Hosting Blog v2.03

Microsoft on Tuesday hit another high-performance computing milestone by placing its server for the first time in the top 10 on the list of the Top 500 super computers as judged by Top500.org. 

Just a year ago, the best Microsoft could do was 116th place based on rankings from Top500.org, which has been benchmarking supercomputers since 1993 with its bi-annual tests it calls “runs.” 
Download the latest Network World Executive Guide - Virtualization Reality Check

Windows HPC Server 2008, a 64-bit system that shipped Nov. 1, came in at No. 10, achieving 180.6 teraflops with 77.5% efficiency at the Shanghai Supercomputer Center and Dawning Information Industry Co. 

Despite the high ranking, Microsoft’s biggest high-performance computing challenge is likely in front of the vendor – creating easy-to-use developer tools for writing applications for the platform.

 The official launch of HPC 2008 Server will be September 22, blogger Mary-Jo Foley reports. It will launch as part of the 2008 “High Performance on Wall Street” conference in New York.

Hard to believe Microsoft needs yet another version of its flagship server operating system but so be it. This variant went into its final “beta” in June, 2008, and has been mostly overlooked in terms of hype, says blogger Alex Lewis. Features include more management tools, job scheduling and built-in clustering … or for a full list view the chart below. 
Feature 
Implementation 
Benefit

Operating system 
Windows Server 2008 
Inherits security and stability features from Windows Server 2008.

Processor type 
x64 (AMD64 or Intel EM64T) 
The large memory model and processor efficiencies of x64 architecture.

Node deployment 
Windows Deployment Services 
Image-based deployment, with full support for multicasting.

Head node redundancy 
Windows Failover Clustering and SQL Server Failover Clustering 
Requires Windows Server 2008 Enterprise and SQL Server Standard, but provides a fully redundant head node and scheduler.

Management 
New Integrated Administration Console 
Integrates all aspects of node and job management, grouping, monitoring at-a-glance, diagnostics, and reporting in a single application.

Network topology 
Network Configuration Wizard 
Fully automated, with a more intuitive interface, the Network Configuration Wizard facilitates configuring the cluster’s network topology.

Application network 
MS-MPI 
New, high-speed application network stack using NetworkDirect drivers. New shared memory implementation for multicore processors. Highly compatible with existing MPICH2 implementations.

Scheduler 
Job Manager Console 
GUI console is integrated into Administration Console, or is stand-alone. Command line supports Windows PowerShell scripting and legacy command-line scripts from Windows Compute Cluster Server (the first version of Windows HPC offering). Greatly improved speed and scalability. Support for SOA applications.

Reporting 
Integrated into Administration Console 
Standard, prebuilt reports. Extensibility features allow using SQL Server Analysis Services for additional analysis. Historical performance charts.

Monitoring 
Integrated into Administration Console 
Heat map on cluster or node group. Per node charts. Cluster-wide performance overview.

Diagnostics 
Integrated into Administration Console 
In-the-box verification tests and performance tests. Store, filter, and view test results and history.


Source: Microsoft

In June, Alex Lewis lamented HPCs lack of software support, but Microsoft marketing materials list about 30 HPC partners when including hardware and software vendors that have been on board with Microsoft Windows Compute Clusters Server 2003. The assumption is that these same makers will offer support of the WS2008 version of HPC, though it is difficult to find any materials which document who, what, when, why. More news on that will no doubt come during the New York gala. Stay tuned.

The Linpack Benchmark

As a yardstick of performance we are using the `best’ performance as measured by the LINPACK Benchmark. LINPACK was chosen because it is widely used and performance numbers are available for almost all relevant systems.

The LINPACK Benchmark was introduced by Jack Dongarra. A detailed description as well as a list of performance results on a wide variety of machines is available in postscript form from netlib. Here you can download the latest version of the LINPACK Report: performance.ps. A parallel implementation of the Linpack benchmark and instructions on how to run it can be found at http://www.netlib.org/benchmark/hpl/.

The benchmark used in the LINPACK Benchmark is to solve a dense system of linear equations. For the TOP500, we used that version of the benchmark that allows the user to scale the size of the problem and to optimize the software in order to achieve the best performance for a given machine. This performance does not reflect the overall performance of a given system, as no single number ever can. It does, however, reflect the performance of a dedicated system for solving a dense system of linear equations. Since the problem is very regular, the performance achieved is quite high, and the performance numbers give a good correction of peak performance.

By measuring the actual performance for different problem sizes n, a user can get not only the maximal achieved performance Rmax for the problem size Nmax but also the problem size N1/2 where half of the performance Rmax is achieved. These numbers together with the theoretical peak performance Rpeak are the numbers given in the TOP500. In an attempt to obtain uniformity across all computers in performance reporting, the algorithm used in solving the system of equations in the benchmark procedure must conform to LU factorization with partial pivoting. In particular, the operation count for the algorithm must be 2/3 n^3 + O(n^2) double precision floating point operations. This excludes the use of a fast matrix multiply algorithm like “Strassen’s Method” or algorithms which compute a solution in a precision lower than full precision (64 bit floating point arithmetic) and refine the solution using an iterative approach.

