eliminate the need for computer-room air co…

Disposable HPC Nodes

Sat, 29 May 2010 01:48:13 -0700

Via our Twitter feed, we read an interesting article by Douglas Eadline, Ph.D. In the article Douglas makes the argument, based on a linked research paper, that the use of lower cost, lower power nodes would enable an organisation to move to a Recovery Orientated Computing (ROC) platform.

EU Digital Agenda for 2020 - 30Mbit/s and unified digital services for everyone

Thu, 27 May 2010 03:43:21 -0700

As anyone working across Europe will tell you some pan European initialives like GSM has been a great boost to interoperability and the European technology industries. Now the the European Commission has created a concrete Digital Agenda. A five-year that covers speed and access to very fast broadband, creating a digital single market and a rapid response system for cyber-attacks.

European Open Filesystems for Supercomputing

Tue, 21 Dec 2010 00:49:59 -0700

A European Open Filesystem (OSF) was kicked off at the headquarters of ParTec Center GmbH in Munich, Germany. The purpose of this enterprise is to promote the establishment and adoption of an open source parallel filesystem, sustain and enhance its quality, capabilities and functionality and to ensure that the specific requirements of European organizations, institutions and companies are considered.

Benchmarking the JVM with JBenchmark

Thu, 03 Jun 2010 02:56:26 -0700

Attempting to benchmark is akin to reglious heracy where web fanboys will tie you to the metaphorical stake and flame war you you to ashes. Undaunted this blog disucsses the use of java benchmark framework that helps provide consitant reproducable benchmarks on the JVM.

Google+ scales using Google closure for Javascript (no sign of GWT)

Fri, 15 Jul 2011 06:58:18 -0700

Google+ scales using Google closure for Javascript (no sign of GWT) Joseph Smarr, a technical lead on Google+ spoke regarding some of the network’s technical details in an AnyAsk interview. According to Smarr, Google+ uses Java with Guice on the server side and the JavaScript relies heavily on the open source cross-browser Closure framework. Closure is a set of tools for JavaScript including a templating framework (soy), a JavaScript “compiler” which is really a JavaScript optimizer that …

GPU Network router delivers breakthrough performance

Fri, 03 Sep 2010 03:57:55 -0700

We've been covering GPU's for some time. They are turning up in all sorts of interesting high performance areas. In a new twist, researchers in South Korea have now built a GPU based networking router that can transmit and interrogate data at record speeds.

GPU performance can it now Conquor the Database?

Wed, 22 Dec 2010 05:43:38 -0700

GPU programming gained ubiquity in 2010 with nearly all new super computing complexes introducing full or partial gpu solutions. Now we are seeing the reach of GPU computing extending into other areas including the Database/Data Warehousing and Data Mining. In addition new conference has been launched including the first ADBIS workshop on GPUs In Databases September 19th, 2011, Vienna, Austria. The topics discussed include

GPU Processing goes mainstream

Sun, 23 May 2010 00:18:26 -0700

For years, GPU processing (the use of high performance video cards for general purpose computing) has been the preserve of academic and niche commercial research. As the tools have improved, the march towards this technology has become more pronounced.

Hadoop is Big, I mean really BIG

Sat, 17 Jul 2010 10:04:57 -0700

Hadoop, for those not up on it, is a top level Apache project. That means it's right up there with the Apache Web Server in terms of importance. Hadoop allows you to distribute data-intensive applications across many nodes. These nodes can scale across many machines and sift through peta-bytes of data very quickly.

High-Frequency Trading Is 77% of U.K. Market

Wed, 26 Jan 2011 05:07:36 -0700

Source: Bloomberg (January 24, 2011) High-frequency trading accounts for 77 percent of transactions in U.K. markets, according to a study by research firm Tabb Group LLC. Orders from long-only funds that bet stocks will rise, hedge funds and retail investors account for 23 percent of activity in continuous markets, the group said in a report today. High-frequency trading, in which firms may transact thousands of times a second, accounts for the rest. The practice makes up 35 percent of the 3.9…

HPC FlashForward - Nvidia, Mellanox and Intel

Tue, 01 Jun 2010 02:19:49 -0700

[Die shot of Aubrey Isle silicon. Aubrey isle is the codename of the silicon chip included the 'Knights Ferry' Intel Many Integrated Core (MIC) architecture development platform.] We've been a great fan of the FlashForward series, which finished it's current series in the UK last night. Trying to guess what the writers have in store for us, having already told us, is an interesting challenge.

