Tag: Nvidia

War between Nvidia and Samsung gets ugly

Newspaper Seller, 1939The war between Samsung and Nvidia has escalated and handbags are being distributed to the troops.

Samsung says that Nvidia has infringed several of its semiconductor-related patents and for making false claims about its products. This is a counter-suit following Nvidia’s lawsuit against the Korean company in September.

Samsung, which filed its lawsuit on Monday, is seeking damages for deliberate infringement of several technical patents, including a few that govern the way semiconductors buffer and use data.

Samsung said that Nvidia is guilty of false advertising when it says its “Shield” tablet sports the world’s fastest mobile processor, the Tegra. Samsung cites benchmarking studies performed by researchers at Primate Labs as proving that claim false.

Nvidia said it would review and respond to these new claims against it, and looked forward to presenting its case on how Nvidia GPU patents are being used without a licence.

Nvidia also pointed to a benchmarking study that supported its claim that the Tegra was the fastest mobile processor on the market.

Everyone knows that industry claims that “our chip is faster because we use x benchmark” always goes nowhere. It will be interesting to see what a court will make of them.

Nvidia and IBM get it together

Jülich SupercomputersIBM and Nvidia will work in collaboration with the Jülich Supercomputing Centre – a German institute for supercomputer simulation – to push the creation and optimisation of research apps on GPUs (graphics processing units) accelerated OpenPower systems.

A new centre will be opened to develop the high performance computing (HPC) space combining researchers from Nvidia and IBM and using the Jülich centre.

There are currently 70 members of the OpenPOWER Foundation formed late last year looking at new ways to develop supercomputers.

Stefan Kraemer, director of HPC business development at Nvidia said: “By providing systems combining IBM Power CPUs and Nvidia’s Tesla GPU accelrators via the NVlink high speed interconnect technology, we can help the new centre address both areas, and enable scientists to achieve new milestones in their research.”

IBM and Jülich have worked together since 2011 to create exascale architectures, and Nvidia has worked with Jülich since 2012.

The aim of both Nvidia and IBM is to create systems that will tackle the challenges of big data.

Nvidia turns in good results

nvidia-gangnam-style-330pxGraphics chip maker Nvidia posted higher fiscal third quarter revenue than many of the cocaine nose jobs of Wall Street expected.

Revenue in the fiscal third-quarter ended Oct. 26 was $1.225 billion, up 16 percent from the year-ago quarter, compared with Wall Street’s random guess of $1.202 billion.

For the current fourth quarter, Nvidia said it expects revenue of $1.20 billion, plus or minus two percent. Analysts on average expected fourth quarter revenue of $1.198 billion.

Third-quarter net profit was $173 million compared to $119 million a year ago.

Nvidia did better by focusing on using its Tegra chips to in entertainment and advanced navigation systems in cars made by companies including Volkswagen’s Audi, BMW and Tesla.

In the third quarter, revenue from Tegra chips for automobiles and mobile devices jumped 51 percent to $168 million. Nvidia’s PC graphics chip business expanded 13 percent to $991 million.

Qualcomm has China crisis

china-syndrome-one-sheet1Qualcomm is facing a little trouble in Big China as it is starting to look like its antitrust investigation is going pear shaped. Meanwhile problems collecting royalties could harm its business in China next year.

To make matters worse it is facing similar investigations in the United States and Europe.

Qualcomm should be making a large profit in China. The country is expanding high-speed 4G network is driving demand for smartphones with leading-edge technology.

But it looks like Qualcomm could face a fine of more than $1 billion in China as a result of the National Development and Reform Commission (NDRC) investigation, and the company could be forced to make concessions that would hurt its highly profitable business of charging royalties on phones that use its patents.

Qualcomm admitted that it faces a new probe by the European Commission about rebates and other financial incentives in the sale of its chips. Another preliminary investigation by the U.S. Federal Trade Commission concerns a potential breach of licensing terms.

Qualcomm President Derek Aberle said that his company was co-operating with the Chinese to come up with potential ways to resolve the problem.

Qualcomm has also been struggling to collect licensing revenue from some device makers in China, including local manufacturers the US chipmaker has done little or no business with in the past.

But the fear is that concessions on royalties that Qualcomm is forced to make in China could spread to manufacturers in other countries.

Qualcomm said it was difficult to predict the outcome of the U.S. and European investigations.

