Articles about Nvidia

Zombie Moore's Law shows hardware is eating software

After being pronounced dead this past February - in Nature, no less - Moore’s Law seems to be having a very weird afterlife. Within the space of the last thirty days we've seen: Intel announce some next-generation CPUs that aren’t very much faster than the last generation of CPUs; Intel delay, again, the release of some of …
Mark Pesce, 22 Sep 2016
Array in a rack. Image via Shutterstock

Rackspace punts piscatorial Power platform at service providers

Freshly acquired hosting slinger Rackspace has built a custom-developed server based on IBM’s Power chips and the Open Compute Project (OCP) specifications. The company hopes to flog it to service providers and enterprise customers looking for an alternative to x86 systems for high performance computing. Known as Barreleye …
Dan Robinson, 21 Sep 2016

Nvidia: Eight bits ought to be enough for anybody ... doing AI

Analysis Nvidia has designed a couple of new Tesla processors for AI applications – the P4 and the P40 – and is talking up their 8-bit math performance. The 16nm FinFET GPUs use Nv's Pascal architecture and follow on from the P100 launched in June. The P4 fits on a half-height, half-length PCIe card for scale-out servers, while the …
Chris Williams, 13 Sep 2016
Woman motorist waves from a red car. Photo by Shutterstock

Apple, Google auto auto woe

Apple and Google's dreams of people buying self-driving cars and being driven around by cyber-chauffeurs by 2020 are bursting, it seems. First, we reported that Google's auto autos find everyday things such as some traffic lights and junctions hard work, and that ideally, as a commercial product, the technology needs a …
Chris Williams, 12 Sep 2016
empty_box

IBM lifts lid, unleashes Linux-based x86 killer on unsuspecting world

HPC Blog IBM is mounting its strongest challenge yet to x86 hegemony with the unveiling of its spanking new Linux based S822LC system. What makes this system different is that it’s based on a new processor/motherboard configuration, complete with a sporty NVIDIA NVLink connector. According to my research, close to half of you HPC users …
Android Nougat

Google emits three sets of Android patches to fend off evil texts, files

It's a smaller-than-usual Android patch bundle from Google – just 47 patches for 57 flaws. These software bugs can be exploited by installed apps or malicious code smuggled in multimedia messages and files to gain total control of vulnerable phones, tablets, internet-connected fridges and other Android gadgets – allowing …
Iain Thomson, 06 Sep 2016
An artist sketches with Yoga Book

Lenovo's tablet with a real pen, Acer's monster laptop, Samsung Galaxy S3 watch

IFA 2016 August 31st was press day one ahead of the IFA event in Berlin, which means vendors lining up to show off their latest efforts in (mainly) consumer electronics. Acer kicked off the madness, unveiling the "world's first laptop with a curved screen," though you should think twice before letting the 8 kg Predator 21X gaming …
Tim Anderson, 01 Sep 2016

Stop lights, sunsets, junctions are tough work for Google's robo-cars

Hot Chips After cruising two million miles of public roads, Google's self-driving cars still find traffic lights, four-way junctions and other aspects of everyday life hard work. To be sure, the hardware and software at the heart of the autonomous vehicles is impressive. But it's just not quite good enough yet to be truly let loose on …
Chris Williams, 24 Aug 2016

If you want to build your own Nvidia-powered self-driving car – or hack one – here's a blueprint

Hot Chips Nvidia is channelling Thunderbirds legend Parker for its latest system-on-chip for self-driving cars. Two Parker processors are used in the PX 2 box we saw at the start of the year. This hardware adds a super-cruise-control to vehicles by hoovering up video feeds from onboard cameras and other sensor data, feeding it through …
Chris Williams, 23 Aug 2016
Dog waits on a Welcome mat. photo by sHutterstock

Nvidia welcomes Intel into AI era: Fancy a benchmark deathmatch?

