Nvidia appears to be readying new members of its professional Quadro range of graphics cards in time for the August SIGGRAPH 2015 event.
According to VideoCardz the latest Nvidia graphics driver for the Quadro range of graphics cards includes a telltale text string revealing Nvidia’s dark satanic intension.
The new driver adds support for:
NVIDIA_DEV.13F0.1152.103C = “NVIDIA Quadro M5000″
NVIDIA_DEV.13F1.1153.103C = “NVIDIA Quadro M4000″
NVIDIA_DEV.13F2= “NVIDIA Tesla M60″
At the same time last year Nvidia revealed details of the Quadro Kxx2 and since SIGGRAPH 2015 scheduled to start on 9th August (runs until 13th Aug, takes place in downtown LA) it looks like the latest Quadro graphics cards will be also launched.
Of course driver strings do not reveal much detail however the Quadro professional graphics range will be based upon the Maxwell architecture GM204GL graphics processor. We are expecting a 256-bit GPU with a a 4GB or 8GB frame buffer versions.
Then there is the Tesla M60. This GPU based general purpose computing product is expected to be based upon a fully-fledged GM204 GPU with 8GB of GDDR5 memory.
Nvidia is about to release a range of price cuts in a bid to see off AMD in the longer term.
While the price cuts have already happened in the US, in the EU Nvidia’s GTX 980, GTX 980Ti and Titan X are still kept high because people were buying them at the prices the Green Goblin was asking.
In the US where competition between AMD and Nvidia is tighter, the prices dropped by 10 per cent. Ironically since Europeans are more loyal to Nvidia in the high-end graphics cards market the outfit decided they could continue to pay.
According to Kitguru the new R9 Radeon 300 series appears to have upset the apple cart. The cards have been launched at similar prices or lower than Nvidia’s top tier products. Apparently Europeans were thinking of going cheaper since the Green Goblin did not seem to admire their loyalty.
It could force AMD to drop its prices as it can’t remain competitive selling top-end graphics at prices higher than Nvidia’s while having weak selling figures in non-US countries.
It will force AMD to sell its freshly launched Fury X at prices lower than planned, and for such a new card this move damage AMD. Nvidia was expected to drop prices of course, but only for its lower-end products like the 700 or 600 series.
Red Hat has announced the release of OpenShift Enterprise (OSE) 3, a new version of its Platform-as-a-Service offering.
Based on Red Hat Enterprise Linux (RHEL)7, Openshift is built on Docker Linux containers with Kubernetes orchestration using technology developed in collaboration with Google.
The news comes in a busy week for Red Hat, which has also announced a new productivity tie-up with Samsung and taken a leading role in the formation of a new alliance known as the Open Container Project to standardise containers.
Users will have access to a wide range of apps via the Red Hat Container Certification Programme. Middleware solutions including Red Hat JBoss Enterprise, Web Server (Tomcat) and JBoss A-MQ messaging are also included.
Included are a number of tools to help developers create and collaborate, with web, command line, and integrated development environment interfaces. Options include direct code-push from GIT and source to image building. There is also flexibility for deployment, rollback and integration.
In addition, a preview of Openshift Dedicated has been released. The public cloud service based on OpenShift 3 will succeed Openshift Online, which already hosts 2.5 million applications online, allowing businesses to quickly build, launch and deploy bespoke apps.
Ashesh Badani, vice president and general manager, OpenShift, Red Hat, said, “This release of OpenShift Enterprise 3 employs open source containers and orchestration practices to change the developer experience and move the platform in the direction of what customers are asking for – a flexible platform for a microservices architecture.
“Our continued upstream work in the Docker and Kubernetes communities enable us to deliver the most updated technology platform for developers and operators, enabling them to remain competitive through quicker innovation.”
To assist users, Red Hat is offering a range of enterprise administrator courses to teach users how to deploy, configure and manage the system, which can result in a Red Hat Certificate of Expertise in Platform as a Service – a worthy certificate for any office wall.
OpenShift 3 is available now with bespoke pricing models based of socket and core pairings.