Microsoft Breaks Into Top 10 of World’s Most Powerful Supercomputers

AUSTIN, Texas — Nov. 18, 2008 /PRNewswire/ — Today at the Supercomputing 2008 conference, Microsoft Corp. debuted in the top 10 of the world’s most powerful supercomputers with Shanghai Supercomputer Center and Dawning Information Industry Co. Ltd., which ranked at No. 10 with 180.6 teraflops, the parallel computing speed, and 77.5 percent efficiency. A truly incredible achievement considering that 12 months ago in Reno, Nevada, Microsoft was at 116 on the Top500 list at Top500.org. This is on the heels of Windows HPC Server 2008 releasing to the manufacturing industry in September.
Reduces costs and complexity of high-performance computing
Windows HPC Server 2008 makes supercomputing more accessible to end users by allowing them to harness computing power through a familiar Windows desktop environment. It also reduces the complexity of Top500 runs and increases efficiency. Microsoft announces the availability of the Top500 Excellence Kit, which includes a Top500 guide containing best practices and internal knowledge from Microsoft developers on how to achieve the highest-efficiency LINPACK runs. As a part of the kit, Microsoft is including several management and performance tools used in its 180.6-teraflop run, a High-Performance LINPACK (HPL) Wizard that automatically tunes HPL for your cluster environment. More information is available at http://windowshpc.net/Resources/Pages/Programs.aspx.
A broad platform for software vendors and an expanded playing field for hardware manufacturers
Microsoft and Cray Inc. teamed up in September with an announcement to drive high-productivity computing further into the mainstream in a broad array of markets with the Cray CX1 supercomputer. Now they’re giving one away! More information is available at http://www.superduosupersweeps.com.
Deep investments in HPC and commitment to driving innovation
Parallelizing code is not easy given that programming languages, frameworks, developer tools, and even the majority of developers have grown up in a largely serial age. So the software development industry is taking strides to make parallelism more accessible to developers, and Microsoft is leading that charge. With Visual Studio 2010, Microsoft is delivering the first wave of powerful developer tools such as Task Parallel Library, Parallel LINQ and Coordination Data Structures for managed code to ease the transition to parallel code. These technologies, along with MPI, MPI.Net and Cluster-SOA, extend parallelism to clusters of thousands of nodes using Windows HPC Server 2008. More information on taking parallelism mainstream is available at http://msdn.microsoft.com/en-us/concurrency/default.aspx.
Executive Quotes

Vince Mendillo, director of the HPC Division at Microsoft:

“Beyond the top 10 number itself there are two things that are very impressive — Windows HPC Server 2008 worked at this scale, and we demonstrated that we have expertise inside Microsoft to lead the integration of a top 10 HPC system. We have come a long way in the past four years from building our first 10-gigaflop cluster to the top echelon of the world’s fastest supercomputers.”
Yao Jifeng, technical director of HPC at Shanghai Supercomputer Center:

With the adoption of Windows HPC Server 2008 for our Top500 run, Shanghai Supercomputer Center was able to accelerate the pace of technology innovation in China. Our growing partnership with Microsoft benefits the many colleges, universities, research institutes and corporations that are our customers.”
Sha Chaoqun, vice general manager at Dawning:

“During our Top500 run we achieved 180.6 teraflops and 77.5 percent efficiency while using the familiar and intuitive Windows operating system. We chose to do the Top500 on Dawning 5000A run using Windows HPC Server 2008 for two reasons: performance and ease of use. Our growing partnership with Microsoft helps Dawning accelerate the pace of technology innovation in China.”
Sash Sunkara, vice president of marketing at Mellanox Technologies:

“We are pleased to partner with Microsoft to deliver the world-leading performance, efficiency and scalability of Mellanox InfiniBand while keeping the simplicity and the ease-of-use capabilities of Windows Server platforms. The tight integration of Windows HPC Server 2008 and Mellanox InfiniBand end-to-end hardware and software solutions not only enabled the tremendous performance achievement at the Shanghai Supercomputer Center, but also ensured the needed application productivity at scale.”

There’s no shortage of demand for faster wireless, but today’s fastest technologies–Wi-Fi, 3G cellular networks, and even the upcoming WiMax–max out at tens or hundreds of megabits per second. So far, no commercial wireless system can beat the raw speed of optical fiber, which can carry tens of gigabits per second.

One way to achieve faster speeds is to harness the millimeter-wavelength frequency of the wireless spectrum, although this usually requires expensive and very complex equipment. Now, engineers at Battelle, a research and development firm based in Columbus, OH, have come up with a simpler way to send data through the air with millimeter-wave technology. Earlier this year, in field tests of a prototype point-to-point system, the team was able to send a 10.6-gigabit-per-second signal between antennas 800 meters apart. And more recently, the researchers demonstrated a 20-gigabit-per-second signal in the lab.