IBM and Revolution team up to create a (R)evolution in the Database using the "R" ...

Tue, 15 Mar 2011 04:11:38 -0700

[Netezza Appliance] Revolution Analytics and IBM Netezza have announced a partnership to integrate Revolution R Enterprise and the IBM Netezza TwinFin Data Warehouse Appliance. As a result customers will be able to run performance and full-scale predictive analytics from within a data warehouse platform will be able to directly leverage the power of the open source R statistics language. The companies are working together to create a version of Revolution’s software that takes advantage of IBM …

Intel Not Going to Quit the GPU Acceleration Race

Fri, 03 Dec 2010 11:07:47 -0700

Nvidia has been making significant in-roads into the high performance supercomputing market, making the GPU acceleration almost implicit in the latest generation of it GPU's. By contrast, Intel has struggled to land the same sort of dominance it enjoys in the processor market in the GPU market.

Is AMD's Fusion Processor a Performance Game Changer?

Tue, 01 Feb 2011 04:20:00 -0700

AMD has recently launched the Fusion' processors incorporate a graphics card (GPU) and microprocessor (CPU) on a single low power chip. These Accelerated Processing Units (APUs) utilise a single-die design that combines multi-core CPU (x86) the graphics and parallel processing engine. They also include a high-speed bus that transmits data across processor cores.

Is Arm the new black - Calxeda to offer a 480-core ARM Server

Thu, 17 Mar 2011 10:17:41 -0700

Arm is popping up in all types of smart devices but it is not commonly seen in high-end servers. However Calxeda 480-core, low power server in the works. With energy and space becoming critical data center criteria, Calxeda can claim a factor of 10 reduction in costs and a 5x, or even 10x performance gain over what is currently available

Jato - an open source JVM aiming at better performance

Tue, 25 May 2010 02:02:59 -0700

Jato is an open source JVM developed by Pekka Enberg, Tomek Grabiec Arthur Huillet and Eduard-Gabriel Munteanu. That aims at better performance especially at start time Early versions of Jato are available in 32-bit x86 versions of Linux (the processor will handle SSE2). Unlike the (hot)spot compiling of the Sun jvm, Jato provides compilation on the fly ( just-in-time ) for the whole of the code. This provides good performance even at lauch of the Java application. Although cited ad a boost f…

Java (JVM) Processor Affinity - A Quick Hack

Tue, 29 Jun 2010 14:37:21 -0700

With muti-processor/muticore systems now ubuitious it is unfortunate that Java does not provide JVM launch with core or processor affinity. Most operating system now have commands to support this feature: * procaff on Windows (2000, Vista, 7) * taskset - on Unix, Linux

JPMorgan Chase - On Demand Risk Analysis - amazing use of FPGA's

Thu, 14 Jul 2011 00:41:03 -0700

JPMorgan Chase - On Demand Risk Analysis - amazing use of FPGA's JPMorgan Chase has a massive amount of data to analysis on a constant basis, especially around risk and fraud detection. In the presentation below, they take us through their reasoning and usage of FPGA's, over other technologies such as GPU's.

JVM Options and their Performance Implications

Mon, 17 May 2010 06:33:15 -0700

JVM Options and their Performance Implications Dozens of JVM options can affect performance and benchmarking. Some interesting ones are: * Type of JVM: server (-server) versus client (-client). * Ensuring sufficient memory is available (-Xmx).

Free Live Financial Pivot Tables for Stocks and Indexes using Google Finance

Mon, 28 Jun 2010 09:02:06 -0700

[Google Spreadsheets showing live data] Live data feeds (in general) cost money. Bloomberg Terminals costs are far beyond most non-professional investors' means. But slowly a data revolution is starting, with Google leading way. And the aim is to create the Poor Man's Trading Terminal.