The European probe is separate from a four-year-old complaint to the European Commission from a subsidiary of Nvidia over alleged patent-related incentives and exclusionary pricing by Qualcomm.

Qualcomm forecast revenue for fiscal 2015 of between $26.8 billion and $28.8 billion. Analysts on average expected $28.91 billion.

The chipmaker reported revenue of $6.69 billion for its fiscal fourth quarter, ended Sept. 28, up 3 percent from the year-ago period. Analysts on average had expected $7.016 billion.

Qualcomm posted fourth-quarter net income of $1.89 billion, up 26 percent from a year ago.

AMD faces Nvidia threat

AMD, SunnyvaleNvidia is ahead of AMD on the graphics front and it won’t be until next year that the Sunnyvale firm catches up.

That’s according to financial analyst Sean Chandler, who works for Seeking Alpha.

He said in a note to his clients that the Nvidia “Maxwell” architecture has put AMD under heavy pressure in the consumer graphics arena.  And that’s worry investors and taking its toll on the AMD share price.

Chandler said that while AMD’s restructuring are widely seen as positive, the firm “still needs to release competitive technologies to remain relevant”.

Nvidia Maxwell, he said, means “monumental advancements” in both efficiency and performance.

Nvidia’s 60watt 750Ti is comparable to AMD’s 150 watt R7 260X, he said.  And Maxwell also outforms AMD in performance efficiency per die size.

He added that the rumour mill suggests AMD may respond with 20 nanometre chips now chip foundry TSMC has got the shrink down pat.

He suggests to investors: “AMD is almost certainly not out of the game, but be cautious and don’t pour all of your eggs into one basket.”

Tesla car is a big computer

Tesla Model SThe people at market research company IHS are known for taking apart PCs, iPhones and iPads to find out what the bill of materials (BOM) is.

Now IHS has gone that little bit further and taken a Tesla Model S apart.  We guess a Tesla Model S would make a reasonable meal for Monsieur Mangetout but it doesn’t appear he was called in to have a snack.

IHS suggests that the Model S is more of a computer than a car.  Andrew Rassweiler, director for materials at IHS, said: “It’s like looking at the components from the latest mobile device from an Apple iPad or Samsung Galaxy. When it comes to the user facing segment of the Model S’s electronics, the company has radically departed from business as usual in the automotive market.”

He said the cost structure of the electronics, the use of large displays, mobile microchips suggess the Model S is more like a smartphone than a traditional car.

Components include a 17-inch display, an Nvidia Tegra 3 1.4GHz quad core microprocessor and a OM twice the cost of the highest end “infotainment” unit.

Rassweiller said that the display is 10 inches larger than most car head units, with a resolution of 1,920 by 1,200.

IHS said it hasn’t finished taking the car apart yet, so perhaps Monsieur Mangtout – who once ate a plane – has a treat in store.

Nvidia wrestles with ARM connections

arm-wrestlingARM Holdings Chief Executive Officer Simon Segars defended his smartphone graphics technology which Nvidia claims it invented.

Nvidia is currently taking Samsung Electronics and Qualcomm to court for using the technology in its phones and accusing both companies of infringing its property patents on graphics chip technology.

Nvidia said Samsung devices made with graphics technology from ARM, Qualcomm and Imagination Technologies illegally use its intellectual property, or IP.

Segars said that the company stood behind its IP and will work with its partners when something like this happened.

Nvidia is not suing ARM or Imagination yet but it did say it would ask the US International Trade Commission to prevent shipments of Samsung devices containing ARM’s Mali or Imagination’s PowerVR graphics architectures, as well as Qualcomm’s graphics technology.

Nvidia has to play this carefully. Nvidia depends on ARM’s technology to make its Tegra chips for tablets and cars.

Segars said that it did “create a bit of a curious situation… But we do a lot of business with a lot of people.”

AMD ready to bang bang Maxwell’s silver hammer

maxwellMystery surrounds the launch of a new AMD product this week, with pundits suggesting that it might be a new GPU.

AMD released a GPU related teaser comes a day after NVIDIA showed off its new  Maxwell graphics card which include the GeForce GTX 980 and GeForce GTX 970. The launch was important because it is supposed to be a new era of powerful and highly efficient graphics cards. AMD might have an answer to that.

It released a tweet connected to its matrix pills campaign to market the new GPUs — one is blue and one is red. A similar marketing campaign was used during the launch of the HD 7770 GHz graphics card.