Richard Ridgway, a senior researcher at Battelle, says that the technique could be used to send huge files across college campuses, to quickly set up emergency networks in a disaster, and even to stream uncompressed high-definition video from a computer or set-top box to a display.

Whereas Wi-Fi and cellular networks operate on frequencies of 2.4 to 5.0 gigahertz, millimeter-wave technology exploits a region from about 60 to 100 gigahertz. These waves can carry more data because they oscillate faster. Much of the millimeter region is unlicensed and open for use; it has only been neglected because of the difficulty and expense involved in generating a millimeter-wave signal, encoding information on it, and then decoding at the other end. Usually, data is encoded by first generating a low-frequency wave of around 10 gigahertz, then converting it into a higher-frequency signal. The drawback is that encoding data on a 10-gigahertz signal limits the data rate to about one gigabit per second.

The Battelle team was able to better this by more than a factor of 10 using off-the-shelf optical telecommunication components. The researchers modulated data on two low-frequency laser beams, then combined the two. When these two beams combine, they create a pattern of interference that acts as a 100-gigahertz signal. “It looks as though we have a laser beam that has a 100-gigahertz frequency,” Ridgway says.

In the past few years, researchers at Georgia Tech, MIT, Intel, and elsewhere have made great strides in developing millimeter-wave devices. Companies such as Intel have even started pushing for standards that could help develop interoperable technologies that operate at 60 gigahertz. And one company, Gigabeam, has rolled out products that can achieve around one gigabit per second using a point-to-point link over a few hundred meters.

Ridgway explains that using telecommunication lasers has two big advantages. First, they are high power, so the resulting millimeter wave is also of relatively high power. Second, the lasers have been engineered to be stable and dependable, producing a signal that doesn’t fluctuate much compared with standard millimeter-wave sources.

Alan Crouch, director of the Communications Technology Lab at Intel, says that the Battelle work is further evidence that millimeter-wave technology could become increasingly important. “There’s demand for more and more wireless communication solutions in this space,” he says, adding that “there is strong industry interest.”

But the research may be years away from being deployed in a product. Ridgway explains that, since the system has been put together from existing components, it’s much larger than it ultimately needs to be. In addition, a property of the signal called polarization, which plays a role in encoding data, tends to drift during operation, which means that the system requires attention when running. But Ridgway hopes that, with some more engineering, these problems can be ironed out. “We’d like to get it to a point where you could just turn on and go,” he says.

Amazon Plans an Online Store for Movies and TV Shows

SAN FRANCISCO — In a significant step toward vanquishing the local video store and keeping couch potatoes planted firmly in front of their televisions and computers, Amazon.com will introduce a new online store of TV shows and movies on Thursday, called Amazon Video on Demand.
Skip to next paragraph
Related
Times Topics: Amazon.com Inc.

Customers of Amazon’s new store will be able to start watching any of 40,000 movies and television programs immediately after ordering them because they stream, just like programs on a cable video-on-demand service. That is different from most Internet video stores, like Apple iTunes and the original incarnation of Amazon’s video store, which require users to download files to their hard drives.

“For the first time, this is drop dead simple,” said Bill Carr, Amazon’s vice president for digital media. “Our goal is to create an immersive experience where people can’t help but get caught up in how exciting it is to simply watch a movie right from Amazon.com with a click of the button.”

Amazon, which is based in Seattle, is also pursuing the technology and media world’s holy grail — an Internet pipeline to the TV. It has struck a deal with Sony Electronics to place its Internet video store on the Sony Bravia line of high-definition TVs.

The video store will be accessible through the Sony Bravia Internet Video link, a $300 tower-shaped device that funnels Web video directly to Sony’s high-definition televisions. That is an awkward extra expense, for now. But future Bravias are expected to have this capability embedded in the television, making it even easier to gain access to the full catalog of past and present TV shows and movies, over the Internet, using a television remote control.

Mr. Carr said Amazon would pursue similar deals with other makers of TVs and Internet devices. “We can support both streaming and downloading,” he said. “Our goal is to continue to establish partnerships with all companies who have a connected device.”

Amazon Video on Demand will be accessible to a limited number of invited Amazon.com customers on Thursday before it opens more broadly to other users later this summer.

Films and TV shows from almost all the major studios and television networks are available for sale or rental to Amazon’s customers in the United States, at varying prices depending on the program and whether people buy or rent it. The lone holdouts are Walt Disney and ABC, which Disney owns. Both have close relations with Amazon’s digital rival, Apple.

Although Amazon does not release revenue numbers for its digital initiatives, its 10-month-old digital music store, Amazon MP3, is viewed favorably as a solid runner-up to iTunes from Apple. But it is far behind iTunes, which recently surpassed Wal-Mart Stores as the leading supplier of music in the United States.

Amazon Unbox, the company’s original download-only video store, was largely seen as a disappointment because it required customers to download special software to watch the programs they bought. The service also worked only on Windows PCs and TiVo set-top boxes.

To make the new service more enticing, the first two minutes of all movies and TV shows will begin playing for users on Amazon.com immediately when they visit a title’s product page on the digital video store.

