Subscribe to:

Subscribe to :: TheGuruReview.net ::

Is AMD’s RX 560 A Re-branded Card

December 14, 2017 by  
Filed under Computing

AMD has admitted that the Radeon RX 560 which is in the shops now is essentially a rebranded older model.

Yesterday we revealed that AMD had changed the Radeon RX 560 so that it had a similar spec to older cards. Unfortunately, it had mixed the downgraded cards with the older ones which meant you had to be a real geek to spot the difference.

What made the whole thing worse was that it had carried out the change after all the reviews were out and it failed to tell anyone about it.

Now AMD has released a statement which confirmed that there were hobbled versions of the Radeon RX 560 with 14 Compute Unit (896 stream processors) and 16 Compute Units (1024 stream processors) released onto the market.

“We introduced the 14CU version this summer to provide AIBs and the market with more RX 500 series options”, AMD said.

However, now AMD said that it had “come to our attention that on certain AIB and etail websites there’s no clear delineation between the two variants.

“We’re taking immediate steps to remedy this: we’re working with all AIB and channel partners to make sure the product descriptions and names clarify the CU count so that gamers and consumers know exactly what they’re buying. We apologize for the confusion this may have caused.”

So basically, it is saying that there is a Radeon RX 560 which got all the reviews and a Radeon RX 560 which is a rebranded earlier model and it is difficult to tell them apart. It is having a word with its partners to make sure that they are branded correctly. No mention of giving people their money back if they bought the wrong card.

Courtesy-Fud

Toshiba AND Western Digital Settle

December 14, 2017 by  
Filed under Around The Net

Toshiba and Western Digital have agreed in principle to settle a dispute over the Japanese firm’s plans to sell its $18 billion chip unit and aim to have a final agreement in place next week.

Word on the street is that the Toshiba board has approved a framework for a settlement.

Western Digital had been able to block a deal to selling the unit to a Bain Capital-led consortium.

The settlement under discussion calls for Western Digital to drop arbitration claims seeking to stop the sale in exchange for Toshiba allowing it to invest in a new production line for advanced flash memory chips that will start next year.

A Toshiba spokesman said that while the company was open to a settlement, it would not disclose discussion specifics or details of board of directors meetings. “It is not a fact that we have reached an agreement with Western Digital,” he said.

Western Digital is not saying anything.

Toshiba was forced to put the unit – the world’s no. 2 producer of NAND chips – on the block to cover billions of dollars in liabilities arising from its now bankrupt US nuclear power unit Westinghouse.

The deal with the Bain-led consortium will, however, see it reinvest in the unit and together with Hoya a maker of parts for chip devices, Japanese firms will hold more than 50 percent of the business – a keen wish of the Japanese government.

As part of the planned settlement, Toshiba and Western Digital would extend existing agreements for their chip joint ventures in Yokkaichi, central Japan, one of the sources said. The current agreements are set to start expiring from 2021.

Western Digital would also invest in a completely new chip plant that Toshiba will start building next year in northern Japan, the source said.

Western Digital, one of world’s leading makers of hard disk drives, paid some $16 billion last year to acquire SanDisk, Toshiba’s chip joint venture partner since 2000.

With data storage key to most next-generation technologies from artificial intelligence and autonomous driving to the Internet of Things, NAND chips have only grown in importance and Western Digital has been desperate to keep the business out of the hands of rival chipmakers.

The sale still needs to clear the snarling mauls of the regulatory watchdogs but they are not expected to rip the trousers of the deal.

Courtesy-Fud

Will Windows 10 On Snapdragon Succeed

December 11, 2017 by  
Filed under Computing

We had a chance to ask Erin Chappelle,  General Manager of the Windows and Device group of Microsoft, and lead player of Windows 10 on Snapdragon about the lack of 64-bit emulation support at launch.

She was the right person to ask why Windows 10 on Snapdragon is missing  64-bit emulation application support. She also leads the team delivering the base components of the operating system including the Kernel, Hypervisor, Containers and Storage and was certainly the right person to address this “elephant in the room”.

64 Bit on ARM in the future

Erin confirmed to Fudzilla that Microsoft is considering 64-bit emulation support in the future, but it was a time to market and executive decision that prevented 64-bit being supported at launch. Microsoft will enable 64-bit SDK for developers and this might be the way to optimize the applications in the future. Our understanding is that if Adobe comes up with an idea that it wants to make a native Photoshop for ARM, Microsoft and its SDK will be there to help.