AMD’s Project Quantum PC system, with graphics powered by two of the new Fiji GPUs may have got the pundits moist but it has been discovered that the beast has Intel inside
KitGuru confirmed that the powerful tiny system, as shown at AMD’s own event, was based upon an Asrock Z97E-ITX/ac motherboard with an Intel Core i7-4790K ‘Devil’s Canyon’ processor.
Now AMD has made a statement to explain why it chose to employ a CPU from one of its competitor in what is a flagship pioneering gaming PC.
It told Tom’s Hardware that users wanted the Devil’s Canyon chip in the Project Quantum machine.
Customers “want to pick and choose the balance of components that they want,” and the machine shown off at the E3 was considered to be the height of tech sexiness right now.
AMD said Quantum PCs will feature both AMD and Intel CPUs to address the entire market, but did you see that nice Radeon Fury… think about that right now.
IT is going to be ages before we see the first Project Quantum PCs will be released and the CPU options might change. We would have thought that AMD might want to put its FinFET process ZEN CPUs in Project Quantum with up to 16 cores and 32 threads. We will not see that until next year.
Based around the Cortex-A7 cores and Cortex-M4 MCUs, the pair have lower power consumption than the predecessor the i.MX6.
The single-core, 800MHz i.MX7 Solo (i.MX7S) and dual-core, 1GHz i.MX7 Dual (i.MX7D) are the first use the Cortex-A7.
The reduced power consumption has happened at the expense of a performance reduction. The up-to-1GHz Cortex-A7 cores are slower than the i.MX6′s up to 1.2GHz Cortex-A9 cores. In addition, there’s no mention of the earlier Vivante GPUs or 3D acceleration. Like the UltraLite, there’s only a simple 2D image processing engine.
Freescale said the i.MX7′s Cortex-A7 and Cortex-M4 cores have a core efficiency levels of 100 ?W/MHz and 70 ?W/MHz, respectively. The SoC’s overall power efficiency is 15.7 DMIPS/mW, and a new Low Power State Retention (LPSR) mode runs at 250 ?W. In LPSR sleep mode, the i.MX7 consumes only 250 ?W, while supporting DDR self-refresh mode, GPIO wakeup, and memory state retention.
The savings are down to the newer Cortex-A7 architecture and a 28nm “ultra low leakage process,” as compared to the i.MX6′s 40nm process. The i.MX7 also features a new discrete power domain architecture.
The i.MX7 ships with Linux, and supports Android, and is aimed at wearables, Point-of-Sale gear and smart home controls.
The i.MX7 SoCs are paired with a new Freescale PF3000 PMIC which has features up to four buck converters, six linear regulators, an RTC supply, and a coin-cell charger. The chip is supposed to optimize peripheral power delivery, system memory and processor cores. The PMIC also supports one-time programmable memory for controlling startup sequence and output voltages.
The i.MX7 has a Cortex-M4 microcontroller unit (MCU) core for offloading processing. The Cortex-M4 can run Freescale’s own MQX, at up to 266MHz, compared to 200MHz on the SoloX.
AMD has denied rumors that it will split up its business in a bid to rival chip giant Intel.
Sources close to AMD said that the company is planning to divide into two, or spin off certain parts of the business, in order to make it a stronger force in the chip industry.
Three sources familiar with the matter told Reuters that AMD is at the initial stage of reviewing, and has asked a consulting firm about its options in trying to turn the company around. The sources said that the deliberations are preliminary and that no decision has been made.
Apparently, one of the options under consideration is separating AMD’s graphics and licensing business from its server business, which sells processors that power data centres.
However, AMD has since denied such speculation. “While we normally would not comment on such a matter, we can share that we have no such project in the works at this time,” a company spokesperson said.
“We remain committed to the long-term strategy we laid out for the company in May at our Financial Analyst Day.”
If the rumours happen to be true, though, and AMD does split its company into two, it would make it the second major technology firm to do so in the past year.