It will also let users buy a TV show or movie without actually downloading the video file to the PC’s hard drive. Amazon will store each customer’s selection in what it calls “Your Video Library.” Customers can then watch that show or movie whenever they return to Amazon, even if it is from a different computer or device, a solution that neatly gets around studio concerns about piracy.

“I can be at my office, purchase a movie, and then it will be available on my television at home,” said Robert Jacobs, a senior manager at Sony Electronics. “Creating this on-demand available-everywhere access to premium content is going to be very attractive to consumers.”

Amazon will have some formidable rivals if it hopes to dominate the emerging world of digital video. Apple, Microsoft, Google and Netflix are all looking to capture the coveted real estate in the living room as well. Apple has had the most success with video on its iTunes video store and its Apple TV set-top box. It recently added content from several movie studios and introduced video rentals to the service.

Amazon Video on Demand is not expected to generate significant profits for Amazon, which must pay large royalties to Hollywood studios and develop the costly technical infrastructure required to make the service operate reliably.

But Jeff Bezos, Amazon’s chief executive, may have another goal in mind. Establishing a foothold on televisions could be a way to let couch potatoes and television advertisers link up to the rest of Amazon’s online store with a click of the remote control.

“That is certainly a possibility for the future,” Mr. Carr said.

Amazon is entering the video on demand arena with Sony as its first partner to deliver the experience directly to the TV.

Amazon already offers the ability to download movies and TV shows with their Amazon Unbox Service.

The Amazon video streaming service will be available on the Sony Bravia Internet Video Link system. Sony BIVL (BRAVIA Internet Video Link was announced early last year and connects to the back of BRAVIA TVs.

Amazon will offer 40,000 movies and TV shows over the service with instant streaming capability after purchase.
Amazon will aim to strike deals with other TV and set-top box makers.
There is no date yet when the service will be available.
More details on the NY Times.

Amazon will compete with Netflix again on this service. Netflix streaming is available on PCs, Netflix Roku Player and soon on the Xbox 360.

With the popularity of those new Netlix-playing Roku devices and this week’s news that Xbox will also work with the Netflix Watch Instantly service and yesterday’s launch of the documentary site SnagFilms and the fact that everyone including your grandma has a site that streams movies, it’s no surprise that Amazon has finally also launched a stream service called Amazon Video on Demand. And it’s got its own special Roku-type device to be manufactured by Sony, only it’s much more expensive at $300 (compared to $99 for the Roku). And it seems you also need a Sony-made hi-def television to use it. The good news: eventually Sony’s Bravia TVs will cut out the middle man (aka that $300 device) and Amazon may be able to work out future deals with other TV manufacturers.

Because it comes a little too late in the game, and because iTunes will likely forever keep its rival from offering Disney and ABC titles, I don’t see Amazon Video on Demand being hugely successful. But the service at least sounds better than Amazon Unbox, which sold movie downloads. Because everyone and your grandmother subscribes to Netflix, if anyone wants to stream a movie they’re going to just go with Watch Instantly and Roku. And because the main downside to that service is that it doesn’t transfer to portable devices nor does it allow offline viewing, movie consumers are better off with iTunes for their non-Netflix needs. Hey, at least Amazon is still the preferred place to buy books, whether in paper or digital format.

S’pore firm claims patent to image search
By Eileen Yu, ZDNet Asia
Tuesday, May 27, 2008 07:22 PM

Singapore company says it owns patent to technology used by millions of online sites worldwide to link graphics and pictures to other Web pages.

SINGAPORE–A local company has laid claims to a technology that Web sites across the globe deploy to link images to other Web pages, and sent out notification letters to several companies demanding to be paid licensing fees.

Dubbing itself “pioneers of visual search technology”, Vuestar Technologies said it owns the patent to the technology that enables “Internet searching via visual images”.

In sum, the company implied that any Web site that uses pictures and graphics to link to another site or Web page will need a license from Vuestar.

“Those who use visual images which hyperlink to other Web pages or Web sites…whether on the first page or subsequent pages of a Web site require a Vuestar ‘license of use’,” the company said on its site.

Singapore law firm Keystone Law issued a note to its clients Tuesday, urging those that intend to take up a licensing agreement with companies such as Vuestar, to “carefully examine the terms of such licenses and the claims of the licensors”.

Bryan Tan, director at Keystone Law, said in an e-mail interview: “Always examine such claims carefully and take legal advice if you are not familiar [with the company’s claims]. Even if you decide to settle a claim, make sure you know what ‘rights’ you are paying for.” Keystone specializes in technology law.

In his note to the firm’s clients, Tan added: “We understand that Vuestar has sent invoices for around S$5,000 (US$3,676) to various parties in Singapore asking them to enter into such license agreements, or to cease allowing images on a parties’ Web site to be downloaded or used in Singapore.”

“We believe that this development would have a wide-ranging impact on the Internet community in Singapore, given the wide claims made by Vuestar on the intellectual property covered by the patents,” he said. “Parties operating Web sites, offering Web services or developing Web-based and WAP-based products and services need to be especially careful.”