Erin also said that most X64 applications in the market are high end games. This is not what the Snapdragon platform is tailored for. At the Snapdragon technology summit in beautiful Maui, Hawaii, it was pointed out by Erin as well as Terry Myerson, executive vice president of the Windows and Devices Group at Microsoft, that the always connected PC is a completely new market for Microsoft.

Always Connected PC is the new market

Of course Microsoft sees potential here,  as the company has invested a lot of time and resources into the Windows on Snapdragon, always connected PC project. The motivation is very straightforward. Microsoft wants to get into new markets as it realizes that people need the connectivity on modern devices.

HP Envy X2 device comes with 4GB or RAM while Asus NovaGo always connected Snapdragon 835 device comes with up to 8GB memory. While the OS can support 8GB of memory, applications will be limited to 4GB. This might become a problem in some extreme scenarios – but both Qualcomm and Microsoft pointed out that these devices are targeting the casual, every day user market, rather than high end video editing professionals.

You start with a market where you have a good value proposition and in this case this was the consumer market. In the future, Microsoft hopes to bring always connected PC ideas to the enterprise, which will benefit from the security and cloud opportunities for mobility that 4G and 5G offer for enterprises.

Snapdragon 835 devices are, it seems the first of many to come. There is an opportunity for higher as well as lower cost devices. This is what Miguel Nunes, Director of Product management for Windows on Snapdragon and Erin Chapple left us with.

Courtesy-Fud

Samsung Begins Production of 512GB Storage Chip

December 8, 2017 by  
Filed under Mobile

Samsung confirmed that it has begun mass production of a 512GB embedded Universal Flash Storage memory chip for mobile devices, meaning that your next Galaxy phone could have seemingly endless storage.

High storage phones models currently come with 128GB or 256GB of memory, but Samsung’s new 512GB chip is double or quadruple that. Samsung says the chip consists of eight 64-layer 512GB V-NAND chips, but what’s interesting is that although it doubles the storage and density of Samsung’s 256GB chip, it takes up the same amount of physical space.

Of course Samsung’s and other Android phones sometimes have the option of expandable storage through a microSD chip, but internal storage has its benefits. The 512GB chip is able to read and write new data at 860MB per second and 255MB per second respectively, which Samsung claims is eight times faster than your average microSD card. Plus, expandable storage has always been an add-on; internal storage is built directly into your phone.

Samsung says that the chip is intended for use in next-generation phones, which makes us wonder which phones will get it. Will the Galaxy S9 or Galaxy Note 9 come with 512GB of storage? Or will it appear in other manufacturers’ phones, like how some iPhones used Samsung memory chips.

Increasing sizes of internal storage is a good thing. This is especially the case as apps get bigger, operating systems take up more space and photo and video quality increases for built-in cameras. Samsung claims that the 512GB chip can hold approximately 130 10-minute 4K Ultra HD videos, which is good news for all the phones that come with 4K video recording capability.

If you can’t imagine using that much storage, Samsung also announced that it’ll expand production of its 256GB chips, too. So as memory extremes increase on the high end we may see more phones with 256GBs of storage, too.

Will Apple Sell A Cheaper iPad

December 7, 2017 by  
Filed under Around The Net

A report said that Apple will introduce a 9.7-inch iPad next year, with a comparatively low-price tag.

According to Digitimes, which quotes unnamed Apple suppliers, the unit will sell for around $250 or so.

The tablet market is somewhat flat because people don’t have to replace them for quite some time, and those that want tablets generally have one, unless they’re tempted like a magpie for more bright shiny things.

The iPad is slated for the second quarter of next year and the wire said that Compal is likely to make the unit with other firms like Compeq and Unitech contributing to the components. Apple also wants to shift iPad into the industrial sector too.

If the reports are true, it is likely to spur Apple competitors to cut their prices too.

Courtesy-Fud

IS Apple Using Qualcomm Patents Without Paying

December 6, 2017 by  
Filed under Around The Net

Apple’s legal spat with Qualcomm just got even messier as the chipmaker found even more patents which it thinks that Apple is using without permission.

Qualcomm filed three new patent infringement complaints against the fruity cargo cult saying there were 16 more of its patents that Apple was using in its iPhones.

The new complaints represent the latest development in a long-standing dispute and follows Apple’s countersuit against Qualcomm, which alleged that it invented part of Qualcomm’s Snapdragon mobile phone chips.