HPC blog Nvidia just fired the first salvo in what promises to be a classic and long-lived benchmark death match vs Intel. In a webpage titled "Correcting Intel's Deep Learning Benchmark Mistakes," Nvidia claimed that Intel was using outdated GPU benchmark results and non-current hardware comparisons to show off its new Knights Landing …
monty_python_french_knights_648

Intel teases geeks with 2017 AI hyper-chip: Xeon Phi Knights Mill

IDF16 Intel is working on a powerful Xeon Phi processor for servers and workstations that is "optimized" for artificial-intelligence software – and it's codenamed Knights Mill. Chipzilla's data center group boss Diane Bryant flashed up this slide during this morning's Intel Developer Forum keynote in San Francisco: The chip is …
Chris Williams, 17 Aug 2016
Robot AI Woman

Chip giants pelt embedded AI platforms with wads of cash

Analysis Artificial intelligence and machine learning engines are underpinning many emerging applications and services, from making sense of big data for enterprises, to supporting hyper-personalized consumer content, or virtual reality gaming. The current challenge is to move AI from the supercomputer to the mobile device, supporting …
Wireless Watch, 17 Aug 2016
man crying, image via Shutterstock

Xeon Phight! Phight! Phight! Nvidia says Intel cheated benchmarks

Nvidia has fired its latest salvo in the ongoing “benchmark wars”, accusing Intel of fudging Knights Landing (Xeon Phi) comparisons to GPUs. The particular claims that Chipzilla made during ISC 16 in June related to deep learning performance claims, specifically that Xeon Phi ran a deep learning benchmark 2.3 times faster in …
SGI in Jurassic Park

OMG: HPE gobbles SGI for HPC. WTF?

Hewlett Packard Enterprise has bought venerable computing firm SGI – formerly Silicon Graphics – for $275m. Over the past few years, SGI has concentrated on the big-iron end of the server market and high-performance computing. It's this technology that HPE wants to bolster its hardware lineup. "At HPE, we are focused on …
Iain Thomson, 11 Aug 2016
NVidia aims to gives games power to handhelds

PCs? Eh, who needs 'em? Nvidia asks while hoovering up more cash

Nvidia says its last quarter set record high marks for revenues, as returns were up across the board. For the second quarter of fiscal 2017 – the three months to July 31 this year – the California-based GPU specialists logged: Revenues of $1.43bn, a 24 per cent increase over last year's $1.15bn. Net income of $253m was up …
Shaun Nichols, 11 Aug 2016
Cloud desktop

Azure courts DaaS with embiggened GPU-powered instances

Microsoft's taken the wraps of some new, GPU-infused, Azure compute instances that it's aiming at heavy compute users and the emerging desktop-as-a-service market. The NC series of instances is aimed at those who like to add a little GPU grunt to their computing, so puts the Tesla K80 in harness with either six, 12 or 18 Xeon …
Simon Sharwood, 05 Aug 2016
Android patch

Android's latest patches once again remind us: It's Nexus or bust if you want decent security

Another month means another double bundle of security vulnerability patches for Android. Google is sticking to the twin-release pattern it used last month: the first batch addresses flaws in Android's system-level software that everyone should install, and the second squashes bugs in hardware drivers and kernel-level code that …
Iain Thomson, 02 Aug 2016
Nvidia GeForce

Bought a GTX 970? Congrats, Nvidia owes you thirty bucks

Graphics goliath Nvidia has agreed to a settlement that will see it pay $30 to American gamers who purchased its GTX 970 graphics cards and can file a valid claim. The pay-out deal [PDF] settles a class-action claim accusing the company of misleading consumers on the specs of the gaming card's memory and processor capabilities …
Shaun Nichols, 28 Jul 2016
Far Cry 4 in game at 4K

AMD, Nvidia in pro GPU battle

Rival chip designers AMD and Nvidia continue to fight for dominance in the professional graphics processor market. Here's a summary of what's gone down this week: AMD is cooking up a Radeon Solid State Graphics (SSG) card that has 1TB of M.2 flash memory. It'll be powerful enough to scrub through 8K video at 90 frames per …
Chris Williams, 26 Jul 2016
android_money_648