Vuestar said it does not intend to claim licensing fees from charities and government agencies. The company added that its patent extends beyond Singapore and to “several parts of the world”.

According to Tan, Vuestar’s patent–tagged under publication number 95940–appears also to have been granted in Australia, New Zealand and United States.

It is not clear how the company’s patent will impact other visual search technology companies such as Like.com, which claims to have developed the “first true visual search engine”.

Established in August 2004, Like.com also said it owns almost 12 patents in the areas of visual recognition and search.

———————————————————-

Patent breach by ‘virtually all websites’? Pay up, firm demands
[2008] 27 May_ST

close this window Close

Title:        Patent breach by ‘virtually all websites’? Pay up, firm demands
Source:        Straits Times
Author:        Chua Hian Hou

Legal News Archive

A SINGAPORE firm has threatened to sue websites that use pictures or graphics to link to another page, claiming it owns the patent for a technology used by millions around the world.

In a move that has come under fire from the online community, VueStar Technologies has sent ‘invoices’ to local website operators asking for thousands of dollars in licensing fees.

The company, which said ‘virtually all websites’ are infringing on its patent, is also planning to take on giants like Mircosoft and Google.

It is a battle that could, at least in theory, upend the Internet, though intellectual property experts have some doubts that VueStar can actually enforce its claims.

The company said it has been awarded a patent here and in several other countries, including Australia, New Zealand and the United States, for the method of ‘locating Web pages by utilising visual images’.

In other words, clicking, scrolling or streaming over a visual image to connect with a website or Web page is an infringement, the company claims on its website.

The technique is the de facto method used to connect websites across the globe, from personal blogs to the biggest search engines.

VueStar managing director Paul Smith said if sites want to keep using images as links, they will have to pay his company - located in a single-unit office at The Adelphi off Coleman Street - between ‘$200′ and ‘millions’ annually.

It is a claim, however, that has its doubters.

Technology and intellectual property lawyer Bryan Tan of Keystone Law Corporation said that while VueStar has been granted a patent, it is an extremely wide one.

In fact, ‘if the patent is allowed to stand, it will probably bring the (Internet) industry to its knees’. And VueStar’s patent may be contested and overturned in court, he said.

The firm has been sending out invoices to Singapore companies since last week asking them to pay up, said Mr Smith. He declined to say how many have gone out, but there have been ‘enough to keep my phone busy’.

Those who do not pay up, warned Mr Smith, face legal action, and his company is ‘highly confident that (a court decision) will be in our favour’.

Mr Alvin Koh, who runs the non-profit Arrowana fish website arofanatics.com, received one of VueStar’s invoices last week for $5,350. He does not intend to pay up and said: ‘I would rather close down the site’.

Mr Smith recognises that Mr Koh’s stand will likely be a popular one, and his firm is already girding up for a public backlash.

‘Website owners are just upset because they never had to pay for it before,’ said Mr Smith.

VueStar will begin enforcing its patent claims in Australia and the US ’soon’, he said, and the firm is also working on invoices for Internet heavyweights like Google and Microsoft.

While governments and charities will need a licence, VueStar will not be asking for payment from these parties, he said.

Mr Tan urged companies to contact their lawyers before ‘paying VueStar anything’.

Source: Straits Times © Singapore Press Holdings Ltd. Permission required for reproduction.
return to top of page

INTERESTING SIMILIAR STORY :

BT in court to enforce hyperlink patent

* 10:06 11 February 2002
* From New Scientist Print Edition. Subscribe and get 4 free issues.
* Kurt Kleiner

Printable versionEmail to a friendRSS FeedSyndicate

Tools
digg thisAdd My YahooAdd Google Reader reddit submitNewsvineciteulike submit
Related Articles

* “Visual” search engine could find useful information much more quickly
* 13 June 2001
* Company that pays people to poke holes in patents is about to claim its first biotech victim
* 8 May 2001
* Search New Scientist
* Contact us

Web Links

* BT’s patent
* Internet Patent News Service
* Consumer Project on Technology

BT Group PLC will appear in a New York federal court on Monday to try to enforce a patent it says covers all hyperlinking. BT claims it invented the hyperlink in 1976, and anyone who uses the World Wide Web owes them money.

If upheld, the patent would give BT the right to collect royalties from any Internet Service Provider.

BT sued Prodigy Communications after the ISP refused to purchase a license for the right to use hyperlinks, the clickable links that connect one web page to another. If Prodigy loses the case, other ISPs would also be charged royalties.

BT was granted the US patent back in 1989. It had been filed when the then-Post Office was developing an online information service, called Prestel. Patents were also granted in other countries, but have since expired.

BT sent letters to 16 ISPs in June of 2000 after someone in the company rediscovered the patent in its files. The patent expires in 2006 but, if upheld, ISPs could owe huge back royalties.