Qualcomm in July accused Apple of infringing several patents related to helping mobile phones get better battery life.

That case accompanied a complaint with the US International Trade Commission seeking to ban the import of Apple iPhones that use competing Intel Corp (INTC.O) chips because of the alleged patent violations.

The three cases filed Thursday were all filed in the US District Court for the Southern District of California in San Diego. One of the cases is a companion civil lawsuit to a new complaint also filed Thursday with the ITC that seeks the same remedy of banning iPhones with Intel chips. The other two cases are civil patent infringement lawsuits.

It all started when Apple tried to increase the margins for its dying iPhone cash cow.  It decided that Qualcomm was asking too much and demanded a reduction.

In January, Apple sued Qualcomm for nearly $1 billion in patent royalty rebates that Qualcomm allegedly withheld from Apple.

In a related suit, Qualcomm sued the contract manufacturers that make Apple’s phones, but Apple joined in to defend them.

Qualcomm in November sued Apple over an alleged breach of a software agreement between the two companies. Apple emailed Qualcomm to request “highly confidential” information about how its chips work on an unidentified wireless carrier’s network, Qualcomm alleged, but Apple had copied an Intel engineer in the email for information.

Courtesy-Fud

Windows 10 Hits 600 Million Mark

December 5, 2017 by  
Filed under Computing

Software king of the world Microsoft says that its Windows 10 is on 600 million devices.

CEO Satya Nadella referenced the new number for the first-time moments ago at the company’s annual shareholders meeting, where he is giving analysts and investors an update on Microsoft’s progress and strategy.

This is an increase from the 500 million devices touted by Microsoft earlier this year, but it’s still well short of the company’s original goal of one billion Windows 10 devices within two to three years of its 2015 release.

Microsoft has admitted that the original goal is a little too optimistic, but it has seen Windows XP use dropping away lately which is at least some comfort.

Courtesy-Fud

Chip Market Soars Past 400 Billion

December 1, 2017 by  
Filed under Computing

The World Semiconductor Trade Statistics (WSTS) body forecast that the chip market will increase 20.6 percent this year compared to 2016.

And the WSTS can’t see any signs of the upwards trend slowing, because based on its current figures it’s forecasting the market will be worth $437 billion.

The rise in the market of 20.6 percent is the largest year of growth since the decade began.

There’s growth in all major categories but the way out leader is memory at 60.1 percent, followed by sensors with 15.9 percent growth.

The WSTS said all geographical regions will grow.

Courtesy-Fud

Los Alamos Lab To Build Raspberry Pi Supercomputer

November 30, 2017 by  
Filed under Computing

The National Laboratory’s High Performance Computing Division Los Alamos has access to 750-node Raspberry Pi clusters which it is using in development work.

Super-computing 17, an annual conference on high-performance computing, was told that the Raspberry Pi-based clusters were part of the first step towards a development program to assist in programming much larger machines.

The platform at LANL uses a modular cluster design from BitScope Designs, with five rack-mount Bitscope Cluster Modules, each with 150 Raspberry Pi boards with integrated network switches.

With each of the 750 chips having four cores, it offers a 3000-core highly parallelisable platform that emulates an ARM-based supercomputer, allowing researchers to test development code without requiring a power-hungry machine at significant cost to the taxpayer.

The full 750-node cluster, running 2-3 W per processor, runs at 1000W idle, 3000W at typical and 4000W at peak – with the switches – and is substantially cheaper, if also computationally a lot slower.

After development using the Pi clusters, frameworks can then be ported to the larger scale supercomputers available at LANL, such as Trinity and Crossroads leader of the High Performance Computing Division at Los Alamos National Laboratory, Gary Grider, told the conference: “It’s not like you can keep a petascale machine around for R&D work in scalable systems software. The Raspberry Pi modules let developers figure out how to write this software and get it to work reliably without having a dedicated testbed of the same size, which would cost a quarter billion dollars and use 25 megawatts of electricity.”

The collaboration between LANL and BitScope was formed after the inability to find a suitable dense server that offered a platform for several-thousand-node networking and optimization – most solutions on the market were too expensive, and anyone offering something like the Pi in a dense form factor was ‘just people building clusters with Tinker Toys and Lego’.

BitScope says that each node will be about $120 fully provisioned using the element14 version of the latest Raspberry Pi (normally $35 at retail). That means that a 150-node Cluster Module will fall in around $18k-$20k each.