PC maker HP confirmed in October that it would tear itself in half, making two new publicly traded Fortune 50 companies. The decision was down to the firm wanting to focus on the faster growing side of the business.
HP’s split is expected to be completed by the end of this year. HP’s enterprise technology infrastructure, software and services businesses will do business as Hewlett Packard Enterprise, and the firm’s market-leading personal systems and printing businesses will operate as HP Inc and retain the current logo.
Intel’s 14nm Skylake processors and 100-series chipsets for desktops are expected to be unveiled in early August at Gamescom in Germany.
The dark satanic rumour mill has manufactured a hell on earth rumour that Intel will lift the kimono in its Core i7-6700K and Core i5-6600K as well as Z170 chipsets.
The reason that Gamescom was selected was because it is initially targeting gamers.
Intel is expected to unveil its Skylake-based Core i7-6700/6700T, Core i5-6600, 6500, 6400, 6600T, 6500T and 6400T, and H170 and B150 chipsets between August 30-September 5.
The H110 chipsets are scheduled to be released between September 27 and October 3, the Q170 and Q150 chipsets are set for October or November.
Motherboard makes will start announcing their 100-series-based motherboards on August 5. The Skylake platform adopts an LGA1151 socket design and supports DDR4 memory.
Notebook versions of Skylake will be in the shops in the run up to Christmas. Chipzilla’s notebook reference design for the Skylake platform includes support for USB 3.1 Type-C ports.
The Fiji GPU behind the AMD Radeon Fury line of graphics cards is definitely a one big chip with 8.6 billion transistors, but during our short chat with, Raja Koduri, Corporate Vice President and CTO, Visual Computing at AMD and we learned that the number is even higher.
Raja tells us that when we add the number of High Bandwidth Memory transistors that are sitting on the interposer, you get to more than 10 billion transistors for the Fiji GPU interposer chip.
This is a lot of transistors but again, the Fury graphics card is much smaller than the one based on GDDR5 memory, and this is what enables new form factors such as Fury Nano that fits on 6-inch (15.24 cm) size PCB.
The Fiji GPU packs 4096 Stream Processors and this is the highest number we saw until today and with aggressive prices, AMD will certainly put a lot of pressure on Geforce GTX 980 Ti sales. We still have to see what happens with the actual real-world performance as we expect to see the actual Fury X graphics cards in our systems before the end of this week.
Just for reference, Nvidia’s Titan X as well as the Geforce GTX 980 Ti both have 8 billion transistors, some 600 million less than the Fiji GPU.
Samsung Electronics has told the world that owns the largest number of patent rights essential for long-term evolution (LTE) technology in the world.
Writing in its official blog “Samsung Tomorrow” that it has more than 3,600 standard essential patents (SEP) for the LTE and LTE-Advanced (LTE-A) technology. That is 17 percent of all LTE-related SEPs.
We guess this means that if someone buys an LTE phone more than 17 per cent of the money which goes to buy patents should end up in Samsung’s bank account.
Samsung Electronics Digital Media & Communication Laboratory’s intellectual property application team head Lee Heung-mo said Samsung Electronics has established a solid foothold as the global leader and the first mover in the fourth-generation mobile telecom market.
“This also means that the company has become able to provide more convenience to customers by developing the latest technologies.”
The Taiwanese patent office conducted market research for the nation’s state-run National Applied Research Laboratory based on about 6,000 patent rights listed at the Patent and Trademark Office in the United States during the last two years.
LG Electronics and Qualcomm followed Samsung Electronics in second place with 14 percent of SEPs, each. Ericsson, Panasonic, Nokia and NTT DoCoMo hold the third spot with 5 percent, each.
Pantech, the nation’s third-largest handset maker which currently faces bankruptcy, held only one percent, while Korea’s state-run Electronics and Telecommunications Research Institute owned less than 1 percent, the report showed.
During the patent dispute with Apple, the U.S. International Trade Commission said Apple had infringed on Samsung Electronics’ SEPs though they had to be shared under a “fair, reasonable and non-discriminatory” principle.