“We believe we have a duty to protect our intellectual property and we would expect companies to pay a reasonable royalty based on the revenues that they have enjoyed through the use of that intellectual property,” a BT spokeswoman told Reuters.
Exclusive right

But critics say the patent claim is overly broad and could drive up the cost of the internet access for users. It could even allow BT to decide who gets to use the web and who does not, they say.

“Is it good public policy to allow patents like this? I think this claim illustrates the problems you’ve got with the patent system,” says James Love, director of the Consumer Project on Technology, a Washington-based advocacy group. “They would have an exclusive right, the right to stop you cold from using the Web.”

Other critics do not think the patent will hold up. Gregory Aharonian, publisher of the Internet Patent News Service, has written that the patent will probably be invalidated because some people were using similar technology before 1976.

One potential piece of “prior art” can be seen on a 1968 video clip, where Stanford researchers demonstrated an online editing system that uses hyperlinks.

Apple’s MacBook Air, iPhone and iPod may be all the rage, but the company’s desktop computers are still relative rarities: although analysts’ figures vary, Apple desktops certainly comprise less than 10 per cent of the worldwide market.

This low market share is often attributed to the relatively high prices of Apple computers. However, another reason could be the fact that users rarely have an opportunity to experience Mac OS before buying a computer. This would require a visit to one of the Apple retail stores or an accommodating Mac-owning friend. But there is another way: for some time, various different images of Mac OS X have been available on the internet that have been modified to circumvent its tight coupling to Apple’s hardware. These allow the operating system to be installed on a standard PC. To date, Apple has taken no action against such activities, which perhaps indicates that the ‘buzz’ around its OS is not entirely unwelcome.

However, Apple’s licence agreement does state that Mac OS X should only be installed on Apple hardware, so this is a controversial subject. The versions available for download that we discuss in this article are unauthorised. You might feel morally justified in using one of these after you’ve bought a legitimate copy of Mac OS X Leopard, but you will still be breaking copyright law. Combine that with the dangers of downloading software from unofficial sources, and it’s clear that taking this route is only for the adventurous, on securely isolated test computers. We certainly don’t recommend you do this — however interesting the results may be.

In the best case, Mac OS X Leopard on standard PC hardware can use full graphics acceleration (Quartz, Core Image) and all of the CPU features

Motherboard, CPU & chipset

Modified Mac OS X Leopard images called Kalyway and Iatkos are ideal for Intel-based systems, while PCs with AMD processors are best served by the Zephyroth variant. A recently published image that covers all platforms well is Leo4All. A successful installation of Mac OS X Leopard is by no means guaranteed, but it is very likely.

For the Intel platform, the closer the components are to those in a real Mac, the greater the chances of success. Specifically, this means a motherboard with an Intel CPU/chipset combination that supports SSE3. However, there can still be problems. In our tests, for example, we couldn’t convince the Esprimo P5925 with Intel’s Q35 chipset to cooperate with Mac OS. Perhaps the Q35’s integrated vPro technology for remote client management is the culprit. However, we were successful with motherboards using Intel’s 975, 965 and X38 chipsets.

When it comes to the Zephyroth image for AMD platforms, some limitations must be accepted. For example, it’s not possible to execute 64-bit programs with an AMD processor. Nor are the extensions of virtualisation programs such as Parallels Desktop for Mac and VMware supported. We also failed to install Leopard on AMD’s quad-core Phenom processor, although some users report success if EFI emulation is bypassed.

It’s also important to note that hard disk and DVD drive installation may be problematical. Success is most likely if both components use a Serial ATA (SATA) interface. This is because modern Intel chipsets no longer integrate an UltraATA controller, so motherboard makers use an additional chip from JMicron or another manufacturer. However, these solutions are far from compatible with the Parallel ATA (PATA) specification (older chipsets, such as the Intel 975, have a PATA-compatible interface, and the installation succeeds). In addition, all input devices such as keyboard and mouse, must attach to USB ports.

There are stability issues with some motherboards. For example, in our tests the Gigabyte 965P-DS3 crashed with more than 2GB of RAM fitted. However, once AHCI (Advanced Host Controller Interface) is enabled on the SATA controller, the 965-DS3 runs Mac OS with 4GB of RAM without any problems. For a stable operation, you should also ensure that the CPU extension for protecting against buffer overflow (No Execute/Execute Disable or NX/XD) is enabled in the BIOS.

Network, sound and graphics cards

Hardware support for network, sound and graphics cards is generally good, although problems can still arise. Realtek and Marvell network chips generally work, but some newer Intel chips can prove troublesome. Drivers are not usually available, so an additional network card will be needed. The same is true for Wi-Fi: Atheros and Broadcom chips are mostly OK, but Intel’s 3945ABG is not supported. Some manufacturers, including Ralink and ZyXEL, now even offer Mac OS drivers for download.

Integrated audio, as often used on motherboards, is supported by default in the latest images. Drivers for a number of ALC88x-family chips are on the setup DVD. However, with Sound Blaster products, you’ll have to accept some reduced functionality. Frequently, the 5.1 analogue output does not work, although the digital output does support this standard. Also, the sound inputs only rarely function.