Courtesy-Fud

Intel’s Find Security Flaw In Management Engine

November 29, 2017 by  
Filed under Computing

Security problems have plagued the Intel Management Engine for years, but the chipmaker is only just confirming the severity of the flaws.

The technology, which is a core part of Intel Active Management Technology (AMT), is present on many of the company’s CPUs and can even remain active when a PC is turned off.

But for years, security specialists have identified a conveyor belt of exploitable security flaws.

On Monday, Intel posted a new security advisory warning manufacturers and users of a string of new vulnerabilities and bugs found in the management engine, along with Server Platform Services and the Trusted Execution Engine.

Intel recently completed a security audit to identify and explore potential vulnerabilities affecting the Management Engine, following warnings from security researchers.

The company is clearly trying to take these vulnerabilities seriously and has since unveiled a detection tool. This is intended to provide Windows and Linux systems administrators with the ability to check their systems for flaws.

Attackers have been capitalising on the fact that the management engine has access to important system processes, and Intel has confirmed that the worst-case scenarios could become a reality.

In particular, cyber criminals can cause instability with complete system crashes by exploiting the management engine. They’ve also found a way to “impersonate” the engine and, in the process, kill existing PC security mechanisms. 

Then arbitrary code can be deployed. In most cases, the user will be unaware of what’s happening to their system. 

The affect Intel chips, servers, PCs and other connected devices. Intel has the ability to issue updates, but it’s up to hardware manufacturers to implement them – which means that most devices won’t be updated, especially consumer PCs. 

Speaking to Wired, a spokesperson for the company said: “Businesses, systems administrators, and system owners using computers or devices that incorporate these Intel products should check with their equipment manufacturers or vendors for updates for their systems.”

In the accompanying advisory, it added: “In response to issues identified by external researchers, Intel has performed an in-depth comprehensive security review of the following with the objective of enhancing firmware resilience:

“Intel has identified security vulnerabilities that could potentially impact certain PCs, servers, and IoT platforms. Systems using Intel ME Firmware versions 11.0.0 through 11.7.0, SPS Firmware version 4.0, and TXE version 3.0 are impacted.

“To determine if the identified vulnerabilities impact your system, download and run the Intel-SA-00086 Detection tool. Contact your system manufacturer to obtain updates for impacted systems.”

Courtesy-TheInq

Samsung Rumored To Show Off New Flagship Phones At CES

November 28, 2017 by  
Filed under Mobile

Samsung lovers may get a sneak peek at the upcoming Galaxy S9 and S9 Plus earlier than expected.

According to VentureBeat, the pair of marquee phones is rumored to make an appearance at CES 2018, the huge consumer electronics trade show held annually in Las Vegas.

Though Samsung is expected to fully introduce the Galaxy S9 and S9 Plus sometime in March 2018 (similar to how it launched the S8 and S8 Plus during a March keynote this year), a cameo at CES in January would push up previews of the devices by two months.

The degree of this sneak peek remains unclear. Citing “someone briefed on the company’s plans,” VentureBeat says the phones would make their “first public appearance” at CES.

However, it’s unusual for Samsung to bring physical demos of upcoming flagships at CES. Instead, the Korean phone maker could tease the device by showing a short video clip or digital image rendering. For now, Samsung said it “does not comment on product rumors or speculation.”

Whatever the debut strategy may be, speculation around the devices are swirling. The Galaxy S9 Plus, for instance, is rumored by the same source to have more RAM and a dual-camera in addition to a bigger display. This is different from the current S8 and S8 Plus, which only differ in size.

The phones are also expected to have Snapdragon 845 processors (and an upcoming Exynos processor for other global models), expandable memory and a 3.5mm headphone jack.

CNET will be on the ground reporting from CES 2018, so if the phones do make an appearance we’ll be there with all the news.

Can IBM’s Quad9 Block Botnets

November 28, 2017 by  
Filed under Computing

IBM has partnered with the Global Cyber Alliance (GCA), an organisation founded by law enforcement and research firms to help reduce cyber-crime, to launch a free public Domain Name Service (DNS) system.

While that might not sound so fascinating, the interesting thing is that the new DNS system, named Quad9, will block domains associated with botnets, phishing attacks, and other malicious internet hosts. This is especially good news for businesses that don’t run their own DNS blacklisting and whitelisting services, as it will make them much safer.