London Stock Exchange (LSE) smashes trade speed record running on Linux

Thu, 21 Oct 2010 01:19:57 -0700

The London Stock Exchange's new Linux-based system is delivering record breaking networking speed, with 126 microsecond trading times. The 126 microsecond speed is “twice as fast” as its main international competitors, the London Stock Exchange said. BATS Europe and Chi-X, two dedicated electronic rivals to the LSE, are reported to have an average latency of 250 and 175 microseconds respectively. Netiher company immediately provided details. But many of the LSE’s older and more traditional ri…

Microsoft tries to perk up their cloud offerings with Java

Mon, 01 Nov 2010 06:49:17 -0700

In an attempt to add some legitimacy to their lack luster cloud offering, Azure, Steve Ballmer in Microsoft’s Professional Developer keynote speech heaped praise on Java and outlined new supports for the platform in Azure. From our perspective such an announcement should be treated with skepticism, especially with a lack of clarity from Oracle on the future direction of Java. Combined with a hostile attitude to the open source java community, perhaps this is a good time for Microsoft to be ma…

Model Mirage Beware the new Technocracy

Sun, 05 Oct 2008 03:43:21 -0700

With rumours that the recent dramatic drop in the market was triggered by [ fat fingers] and computer based trading, I felt it might be time for a history lesson. Sometimes, remembering that some members of the technical community tend to over sell their worth and under sell the risks can provide a fresh perspective.

MongoDB and Morphia Performance

Tue, 25 May 2010 02:48:26 -0700

Internally we've been looking at using for a number of projects. Of course one of the areas of concern when moving to to a NoSQL data store is performance and accessibility. To this end, I started looking at how best to persist our Java Objects in MongoDB. There were a number of options to consider:

MongoDb vs MySql - Fight !!

Fri, 21 May 2010 08:05:51 -0700

MongoDb vs MySql - Fight !! Relational databases, and to a degree MySQL, have become the backbone of commercial organisations. Programmers and database developers have been engaged in a tug-of-war, deciding the best place to store and execute the business logic.

Morphia and Mongo Revisited

Fri, 15 Oct 2010 14:56:18 -0700

Back in May, we took at look at a promising Mongo DAO called Morphia. At the time, we concluded that it may not be ready for prime-time, but as with all things (hopefully) things improve. We had sometime this week to revisit the results, hook up the newest versions and try again.

Munich to switch on Europe's fastest supercomputer

Wed, 22 Dec 2010 09:51:22 -0700

Munich will soon become the proud new owners of Europe’s fastest supercomputer. The Leibniz Supercomputing Centre at the Bavarian Academy of Sciences and Humanities in Munich has announced the order of the SuperMUC, a massive supercomputer that will run at 3 petaflops - at least twice as fast as the current European champion.

New High Performance Real-Time Data Streaming Protocol

Mon, 18 Apr 2011 00:50:08 -0700

New High Performance Real-Time Data Streaming Protocol STREST is a new open source protocol and server from Wiredset, STREST is HTTP-compatible and is designed for real-time data streaming. The protocol was developed as part of creating the the real-time social media analytics service Trendrr. The protocol offer extremely high-volume API calls with low latency, deliver the results in real-time at scale, and do so through a RESTful interface.

New Linux Kernel brings Ceph - A distributed, scalable cloud filesystem

Sat, 22 May 2010 04:21:09 -0700

Ceph PetaByte Storage Linux kernel 2.6.34 adds a scalable Ceph filesystem It is a distributed network filesystem that is designed for massive scalability, capable of managing petabytes of storage. The new filesystem is design for distributed storage from the ground up. Some of the highlights includ

New NoSQL entrant MemBase .. looks hot !!

Sat, 17 Jul 2010 10:08:49 -0700

As regular readers, and our customers will know, we've been quite smitten with NoSQL databases, and MongoDB in particular, for quite some times. The combination of speed and flexibility perfectly matches the type of problems we solve on a daily basis. SQL DB's such MySQL, Oracle or MS SQL would provide limited scaling in these sorts of problems.