The Verdetrol pills were used as a teaser for the launch of the first GHz edition graphic card from AMD but they did another teaser with the Radeon R9 295X2 where they used two packs of chips and a water bottle to indicate their water-cooled dual GPU solution.

Since the marketing is similar there’s a blue pill which might indicate a water cooling solution such as the one that was leaked a while ago and the red pill may indicate the Radeon chip which will be used to power the graphics card based on the GCN architecture.

Smart money is that it is the launch of the new Radeon R9 285X sometime in late September.

AMD is not just messing around with the pills either. AMD has a teaser for a FirePro product where they ask “Can you name our first product that processed graphics independently of the CPU”.

This could be related to either a Radeon or a FirePro product.  It will be a year since AMD has introduced most of their lineup next month and will probably be the best time for AMD to offer new replacements.

Nevertheless, AMD has to do something to tackle Nvidia’s Maxwell and if has new high-end chips ready then it will need to play them fast.

Next year’s server wars may be cancelled

soldiers-2The major battle in the server space planned for next year may be only a minor skirmish with the usual suspects winning.

Intel needs to see off the expected competition from ARM and is going to chuck a lot more cash in the area to keep its position as market leader. What we are seeing from the Intel Developer Forum is that its answer will be a a new Xeon D family of chips.

Xeon D chips will be the first server chips based on the Broadwell architecture, and will go into dense servers starting next year. But these are not your normal server chips, they are effectively systems on a chip which means that they will be deliberately targeting anything “low level and power efficient” that ARM is expected to come up with.

It means that Intel does not think that its Xeon E3 and Atom chips code-named Avoton will be up to the task of taking on ARM. The Xeon D chip will be faster, but more power hungry than Avoton, which is based on an architecture called Silvermont used in mobile chips.

But Intel thinks that the Xeon D will provide more performance-per-watt, which punters will find attractive.

Intel does have some other advantages in any coming server war. Intel’s chips already go into more than 90 percent of servers, and server makers like Dell have said that the chances for success of ARM servers are diminishing due to product delays. Intel also has a head-start on software development over ARM.

ARM’s server chips are based on the ARMv8 architecture, and have integrated networking, storage and I/O controllers. Its key weapon against Intel is still lower power consumption, something Chipzilla is fast catching up on.

A variety of companies had indicated interest in making server processors based on blueprints from ARM,  but so far ARM 64-bit server processors have not been made available commercially.

Chip makers like Applied Micro and Advanced Micro Devices have delayed shipment of ARM-based chips.

Dell is offering prototype ARM servers for benchmarking and application development. Hewlett-Packard announced plans to use ARM processors in its Moonshot “dense” server, which uses x86 chips, but hasn’t announced a definitive release date for the ARM edition.

The other player in any coming war AMD is also expanding its low-power server processor lines,  which could also will hurt adoption of ARM servers.

The other big hurdle for ARM is the fact that most firms already have software and hardware based around x86. To adopt ARM-based servers, companies will not only have to invest in new servers and components, but also port applications to the architecture.

This could make a switch to ARM very expensive in terms of capital and final cost of ownership. Then there are some licensing issues surrounding the adoption of ARM servers, as companies will have to pay more for software per core used in them, Norrod said.

ARM is also finding its allies thin on the ground. ARM server pioneer Calxeda folded operations and earlier this year Nvidia scrapped server chip plans. Samsung has also abandoned ARM server chip development.

EU watchdog bites Qualcomm’s rump

AnubiIt looks as if the EU is going to back Nvidia’s complaint against Qualcomm and investigate the chipmaker for alleged anti-trust shenanigans.

Nvidia has been moaning about Qualcomm for nearly four years and the investigation coincides with a similar case in China into the chipmaker’s monopoly practices.

If found guilty of breaching EU rules, the company could face a fine of up to $2.5 billion.

Reuters said that the Commission may open a case after the summer.

The case centred on the British mobile phone chipmaker Icera which was bought by Nvidia in 2011.

While no one said what happened to Icera, it appears that the company accused Qualcomm of using patent-related incentives and exclusionary pricing of chipsets to discourage customers from doing business with it.

No one seemed to care that much and the issue appeared to have faded from the Commission’s agenda. However, a recent case where Europe’s second-highest court in June upheld a record 1.1 billion euro EU fine against Intel for abuse of its dominant market position made the regulators realise that they were sitting on a nice little earner.

Companies can be fined as much as 10 percent of their global revenues for breaching EU antitrust rules.