Quad9, which is named as such du to its 9.9.9.9 Internet Protocol address, works in the same way as any other public DNS server, such as Google’s, but the difference is it won’t return name resolutions for sites that are identified via threat feeds the service aggregates daily.

“Anyone, anywhere can use it,” said GCA’s president and chief operating officer, Phil Rettinger, in an interview with Ars Technica, adding that the service will be “privacy-sensitive” with no logging of the addresses making DNS requests.

“We will keep only [rough] geolocation data,” he said, explaining that this will be used to track the spread of requests associated with particular malicious domains. “We’re anonymising the data, sacrificing on the side of privacy,” he added.

So where does IBM come in? According to GCA, the computer giant will provide the power behind one of Quad9’s major threat feeds, one of which is IBM’s X-Force. This converts the feeds into a database that is then de-duplicated.

So whenever a Quad9 user clicks on a website link or types an address into a web browser, Quad9 checks the site against IBM X-Force’s threat intelligence database of over 40 billion analysed web pages and images. The other 18 feeds the service taps from include threat intelligence partners including Abuse.ch, the Anti-Phishing Working Group, Bambenek Consulting, F-Secure, mnemonic, 360Netlab, Hybrid Analysis GmbH, Proofpoint, RiskIQ, and ThreatSTOP.

Quad9 then generates a whitelist of domains never to block, using a list of the top one million requested domains, as well as a “gold list” of safe providers, such as major Internet service sites like Microsoft’s Azure cloud, Google, and Amazon Web Services.

The blocked sites, whitelist, and gold lists are then converted into a Response Policy Zone (RPZ) format before being pushed out to the clusters of DNS servers around the world maintained by Packet Clearing House via DNS zone transfers. The DNS server clusters, which are each load-balanced with dnsdist, use a mix of Unbound and PowerDNS servers to deliver responses.

As of launch, there were clusters of DNS servers configured in 70 different locations around the world, and Quad9 expects to have 100 sites up and running by the end of the year. It’s also free, but will need to be continually funded as the GCA is a non-profit.

IBM said that telemetry data on blocked domains from Quad9 will be shared with threat intelligence partners for the improvement of their threat intelligence responses for their customers and Quad9.

Courtesy-TheInq

Is Gaming Growing The GPU Market

November 28, 2017 by  
Filed under Computing

Overall GPU shipments increased 9.3 percent from last quarter, AMD increased eight percent Nvidia rose 30 per cent and Intel, increased by five percent, according to figures from Jon Peddie Research.

 The desktop gain is attributed to gaming and cryptocurrency which mainly helped AMD and Nvidia gain market share.

The third quarter is typically the strongest from the previous quarter in the seasonal cycles of the past. For Q3’17 it increased 9.3 percent from last quarter and was below the ten-year average of 9.52 percent.

AMD’s overall unit shipments increased 7.63 percent quarter-to-quarter, Intel’s total shipments rose 5.01 percent from last quarter and Nvidia’s increased 29.53 percent. The attach rate of GPUs (includes integrated and discrete GPUs) to PCs for the quarter was 144 percent which was down -1.28 percent from last quarter.

Discrete GPUs were in 39.55 percent of PCs, which is up 4.18 percent. The overall PC market increased 10.31 percent quarter-to-quarter, and decrease -2.06 percent year-to-year.

Desktop graphics add-in boards (AIBs) that use discrete GPUs increased 29.05 percent from last quarter. Q3’17 saw an increase in tablet shipments from last quarter.

The normal seasonality has returned to the PC market, albeit in a slowly declining way, the report said.

GPUs are traditionally a leading indicator of the market, since a GPU goes into every system before it is shipped, and most of the PC vendors are guiding cautiously for the fourth quarter.

The Gaming PC segment, where higher-end GPUs are used, was the bright spot in the market in the quarter. AMD’s shipments of desktop heterogeneous GPU/CPUs, i.e., APUs, increased 7.1 percent from the previous quarter.

AMD’s notebook APU shipments were up 2.2 percent. Desktop discrete GPUs increased 16.1 percent from last quarter, and notebook discrete shipments increased 5.2 percent. AMD’s total PC graphics shipments increased 7.6 percent from the previous quarter.

Intel’s desktop processor-graphics shipments increased from last quarter by 5.0 percent, and notebook processors increased by 5.9percent, and total PC graphics shipments increased 5.0 percent from last quarter.