New presentations around Scalability from Fokus 2010

Tue, 21 Dec 2010 00:51:15 -0700

* NoSQL, One size fits all: A concept whose time has come and gone Nati Shalom, GigaSpaces Major Internet companies, including Google, Amazon and Facebook, are taking an alternative approach to the traditional SQL database. The primary reason is that the existing SQL database doesn?t fit their scalability requirements. Does this mark the end of the relational database era?

NoSQL Merger - CouchOne + Membase = CouchBase

Wed, 09 Feb 2011 13:45:52 -0700

Exciting new yesterday, as it was announced that two big open source NoSQL heavyweights had decided to join forces. Both teams showed that they were each dominant in seperate but complimentary areas of the database space. We can really see the synergies from where we sit, having worked with a number of NoSQL solutions for a while.

SGI's aggressive push breaks the boundaries of JVM performance

Mon, 02 Aug 2010 01:05:55 -0700

SGI showed of their brand new Altix UV 1000 system, letting it be known that they have broken the current record for performance on the Standard Performance Evaluation Corporation’s SPECjbb2005 benchmark, a standard measurement of Java-based application performance.

Supercomputers and Medicine

Mon, 09 May 2011 02:31:37 -0700

Supercomputers and Medicine EarthSky had an interesting podcast looking at just one aspect of the emerging bio-informatics and the application of supercomputing in individual genetic mapping and eventual personalized medicine. supercomputing, medicine, podcast, gentics, cloud

T-Platforms - From Russia with High Performance Love

Wed, 22 Dec 2010 01:25:09 -0700

Russian scientific excellence has always been world leading, despite limited resources. Now those resources are getting a boost from T-Platform, a Russian provider of state-of-the-art high performance computing. T-Platform was responsible for building the Lomonosov supercomputer (pictured right) which is the most powerful HPC system in Eastern Europe, delivering 420 TFlops at peak, and 350 TFlops of sustained performance under Linpack. Lomonosov is a currently #13 on the TOP500 list of the world…

The Continuing March of SSD

Wed, 06 Apr 2011 03:38:45 -0700

The Continuing March of SSD [200px] Intel today announced the refresh of it's SSD range of solid-state storage drives. The X25, the original Intel SSD has bit the dust, replaced by the new MLC-based drives. Of course, Intel haven't stood still, and as well as increasing capacity and speed (the drives are now based on the 25nm fab process), they have also managed to drop the price.

The Gigahertz war is over ..nobody told IBM's new 5.2Ghz processor

Fri, 27 Aug 2010 01:31:09 -0700

The Z-series mainframes are going to get the option of the new Z196 processor that has the surprising clock speed of 5.2GHz. The Z196 is a 45nm PD SOI chip which adds up to almost one and a half billion transistors. It also carries a 64KB Level 1 instruction cache, 1.5MB of private L2 cache per core, a 128K Level 1 data cache and even a pair of co-processors dedicated to doing cryptographic work.

The Long Road Back - IBM valued more than Microsoft for the first time in 15 years

Tue, 31 May 2011 02:57:40 -0700

The Long Road Back - IBM valued more than Microsoft for the first time in 15 years In an amusing aside it was interesting to IBM's capitalization over taking Microsoft for the first time in 15 years.On the 25th of May IBM briefly passed 205.2 billion dollars to Microsoft's 203.82 billion.

The Quants By Scott Patterson - Part 1

Fri, 14 May 2010 14:34:04 -0700

With the crash still fresh, and the financial archeology processing, I thought it was be interesting to read someone else's take on what happened. To that end, I've just picked up a copy of Scott Patterson's new book, The Quants, subtitled “How a New Breed of Math Whizzes Conquered Wall Street and Nearly Destroyed It”.

Twitter Boosts Search Performance 300% Drops Ruby on Rails for Java Lucene

Wed, 13 Apr 2011 01:11:26 -0700

Twitter Boosts Search Performance 300% Drops Ruby on Rails for Java Lucene Twitter search serves over one billion queries per day. The unfortunate tsunami in Japan caused such search query load as users experienced significant delays. About the same time Twitter released their latest search technology (Blender) based on Java/Lucene. As a result Twitter were able to see a 300% drop in most search times and significantly the CPU load on front-end servers was cut by 50%.