Samsung: You won't settle? Fine, we'll do this the Huawei

Samsung has filed a countersuit against Huawei in China over allegations of patent infringement. The South Korean electronics behemoth filed allegations of infringement on six patents it holds in China. According to filings seen by Reuters reporters in Beijing, the Samsung claim would seek to collect 161 million yuan (around …
Shaun Nichols, 22 Jul 2016
Qualcomm Snapdragon 820

Huge double boxset of Android patches lands after Qualcomm disk encryption blown open

Google has released two bundles of Android security patches this month: a smaller one to handle bugs in the operating system, and a larger package that tackles a raft of driver-level issues, particularly with Qualcomm's hardware. The first tranche of patches includes eight critical, 11 high severity, and nine fixes that are …
Iain Thomson, 06 Jul 2016

Powerful ARMs, heaving racks ... yes, of course it's this year's student cluster-building contest

HPC Blog It's time for our traditional video look at each of the teams in the ISC (International Supercomputing Conference) 2016 Student Cluster Competition. Let's take it alphabetically this time... Team Boston Green: They're one of two teams utilizing ARM processors to fuel their cluster. A lot of ARM processors, about 800 cores …

Undergrads build 12.6-TFLOPS cluster out of four nodes, 112 cores

HPC Blog The ISC (International Supercomputing Conference) Student Cluster Competition once again cements its reputation as the place where LINPACK records go to fall. Last week we saw not one, but two teams top the current 12.03 TFLOPS record that was established at the ASC'16 spring competition. Team HUST, from the Huazhong …
Exit sign. Pic:  Lukas Kästner

It's all fun and games until someone loses a rack*

HPC Blog This year’s ISC’16 Student Cluster Competition boasts the most diverse set of hardware in the near 10-year history of student cluster competitions. Student teams are running three different system architectures (x86, ARM, and Power) in both traditional and hybrid (hardware accelerated) forms. The configurations of these …

All aboard the PCIe bus for Nvidia's Tesla P100 supercomputer grunt

ISC Nvidia has popped its Tesla P100 accelerator chip onto PCIe cards for bog-standard server nodes tasked with artificial intelligence and supercomputer-grade workloads. The P100 was unveiled in April at Nvidia's GPU Tech Conference in California: it's a 16nm FinFET graphics processor with 15 billion transistors on a 600mm2 die. …
Chris Williams, 20 Jun 2016

You've got a patch, you've got a patch ... almost every Android device has a patch

It's the first Monday of the month, and that means another batch of patches for Android, fixing flaws that can be exploited by apps and webpages to hijack devices. As usual, if you're not using a Google Nexus device, you're at the mercy of your manufacturer and phone carrier to approve and distribute these updates, which may …
Iain Thomson, 06 Jun 2016
Xeon E3-1500 v5

You don't need no STEEENKING GPU, says Intel

Intel's again trying to make the argument that integrated graphics trump dedicated graphics processing units, by cooking a new Xeon tuned to processing video. Chipzilla's been here before: its Iris Pro graphics made it into last year's Xeons. This year's Xeon E3-1500 v5 is pitched at those who want to stream 4K video. It's …
Simon Sharwood, 01 Jun 2016

Surface Book nightmare: Microsoft won't fix 'Sleep of Death' bug

Hands-on In its most recent quarterly earnings report, Microsoft highlighted its increasingly popular Surface line as the growth leader in its More Personal Computing line of business. Surface led the category with 61 per cent growth in constant currency, a rise driven by the top products in the line, the Surface Pro 4 tablet and the …
Addison Snell, 27 May 2016
robot rally car

Safety, pah! Digital Dukes of Hazzard have robot cars powersliding

Video Last month, Google, Ford and Uber started a lobbying group called the Self-Driving Coalition for Safer Streets to promote autonomous vehicles on safety grounds, but boffins in Georgia are having far more fun with their hardware. While Google spends time pootling around Mountain View at 25mph, researchers at Georgia Tech, the …
Iain Thomson, 23 May 2016

Home team wins Asia student cluster competition

HPC Blog For the first time in student cluster competition history, a hometown team has won the Overall Championship award. The team, from China’s Huazhong University of Science and Technology, captured the crown at the recently concluded Asia Supercomputer Community 2016 Student Cluster Competition. This win earns them a berth in the …
Toothpaste image via Shutterstock