Samsung Electronics said it has pushed for securing the SEPs in this sector during the last 18 years and has competed with global telecom giants including Qualcomm, Nokia and Ericsson as a relative latecomer. It said securing leadership in SEPs may change the crisis of facing patent disputes to diversifying income sources.
Now that Intel and Altera have tied the knot, speculation is rife that AMD is going to merge with Xilinx.
The speculation appears to have come from Steve Casselman, the CEO of Reconfigurable Computing but since then others have started to also think it is a good idea.
Thing is that onces Intel has its claws into the Altera product line, it will be able to cheaply produce the chip and drop the price, drastically undercutting Xilinx.
Xilinx get any help from AMD, since AMD is fabless, but it would open up the opportunity for Xilinx to have an x86 and FPGA option. AMD’s server position is pretty pants these days but it would give Xilinz a way in.
What is more likely to interest Xilinz is AMD’s GPUs where it still has some relevance at least in the short term.
Gus Richard, formerly with Piper Jaffray and now with Northland Capital Markets feels AMD is a takeover target and Xilinx is a suitable buyer.
“Investors have decided that AMD has hit the iceberg and they are waiting for the ship to sink,” he said, he told Barron’s. “However, we believe that AMD still has time to avoid the collision, and if all else fails investors will be bailed out by the lifeboat of M&A.”
Xilinx could use AMD as a server/data center play. Again, that overlooks AMD’s pitiful position in the data center, but even that could change with Zen.
Richard thought it possible that the Chinese could buy AMD, which is unlikely because Intel holds a license agreement with AMD on x86 patents which can’t be shared with a Chinese company. Intel and the DoJ would scream blue murder.
Intel has suddenly made some interesting hardware less interesting to open sourcers by insisting that its i915 DRM kernel graphics driver for upcoming Skylake and Broxton hardware demands some binary-only firmware blobs.
According to Phoronix these first i915 DRM firmware blobs are for Skylake and Broxton for the GuC and DMC.
DMC is the Display Microcontroller used by Skylake (Gen9) within the display engine to save and restore its state when entering into low-power states and then resuming. It saves and restores display registers across low-power states separate of the kernel.
Intel said that the firmware blobs are required by the DRM driver rather than being an optional add-on.
The license of these firmware blobs also indicate that redistribution is only allowed in binary form without modification. Beyond that, “no reverse engineering, decompilation, or dis-assembly of this software is permitted.”
Basically this will kill off any desire for Open Source enthusiasts to touch Skylake, although we doubt Intel will be too worried – they are a very nice couple. In any event AMD apparently uses something similar to protect bits of its operation.
Still Intel is shipping these firmware files early so everyone knows they are there.
While MediaTek might be known for its multi-core smartphone processors, the firm was very keen to show off its more adventurous side at Computex 2015.
With a booth almost entirely dedicated to the latest and greatest from its new Labs division, which aims to bring the latest innovations from developers to market, MediaTek offered something a little more unexpected compared to previous years.
Launched in autumn last year, MediaTek Labs is a worldwide initiative to help developers of any background or skill level to create and market wearable and Internet of Things (IoT) devices.
With the firm’s LinkIt Development Platform, based on the MediaTek Aster (MT2502) chipset, sitting at its core, the Labs programme provides developers, makers and service providers with both software and hardware development kits, technical documentation and business support.
Here’s a few of our favourite innovations showed off at Computex, based on either the LinkIt One platform, or the firm’s fresh Helio P10 smartphone family of SoCs.
Heart rate monitoring smartphone camera
This “contactless heartrate monitoring” technology is powered by the firm’s Visual Processing Application in its latest P10 smartphone SoC.
It makes use of a smartphone’s video camera to take a heart rate reading via the front-facing camera by stripping down the layers of the image taken by the camera in real-time to detect the pulse in a user’s temple.
We were rather dubious about how well this might work, so gave it a go. While it took a good few seconds to match up, you can see from the photo that it is almost as accurate as the portable ECG monitoring device we had clipped on our finger. Impressive stuff.