UIUC University Release Parallel Java Tools

Tue, 21 Dec 2010 01:05:16 -0700

Researchers at the University of Illinois at Urbana-Champaign (UIUC) have released a tool ( DPJizer) to facilitate the development of Java applications taking advantage of multicore processors. The program analyzes the code and automatically generate annotations needed to synchronize the tasks running in parallel. According to researchers, this method may be more accurate than manual annotations.

Uk Cloud Company Flexiant helps power the New Pan European Cloud

Fri, 03 Dec 2010 11:10:09 -0700

Pan European standards like GSM have later become standards around the world and help greatly boost European technology companies. Now the European Commission is formulating plans for a pan-european grid standard that will provide on-demand computing to those that need scale and computing power, but don't necessarily want or need to have it in-house.

UK Introduction to an Msc in High Performance Computing

Mon, 21 Mar 2011 01:53:19 -0700

UK Introduction to an Msc in High Performance Computing Dr David Henty, Programme Director of the MSc in HPC. Recorded at the Postgraduate Virtual Open Week 2011 at The University of Edinburgh. The MSc in High Performance Computing is a one-year postgraduate masters course taught by EPCC at the University of Edinburgh.

UK public services ‘too slow’ on cloud adoption

Thu, 10 Mar 2011 07:17:11 -0700

Former CIO for UK government, John Suffolk has suggested public services could be ‘losing billions’ because politicians and parliament strategists are not reacting quickly enough to the potential of cloud computing. In an online debate hosted by Cisco ( here )and also involving David Wild, current CIO for Westminster City Council, and Andy MacLeod, Cisco’s Head of Public Sector strategy, Suffolk stated that the public sector was making unnecessary redundancies and missing out on the possibili…

US Government jumps into the cloud - will the UK follow?

Thu, 03 Jun 2010 02:55:44 -0700

Government IT projects are not known for agility or cost savings. So it is with some surprise that U.S. government agencies have already leapt into the cloud. The primary motivation is increase efficiency, improve security and most importantly considering the current US economy is to save money.

US Supercomputers are being used to police the Internet

Wed, 22 Dec 2010 01:22:26 -0700

Oak Ridge National Laboratory researchers are using the Jaguar supercomputer to analyze Internet traffic, looking for clues that will lead law enforcement officers to sources of illegal content, including child pornography. Oak Ridge has devised algorithms that analyze traffic by studying search terms used on file-sharing networks. The project is planned to be in operation for 12 months and has been apportioned 1 million processor hours on the Jaguar system.

Viking Modular Fusion Memory and SSD Aims to Shake up How we See Memory

Wed, 13 Apr 2011 05:08:29 -0700

Viking Modular Fusion Memory and SSD Aims to Shake up How we See Memory Viking Modular is showing of a new SSDs that plugs into DDR3 memory slot. [ ] SSDs will be optimal for compact or blade servers where space is limited. The memory wont beat top end memory speeds but the not too shabby 520 MB/s performance will be good in a server environments. With the potential to offer far greater memory sizes at lower costs as well as the intriguing option of having all memory persisted.

VMWare challenges Google and Microsoft

Thu, 14 Apr 2011 05:26:17 -0700

VMWare challenges Google and Microsoft [250px] We've always appreciated VMWare's solutions, especially since we moved to OS X and found VMWare Fusion. ESX was always the benchmark of enterprise virtulisation, but never really helped our customers, who were into transparent, large systems.

vSMP Foundation Challenges More Expensive Symmetric Multiprocessing (SMP)

Tue, 31 May 2011 16:12:03 -0700

vSMP Foundation Challenges More Expensive Symmetric Multiprocessing (SMP) ScaleMP vSMP Foundation aims to allow you to combine multiple commodity (read cheap) x86 servers into one larger powerful virtual Symmetric Multiprocessing (SMP) System. The vSMP Foundation has emerged as a very interesting alternative to the often proprietary solutions that are available.

Who saw AAPLs market cap exceeding MSFT?

Fri, 28 May 2010 05:53:00 -0700

Predctions when looking backwards are rarely wrong, unless you want them to be. With that in mind, Apples' eventual overtaking of Microsoft in the market capitalization stakes was seen enviable by many. But was there any indication that is was so pre-determined?