Linux Mint to go DIY for multimedia

The Linux Mint project has decided version 18, scheduled for June 2016, will end out-of-the-box installation of multimedia codecs. The reasoning is straightforward: shipping with codecs involves a lot of work that other mainstream distributions don't bother with, instead leaving users to choose what they want post-install. As …

Nvidia Pascal GTX launch

Unless Nvidia has a massive last-minute change of heart, it will tonight unveil a couple of Pascal-powered GTX cards for gamers. Pascal is Nv's latest GPU design: word of it emerged in 2014, it finally arrived last month, and you can find deep dives into how it works and why here, and over here. So far Pascal has emerged in …
Chris Williams, 07 May 2016

Nvidia NVLink deep-dive

Our sister website, The Next Platform, has a detailed dive into Nvidia's NVLink technology – describing it as "transformative as SMP was for CPUs to GPU accelerators." NVLink glues Nvidia's high-speed number-crunching graphics chips to general-purpose processors, accelerating the rate supercomputers and other big machines can …
Chris Williams, 05 May 2016
Big cloud, image via Shutterstock

Alibaba says its AWS-a-like division embiggened by 138%

Alibaba’s cloud computing division Alicloud reported bumper growth of 138 per cent to $468m (£322m) for its full financial year 2016, as more paying customers flocked to the Middle Kingdom's AWS equivalent. As of March 31, 2016, AliCloud had more than 2.3 million customers, including more than 500,000 paying customers. The …
Kat Hall, 05 May 2016
Fan

Old fashioned engineering: HPC cluster kids would like to thank their fans. No really

HPC Blog Excitement reigned at the Asia Student Supercomputer Challenge Cluster Competition as Zhejiang University set a new student LINPACK record with 12.03 Tflop/s. With this result, they eclipsed the previous record of 10.78 Tflop/s set by Jamia Millia Islamia University (JMI) at last year’s ISC15 competition. Here's a historical …
Moments of perspiration

Sweaty students push hot HPC clusters through benchmark tests

HPC Blog The first task that students have to perform at the ASC16 Student Cluster Competition is to run the venerable HPL and newish High Performance Conjugate Gradients (HPCG) benchmarks. High performance Linpack (HPL), also known as LINPACK, is a routine that measures floating point performance and is the basis for the Top500 list. …
Boxers image via Shutterstock

Some HPC kids just can't help piling on the weight

HPC Blog We've saved the biggest for last in our round-up of the ASC16 Student Cluster Competition competitors. All of the teams below are driving clusters with more than nine nodes, most with several GPU accelerators. They're definitely power hungry beasts, which will require the teams to apply heavy throttling in order to stay under …
Tesla Model S sports sedan

Student cluster warriors build out 'roomy sedan' clusters

HPC Blog Four teams in the ASC16 Asian Student Cluster Competition decided to stake out the middle ground in the competition with clusters that aren't too small and aren't too big. Is their "just right" approach going to get them into the winners' circle at ASC16? Let's meet the teams and see what they're packing... Dalian University …
The Incredible Shrinking Man

HPC kids find bite-sized clusters are just as chewy

HPC blog Although the teams in the ASC16 Asian Student Cluster Competition are all using the same Inspur-provided gear, there's a LOT of variety in what they built. Cluster sizes in this competition range from three to twelve nodes, with anywhere from zero to eight accelerators. Let's meet the teams who believe that smaller is better …
android_toys_648

3-in-4 Android phones, slabs, gizmos menaced by fresh hijack flaws

Google has today issued a bundle of 40 security patches for its Android operating system. A dozen of the fixes correct critical vulnerabilities in versions 4.4.4 of the operating system and above. About 74 per cent of in-use Android devices run Android 4.4.4 or higher. These critical bugs can be potentially exploited by …
Iain Thomson, 02 May 2016
Handshake

Nvidia, Samsung pump brakes in car-crash GPU patent rip-off race

Samsung and Nvidia say they have agreed to a settlement that will end all patent litigation between the pair. Just hours before the US International Trade Commission (ITC) was set to rule on the duo's chip design infringement squabbles, the two companies have suddenly agreed to a deal that will dismiss their complaints before …
Shaun Nichols, 02 May 2016