But the case is a long way off being resolved and anything could happen. In 2010, the EU competition authority scrapped a four-year probe into Qualcomm after Ericsson and Texas Instruments withdrew their objections against the company.

Nvidia does rather well

nvidia-gangnam-style-330pxNvidia posted higher second-quarter earnings and gave a forecast for current-quarter revenues that exceeded what the cocaine nose jobs of Wall Street predicted.

This was surprising given that some analysts were worried that PC shipments were flat in the June quarter.

Nvidia’s graphics chips for PCs make up most of its business but what appears to have saved the company’s bottom line was that it has been selling to car makers and data centres.

In the second quarter, revenue from Tegra chips for automobiles and mobile devices jumped 200 percent to $159 million.

After struggling to compete against larger chipmakers like Qualcomm in smartphones and tablets, Nvidia has increased its focus on using its Tegra chips to power entertainment and navigation systems in cars made by companies including Volkswagen’s Audi, BMW and Tesla.

Nvidia in July launched its own tablet aimed at game enthusiasts, called Shield, with Tegra chips and other high-end components. This went against the industry trend toward commoditized, inexpensive devices.

Nvidia has been doing well in the cloud by flogging its chips to IBM, Dell and HP as part of their datacentre product range.

Predictions are that Nvidia’s GRID graphics technology for data centres will also do well after it has been tested by other potential enterprise customers.

Nvidia reported second-quarter revenue of $1.1 billion, up 13 percent from the year-ago quarter as it expanded its focus on cars and cloud-computing.

For the current quarter, Nvidia said it expects revenue of $1.2 billion, plus or minus 2 percent. Analysts on average expected second-quarter revenue of $1.1 billion and third-quarter revenue of $1.16 billion.

Nvidia’s net income in the second quarter, which ended on July 27, added up to $128 million or compared with $96 million in the year-ago quarter.

Nvidia shows off its SHIELD tablet

nvidia-gangnam-style-330pxNvidia has officially announced its new SHIELD tablet  and SHIELD wireless controller which is powered by its powered by their Tegra K1 System on a Chip.

The graphics chip maker has been developing the Android-based SHIELD portable to create what appears to be a mix of SHIELD portable and the Tegra Note 7, but featuring updated technology and better materials.

All this is supposed to create a gaming-oriented tablet to consumers with the screen de-coupled from the controller.

The original SHIELD portable and the Tegra Note 7, were built around Nvida’s Tegra 4 processor. The SHIELD tablet, however, is powered by the newer Tegra K1, which features quad ARM A15 cores and 192 Kepler-class graphics cores.

The SHIELD tablet’s specifications makes it one of the more powerful tablets on the market. It has 2GB of RAM and an 8”, full-HD IPS display, with a native resolution of 1920×1200. There are also a pair of 5MP cameras on the SHIELD tablet (front and rear), 802.11a/b/g/n 2×2 MIMO WiFi configuration, GPS, a 9-axis motion sensor, and Bluetooth 4.0 LE.

There is a WiFi-only version and a 32GB version coming with LTE connectivity as well. N

Nvida  is using a mostly untouched version of Android to run the thing with a few pre-installed applications, and control panels for the SHIELD’s gaming features.

The gaming features include the ability to stream games directly to Twitch, Nvida’s ShadowPlay technology, PC Streaming, Nvida GRID cloud gaming support, and access to the new SHIELD hub.

The SHIELD tablet is also one of a select few mobile devices that has been certified to stream Netflix 1080p content.

 

Currys and PC World to peddle Tegra Note 7

tegra-tabIn a rather surprising development, Currys and PC World will be selling Nvidia’s Tegra Note 7 tablet under their own brand.

The Advent Vega Tegra Note 7 as it’s known, will be available in stores and online from November 15, but pre-orders will be available tomorrow. The price is £179, slightly less than the Nexus 7, but quite a bit more than Tesco’s Hudl.

Although it has a different brand, the hardware is identical – Tegra 4 SoC, 7-inch 1280×800 IPS display, 1GB of RAM, 16GB storage and a stylus to boot. The only difference is the sticker.

Here is the interesting part. Although we speculated that some players could sell the Tegra Note 7 under their own brand, when it was announced, Nvidia said it would be sold by five AIBs, one or two handling every major region. Nvidia never mentioned retailers.

If Currys and PC World and willing to give it a go, and if Nvidia is happy to offer such an arrangement, they might be the first of many retailers to adopt Nvidia tablets and sell them under their own brand.