Mac OS X is particularly animation-rich, so support for graphics chips is critical. Drivers are present, but work only rarely for newer ATI and Nvidia GPUs (Graphics Processing Units); older graphics chips from the Radeon X1000 series give fewer problems. With newer graphics chips it’s best to perform the installation without selecting a graphics driver, so that Mac OS starts in VGA mode. In our test with an ATI HD 2600 X, however, even that failed. Installation is successful with some compatible graphics cards, though. We succeeded with ATI’s X1400, X1600 and X1900, while newer graphics cards with HD 2600 XT, HD 2900, HD 3850, HD3870 and HD 3870X2 chips required Mac OS to be started in VGA mode.

After starting Mac OS X Leopard with a X1000 graphics card, drivers can be installed to ensure that newer GPUs offer full graphics acceleration. You can even use an ATI HD 3870 X2, although Mac OS X will not make use of the second GPU in this high-end card. Nevertheless the second chip is recognised and you can use the (four) DVI outputs to connect further monitors — similar to a Mac Pro with two graphics cards. Almost all graphics drivers in circulation support digital LCD connections, although it’s possible that in some setups only the analogue VGA mode will be active. Multi-screen operation is no problem: after connecting a second monitor, Mac OS extends the desktop area automatically.

Installing Leopard on a standard PC

For our test installation of Leopard on a standard PC, we used the Gigabyte GA-X38T-DQ6 for the Intel platform, and the GA-MA790FX-DQ6 for the AMD platform. The hard disk was a Seagate Barracuda 7200.10 with a SATA interface and 750GB capacity, while the DVD drive was an ASUS DRW1814BLT, also with a SATA interface. Both systems used the Radeon HD 3870 X2 graphics card and both were fitted with 4GB of memory from Aeneon.

We installed Mac OS X Leopard from the Leo4All image, which provides excellent hardware support and works with both Intel and AMD processors. Leo4all also integrates many updates, including Leopard 10.5.2. The Kalyway image also works well for Intel systems.

Setup DVD
About five minutes after starting the installation, a graphical setup mode appears where you choose, among other things, the partition into which Mac OS is to be installed. Mac OS initially finds no partition in a new drive, and Disk Utility must be used to create one or more partitions. Thanks to EFI emulation, hard disks can be partitioned using the GUID scheme; GUID is the successor to the MBR (Master Boot Record), supporting more partitions and providing greater flexibility. However, most Windows operating systems do not support the GUID partition scheme: you’ll need to install the GRUB bootloader and jump through some hoops to get both Windows and Mac OS onto a GUID-partitioned hard disk — something that’s beyond the scope of this article. With an MBR-partitioned drive, it’s relatively straightforward to install Mac OS alongside Windows.

The setup process takes a good 30 minutes. Thanks to Leo4all’s good hardware support, network and audio devices are immediately available after the installation. However, you’ll need to install a graphics driver later. After completing the installation, Mac OS plays a short video and asks for a user name and password.

It’s worth noting that, using the Netkas EFI emulation, you can install the original ‘vanilla’ Mac OS kernel. However, this is not advisable, as you’re likely to encounter shutdown/reboot problems. Instead, it’s better to use the modified Netkas kernel 9.2.0.

Tips and tricks

As already mentioned, you may encounter difficulties either during setup or installation. For example, if the system hangs shortly after the start of the setup DVD, or if an error message appears, the following switches for the Darwin bootloader can be helpful: - f, - x, -v, -legacy. Also, switching off one of a dual-core processor’s cores in the BIOS can be helpful.

After installation, a flashing cursor can appear on-screen, followed by nothing much at all. In this case, the partition was probably not marked as active, which prevents startup. You can retrieve this situation manually by starting again from DVD, calling up the Terminal program and entering the following instructions:

fdisk -e /dev/rdisk0 (If a hard disk is in the system)
p (Show partitions)
f 2 (Make partition 1 active (GUID), for MBR enter 1 here instead of 2)
w
y
exit

Post-installation problems: orange disk icons
In many cases, the Leopard hard drive is identified as a removable device — especially if the disk controller is an Intel ICH9 in AHCI mode. Accordingly, the disk’s icon is orange instead of grey. That not only looks bad, but can also cause problems with access rights. This can be resolved by replacing two kernel extensions (’kexts’). The Kext Helper tool, which is often installed by the modified image, lets you easily install the AppleAHCIPort.kext and IOAHCIFamily.kext in /System/ Library/Extensions without being in Terminal with super-user rights (sudo -s).

‘About this Mac’
Apple has no conventional desktop processors in its systems (the Mac mini, iMac and notebooks use Intel’s mobile processors, while the Mac Pro is equipped with server CPUs), which means that Mac OS usually misrecognises standard PC processors. In the ‘About this Mac’ info box, it simply says ‘Unknown Processor’. The memory complement is also often not properly recognised, while the CPU functions supported by Leopard no longer appear.