GCC 6.1 hits, uses C++14

Version 6.1 of granddaddy C/C++ compiler suite GCC was released today, and is described as a major update. According to Red Hat's Jakub Jelinek: The C++ frontend now defaults to C++14 standard instead of C++98 it has been defaulting to previously, for compiling older C++ code that might require either explicitly compiling …
Chris Williams, 27 Apr 2016

Grab your Hammer pants – it's the '90s again: Facebook brings Virtual Reality back

Judging by the hype around Facebook VR, you'd think the '90s never happened. Back then, the Social Network™ didn’t exist and it was the era’s video game giants – Sega and Nintendo – rolling out virtual reality with great fanfare. However, both of the Japanese games giants cancelled their efforts before launch: Sega’s Virtual …
Mark Pesce, 13 Apr 2016

Large Hadron Collider gets 4,500 more data-crunching GPUs

Swiss super Piz Daint is getting an upgrade with 4,500 Nvidia Pascal GPUs, replacing 5,200 existing K20x accelerators. The iron, operated by the Swiss National Supercomputer Centre CSCS in Lugano, will have its 7.8 petaflop current performance doubled to nearly 16 petaflops. Announced at GTC16, the upgrade will also include …
Photo by Saul Bromberger for OpenPOWER Foundation

Power9: Google gives Intel a chip-flip migraine, IBM tries to lures big biz

OpenPower Summit IBM's Power9 processor, due to arrive in the second half of next year, will have 24 cores, double that of today's Power8 chips, it emerged today. Meanwhile, Google has gone public with its Power work – confirming it has ported many of its big-name web services to the architecture, and that rebuilding its stack for non-Intel …
Chris Williams, 07 Apr 2016
Venomous snake

Inside Nvidia's AI-in-a-box

Our sister site The Next Platform has more details on the DGX-1, Nvidia's deep-learning system-in-a-box that crunches 42TFLOPS using double-precision math. It costs $129k, consumes 3.2kW, and uses eight Pascal-powered Tesla P100 GPUs plus Intel Xeon CPUs: The system has two 16-core Xeon E5-2698 v3 processors, which run at 2.3 …
Chris Williams, 06 Apr 2016

Google, Rackspace to together unfurl DIY Power9 server designs

OpenPower Summit Google and Rackspace are working together on blueprints for servers that use IBM’s Power9 processors. The designs will be shared publicly via the Open Compute Project, it is hoped. The pair are due to announce their collaboration this week. The hardware will fit the 48V Open Compute racks Google is designing with Facebook. We’ …
Chris Williams, 06 Apr 2016
Harry Potter

Axellio! X-IO conjures up hyper-converged PCIe monster

Comment X-IO Technologies' Axellio* is a rackmount hyper-converged server/all-flash storage platform that will be sold through OEM, ODM and system integrators – and it provides an almost unbelievable combination of performance and density. The company has been working on the technology for two years, developing a third product line, …
Chris Mellor, 06 Apr 2016

Inside Nvidia's Pascal-powered Tesla P100: What's the big deal?

GTC16 So there it is: the long-awaited Nvidia Pascal architecture GPU. It's the GP100, and it will debut in the Tesla P100, which is aimed at high-performance computing (think supercomputers simulating the weather and nuke fuel) and deep-learning artificial intelligence systems. The P100, revealed today at Nvidia's GPU Tech …
Chris Williams, 06 Apr 2016
Large leaver switch on a board of push button switches

Nvidia's supercomputer-in-a-box needs 3.2kW of juice

Nvidia is going to sell a $129,000 3200W supercomputer-in-a-box called the DGX-1 from June. It has eight 16GB Pascal P100 GPUs, two Intel Xeons, 10GbE, and NVLink interconnects. It can run at 170TFLOPS if you use 16-bit floating-point numbers, or 42TFLOPS if you use 64-bit floating point. It sits in a 3U box and is aimed at …
Chris Williams, 05 Apr 2016