In addition to the Tegra Note 7, Nvidia is also working on a few other Android devices. The Shield console is already out, although it lacks the mass market appeal of tablets. However, Nvidia is also working on a reference phone platform (Phoenix) and rumour has it that other tablets are in the works, too.

AMD bullies Nvidia with $399 Hawaii card

radeon-r9-290A few weeks ago AMD introduced its Volcanic Islands products at an event in Hawaii. Most of the line-up were just rebrands, but the flagship R9 290X and R9 290 weren’t. 

The Hawaii cards are based on all new silicon, 6.2bn transistors crammed onto a 28nm die. AMD did not announce the prices at the event, but a couple of weeks later it launched the R9 290X at $549. The price was lower than expected and it forced Nvidia to slash the price of the GTX 780 by $150.

Just as Nvidia countered the R9 290X, AMD decided to make its life miserable once again. The Hawaii Pro version of the card, the R9 290, launched at $399 – making it $100 cheaper than the GTX 780, which went down from $649 to $499 last week.

There is, however, a slight problem for Nvidia. The R9 290 ends up significantly faster than the GTX 780 and in some cases it can even give the $999 GTX Titan a run for its money.
So, the new card is $100 cheaper than what Nvidia has to offer, yet it’s faster. There is one problem though, reviewers report the R9 290 can get very loud, but it seems like a small price to pay considering the price/performance ratio. In addition, it’s only a matter of time before AIB partners come up with non-reference designs, with custom coolers to keep the noise down.

Nvidia was already forced into two price cuts following AMD’s launch. First it slashed the prices of its sub-$199 products to compete with AMD’s rebranded R7-series. Then it slashed the prices of the GTX 780 and GTX 770, only to be undercut by AMD’s new $399 card. Most punters were expecting the R9 290 to end up at ~$449, but like we said last month, AMD had a couple of good reasons to launch it at $399 – and it did.

Nvidia simply doesn’t have much wiggle room left. Perhaps it’s feeling a bit like Guy Fawkes, and hoping bonfire night is merely a damp squib.

Intel seems to have lost its way

Sean MaloneyThe news that Intel’s Galileo is on its way just underlines to me how the chip giant has lost its way.

The “open source” computer costs $70, and uses its Quark microprocessor. Intel clearly thinks it will compete against the highly successful Raspberry Pi but clearly it hasn’t got a chance to play catch up.

The launch mirrors Intel’s late attempt to climb on the tablet bandwagon by cutting the price of its Atom microprocessor to compete with ARM and Nvidia based chips.  But it hasn’t got an earthly here, either.  Manufacturers are very chary about using anything with the Intel name associated with the tin. Again, that’s underlined by vendors’ reluctance to be associated with Intel.

Cheap is everything in the tablet market now and even though Intel’s chips might be, er, cheap as chips, the economics of this don’t really make a lot of sense to anyone. Sure, Intel has heaps of capacity but that in itself is part of the problem. State of the art fabs are really expensive these days and the volume game just doesn’t fit Intel’s business model.

In reality, the chip giant really has very little new to say. The new broom in the shape of CEO  Brian Kzanic appears to be attempting the Herculean task of cleaning the Augean Stables not just of the dung but also of a heap of very good people who have let their legs do the walking.

Datacentre business no doubt is still healthy for Chipzilla, but on the other hand independent market research shows that the notebook market is on the wane.  Sure, enterprises will refresh their notebooks but with the arrival of BYOD, there’s a level of ambiguity which must leave Intel more than a little bemused.

In truth, Intel has had zilch to say in the last three years as smartphones and tablets transformed the “traditional” Wintel model.

As part of the antitrust agreement following the demise of DEC, Intel found itself with StrongARM devices. At the time, we asked top executives from the firm why it didn’t just cut the Gordian Knot and produce a highly portable ARM based device?  The answer, of course, was that Intel was on the Centrino notebook gravy train. Sean Maloney, now a non-executive director at Chinese foundry SMIC, realised that the Atom chip might well cannibalise the notebook market but nobody at Intel appeared to have looked further than the next three quarters and see its dominance becoming more and more eroded.

Of course, Intel has oodles of cash in the bank but oodles don’t last forever.  Re-engineering its business model is, for Intel, a far from trivial task. As an Intel watcher for the last 30 years, I will be most interested to see what happens in the next 12 to 18 months.