Winning first prize in the ITRI Mobilehero competition in Taiwan last year, this nifty IoT wine brewing device was developed by a local start-up called Alchema.
It consists of five sensors thatmonitor the alcohol content and the brewing environment. The results we tasted were, shall we say, interesting, if a little on the sharp side.
Alchema looking to raise more funds on Kickstarter before the end of the year.
Another LinkIt-powered device MediaTek showed off at Computex was a wearable aimed for the elderly. Using Bluetooth and accelerometer sensors, the wristband tracker detects the users’ wrist motions and raises an alarm, alerting those that are linked to the watch via a smartphone app if their elderly family member, loved one or friend’s device has detected a sudden movement that could resemble a fall or accident.
Sitting at the more mature end of the LinkIt developer platform spectrum, but still less than a year old, is an electric-scooter rental company called Skuro Moto. We spoke to its chief executive Frank Chen, who is running the company at the tender age of 24 after developing the idea while at university.
Skuro works with electric-vehicle maker Ahamani EV Technology to provide a rental service at Yuan Ze University in Taiwan. The bikes reduce costs for riders by about 30 percent thanks to a monitoring system enabled by the LinkIt chip that lets riders see their power usage. They can also be started by a swipe of a student identity card, to save the trouble of lost keys.
MediaTek wants to grow its 4G chip market share in China during the second half of 2015 with the aim of scoring half the market.
According to Digitimes the outfit currently has 20 per cent of China’s 4G chip market and is looking to push its 4G chip market share in China to more than 50% in the second half of 2015.
This is a step up from earlier targets as MediaTek president Hsieh Ching-chiang remarked recently that the company is aiming to reach a 40 per cent share. That comment was only a few months ago, and so any 50 per cent target means that its ambitions must be being more than met.
MediaTek just released the P10 series of its Helio family for mid-range 4G smartphones. The Helio P10 is a 64-bit SoC based on eight Cortex-A53 cores and clocked at 2GHz, and will be available in the third quarter of 2015.
MediaTek said that using TSMC’s 28nm HPC+ process, the Helio P10 can save up to 30 per cent more power compared to existing smartphone SoCs manufactured using 28nm HPC process.
MediaTek previously announced its 10-core Helio X20 series, which integrates a multi-mode Cat 6 LTE modem.
TSMC’s 16nm FinFET process has barely gotten off the ground, but the foundry is already talking about 16nm FinFET Plus, which is due to launch by the end of the year.
The improved 16nm FinFET Plus (16FF+) node is supposed to deliver more efficiency and performance, making TSMC’s node more competitive compared to Samsung’s 14nm node. That is the general idea, but TSMC’s first generation 16nm node has failed to impress in terms of design wins.
TSMC president CC Wei said the new 16FF+ node already has 20 tapeouts, ten of which achieved satisfactory yield performance. Wei said the company expects up to 50 tapeouts by the end of the year. TSMC expects 16FF+ to enter commercial production in the second half of the year.
16FF+ is not the only FinFET node coming from TSMC over the next year. The company plans to introduce 16FFC for compact devices sometime in the second half of 2016. In addition, 10nm FinFET is expected to enter risk production by the end of 2015, reports Digitimes.
TSMC has said that it is confident that it can beat Samsung Electronics in ramping up production on its 10nm lines.
Samsung disclosed during a recent technology forum in the US that the company plans to enter mass production of chips using its 10nm FinFET process by the end of 2016,.
But in a statement TSMC claimed it could the outfit said the way things are shaping up it could beat that time table. TSMC continued that in the 10nm FinFET race, Intel will be its major competitor.
We expect to hear a bit more about TSMC’s plans at its Taiwan Technology Symposium 2015 on May 28. At the upcoming event, the foundry is expected to talk about the progress and development of its FinFET manufacturing nodes.
TSMC chairman Morris Chang remarked earlier in 2015 that TSMC expects to gain a majority of market share in the FinFET segment in 2016.
Intel is also expected to release its first chips made using 10nm process technology as early as in the middle of 2016.