Nvidia’s discrete desktop GPU shipments were up 34.7 percent from last quarter, and the company’s discrete notebook GPU shipments increased 22.4 percent, and total PC graphics shipments increased 29.5percent from last quarter.

Total discrete GPUs (desktop and notebook) shipments for the industry increased 23.3 percent from the last quarter and increased 11.7 percent from last year.

Sales of discrete GPUs fluctuated due to timing, memory pricing issues, new product introductions, and the influence of integrated graphics.

Only one percent of Intel’s non-server processors don’t have graphics, and over 66 percent of AMD’s non-server processors contain integrated graphics; AMD still ships integrated graphics chipsets (IGPs).

Courtesy-Fud

Did Kaspersky Hack NSA Staff

November 22, 2017 by  
Filed under Around The Net

Kaspersky has denied it played a role in hacking into the personal computer of a US National Security Agency (NSA) worker.

Kaspersky Lab has published a report detailing an internal investigation it launched examining allegations that its software was used to compromise an NSA employee’s home computer.

In early October, a report published in the Wall Street Journal claimed that the firm’s software was used to download confidential data from an American agent’s home computer.

However, later reports circulated accusing the firm of deliberately taking files from the PC. Following the incident, Kaspersky conducted a full investigation to gain additional evidence of the incident and explore how it happened.

Researchers at the company confirmed that Russian cybercrooks installed software on an NSA contractor’s computer to access and steal sensitive data.

The user, according to the company, was able to download and install pirated software on the machine. The researchers identified a compromised Microsoft Office ISO file, as well as an illegal Microsoft Office 2013 activation tool.

They were able to install the pirate copy of Office 2013 after disabling the Kaspersky security product. If the latter had been left on the PC, it would have identified the illegal activator tool.

This illegal tool was infected with malware, and this was left on the PC while the Kaspersky software was inactive. The malware meant other third-parties could access the user’s machine, causing major security concerns.

However, when the company’s antivirus software was re-enabled, it detected the software with the verdict Backdoor.Win32.Mokes.hvl and stopped it from contacting a dodgy command and control software.

This backdoor approach was first identified in October 2014, but it’s still being used by cybercriminals looking to steal important data. Kaspersky researchers said the antivirus software detected other variants of the Equation APT malware too.

Various variants of the malware, including a 7zip archive, was sent to the Kaspersky Virus Lab for analysis. Researchers found that it contained a number of source codes and classified documents.

At the request of the firm’s CEO, these files were removed from its servers.

“The reason Kaspersky Lab deleted those files and will delete similar ones in the future is two-fold: first, it needs only malware binaries to improve protection and, secondly, it has concerns regarding the handling of potentially classified material,” the firm wrote.   

“Because of this incident, a new policy was created for all malware analysts: they are now required to delete any potentially classified material that has been accidentally collected during anti-malware research.”

“To further support the objectivity of the internal investigation we ran it using multiple analysts including those of non-Russian origin and working outside of Russia to avoid even potential accusations of influence.”

Speaking about other findings, the firm said: “One of the major early discoveries of the investigation was that the PC in question was infected with the Mokes backdoor – a malware allowing malicious users remote access to a computer.

“As part of the investigation, Kaspersky Lab researchers took a deeper look at this backdoor and other non-Equation threat-related telemetry sent from the computer.

Courtesy-TheInq

Linux Appears To Be The King In The Supercomputing Space

November 21, 2017 by  
Filed under Computing

Looking at the November 2017 TOP500 Supercomputer list one thing is particularly clear – the open saucy Linux is king.

In 1998, Linux first appeared on the TOP500 Supercomputer list and it was regarded as unusual – indeed many just said it was because it was really Unix in drag. But the November list showed that all 500 of the world’s fastest supercomputers are running Linux.

There had only been two non-Linux systems left on the list but the pair of Chinese IBM POWER computers running AIX were too slow to rate a mention any more.

Before Linux took the lead, Unix was everywhere but slowly, since 2003, the Linux was the TOP500 main OS of choice. By 2004, Linux had taken the lead for good.

The reason Linux did well in this arena and not the desktop is that most of the world’s top supercomputers are research machines built for specialised tasks, each machine is a standalone project with unique characteristics and optimisation requirements.

Linux means that research teams can easily modify and optimize open-source code to their one-off designs.

Courtesy-Fud

Next Page »