To correctly display processor and memory information, you need to edit the AboutThisMac.strings file. To do this, go to /System/Library/Core Services/loginwindow.app, right-click on the program and select Show Package Contents. Then open another Finder window and navigate to the /Contents /Resources/German.lproj folder. Here you should find the AboutThisMac.strings file, which you can copy to the desktop and open with TextEdit. Scroll down to the line “ABOUT_BOX_MULTIPLE_PROCESSOR_FIELD_FORMAT” and, after the equals sign, replace the two operands “% @” with the correct CPU string — for example, “Intel Core 2 Extreme CPU X9650″. For a single-processor system, do the same for the “ABOUT_BOX_SINGLE_PROCESSOR_FIELD_FORMAT” line, and enter the correct memory information at “ABOUT_BOX_MEMORY_FIELD_FORMAT”. Then save the file and replace the original with the modified version.

With a little judicious editing, you can get Mac OS X to display CPU and memory information correctly.

Assistance, drivers and instructions for installing Mac OS X on standard PCs are available on the internet. The most comprehensive forum is insanelymac.com; there are also useful tips and tricks at osx86scene.com and hackint0sh.org.

Performance

As far as performance is concerned, there’s no difference between Mac OS on a standard PC and an official Apple Mac — so long as the CPU and graphics card are properly supported. The Intel platform manages this better than AMD, as the latter cannot operate in 64-bit mode and lacks the appropriate CPU extensions for optimising virtualisation performance. By contrast, the overclocked 3.6GHz Penryn quad-core (Intel Core 2 Extreme X9650) works in the Gigabyte motherboard with a frontside bus speed of 1600MHz and is fully supported by Mac OS: all CPU extensions can be used.

As our tests show, Mac OS Leopard running on a single 3.6GHz quad-core Intel processor compares very well to its performance on a Mac Pro’s pair of 2.8GHz quad-core chips. It even outperforms Apple’s high-end workstation in some areas — particularly the memory benchmarks. This is partly the result of the Intel system’s 1,333MHz DDR3 non-ECC memory (Aeneon DDR3 1333) and higher 3.6GHz clock speed, compared to the Mac Pro with its fully buffered DIMMs and 800MHz DDR2 memory. More serious, however, is the fact that with two memory modules installed in the Mac Pro, only two of the four available memory channels are used, cutting the theoretical maximum bandwidth by half. As a result, even the AMD system matches Apple’s high-end workstation in the memory benchmarks. With pure arithmetic operations, however, the dual quad-core Mac Pro is superior to the Intel system, despite its lower clock speed.

Mac Pro: 2 x 2.8GHz quad-core. Intel PC:1 x 3.6GHz Penryn quad-core. Longer bars are better.

Mac Pro: 2 x 2.8GHz quad-core. AMD PC: 1 x 3.0GHz Athlon 64 dual-core. Intel PC:1 x 3.6GHz Penryn quad-core. Longer bars are better.

Conclusion

With the right hardware components, a standard PC running Mac OS X Leopard is, at first sight, no different from a genuine Apple Mac. Special CPU extensions such as Intel VT-x provide support for software solutions like Parallels Desktop for Mac. Even Adobe Photoshop, which queries a Mac to verify its authenticity, runs fine on a standard PC thanks to EFI emulation. You can also use .Mac accounts and the integrated Time Machine backup feature, which didn’t work with earlier Mac OS images. Graphics acceleration is no longer an issue, and nor is the digital control of LCD screens. However, you will have to accept some reduced audio functionality, and will look in vain for drivers for most PCI add-on cards.

Although the modified Mac OS images exhibit no stability problems on optimal hardware under normal conditions, not all functions are supported. Sleep mode — which puts the system into a state that consumes only a few watts of power — works only in exceptional cases, when all the relevant kernel extensions are available.

You’ll need to be careful with software updates, too. Although most Mac OS updates work, those affecting the core operating system (kernel) are often incompatible. For instance, the automatic update to version 10.5.2 caused a system crash, which is very difficult to repair. Changes to key drivers following an update may also prevent Mac OS from starting up on a standard PC. Installing Windows using Boot Camp does not work either. Whatever the difficulties involved in installing and setting up Mac OS on a desktop PC, the situation is usually exacerbated on a notebook — for example, it’s not usually possible to replace a graphics accelerator or a network chip.

Installing a modified Mac OS image on a standard PC is unlikely to be a productive exercise for most users because of the technical obstacles likely to be encountered. However, it can be a rewarding experience for enthusiasts and specialists with Unix expertise. As mentioned at the start of this article, this is a legally questionable pursuit that is definitely not to be undertaken without first purchasing a legitimate copy of Mac OS Leopard.
Running Mac OS X on standard PCs

Kai Schmerer ZDNet Germany

Published: 14 May 2008

Anyone looking for work at home opportunity can easily make handsome earning through pay per click program. For this all they need is basic IT knowledge in the form of Microsoft certification and the latest seo tips. Nowadays some quality web hosting companies are offering free tips to their clients. While making the selection of web site hosting, webmasters should focus on this feature as well. Another basic requirement for success in this field is high speed internet connection. In this regard IP phones are considered the best.