The Omnitek deep learning processing unit (DPU) employs a novel mathematical framework combining low-precision fixed point maths with floating point maths to achieve 135 GOPS/W at full 32-bit floating point accuracy when running the VGG-16 CNN in an Arria 10 GX 1150. Scalable across a wide range of Arria 10 GX and Stratix 10 GX devices, the DPU can be tuned for low cost or high performance in either embedded or data centre applications. The DPU is fully software programmable in C/C or Python using standard frameworks such as TensorFlow, enabling it to be configured for a wide range of standard CNN models including GoogLeNet, ResNet-50 and VGG-16 as well as custom models. No FPGA design expertise is required to do this. "We are very excited to apply this unique innovation, resulting from our joint research program with Oxford University, to reducing the cost of a whole slew of AI-enabled applications, particularly in video and imaging where we have a rich library of highly optimised IP to complement the DPU and create complete systems on a chip", commented Roger Fawcett, CEO at Omnitek.
Intel on Tuesday announced that it's acquiring Omnitek, makers of a programmable chip design for visual processing. Financial terms of the deal were not disclosed. FPGAs, or field-programmable gate array (FPGA) fabric, is an integrated circuit that can be further configured after manufacturing. Basingstoke, England-based Omnitek has over 220 FPGA IP cores and accompanying software. Intel plans to merge the technology, plus Omnitek's 40 employees, into its Programmable Solutions Group.
"Were you surprised we went into the cloud?" asks Qualcomm president Cristiano Amon during a chat by video conference. Amon was referring to how Qualcomm, a giant in mobile chips, is now hoping to make it big in machine learning "inference" for data centers. Qualcomm is not a presence in the data center. It entered that market in 2014 with lots of gusto, only to back out last year. When Amon became president, in December of 2017, his team took a look at the enormous cost to compete with the server CPU king, Intel, and how little it had produced in actual shipments for Qualcomm.
While Qualcomm has integrated several gaming-specific technologies into its Snapdragon mobile processors, on Tuesday the company announced something a little different: a version of its Snapdragon 730 optimized for gaming, dubbed the Snapdragon 730G. Though mobile gaming may be an idle pastime with American consumers, it's a way of life overseas. Over 586 million mobile gamers are in China alone--twice the population of the United States, according to Qualcomm's Hiren Bhinde at Qualcomm's technology summit last December. It isn't clear which phones and mobile devices Qualcomm has in mind for the Snapdragon 730G, but recent gaming phones from Asus ROG and Razer indicate that Qualcomm was designing for what they hope will be a trend. Though most premium smartphones use Qualcomm's 8-series CPUs like the Snapdragon 855, the new 7-series chips are designed for a slightly cheaper but still premium phone.
With legitimate economic uncertainties for some of us, paired with predictions of a dystopian end of days at their hands, tech like robots, AI and automation have earned a bad rap. Sony has taken a soft-touch approach when it comes to robotics, perhaps best illustrated by its iconic electronic pet, Aibo. It's just one effort from a company that has seen a resurgence over the last few years, not just through a redesigned robot dog, but also its work on self-driving car sensors. Alongside CES and the other usual tech shows, Sony has been attending Milan Design Week for the past few years, and this time its lofty concepts touch on artificial intelligence, robotic empathy and how we're all going to feel about it. I wouldn't say I walked away with any answers, however.
Flagship features continue to trickle down from $1,000 phones to their more-affordable brothers, and the same is happening with the chips that power them. Qualcomm unveiled new midrange mobile CPUs today that offer advanced features typically reserved for high-end phones, like AI processing and gaming enhancements. The Snapdragon 730, 730G and 665 are supposed to show up in (presumably cheaper-than-flagship) devices in mid-2019, meaning we may have a slate of budget-friendly handsets to look out for. For the first time, Qualcomm is launching a gaming-specific version of a chipset alongside the regular one. The Snapdragon 730G (G stands for Gaming, get it?)
A soft robot, attached to a balloon and submerged in a transparent column of water, dives and surfaces, then dives and surfaces again, like a fish chasing flies. Soft robots have performed this kind of trick before. But unlike most soft robots, this one is made and operated with no hard or electronic parts. Inside, a soft, rubber computer tells the balloon when to ascend or descend. For the first time, this robot relies exclusively on soft digital logic.
The history of AI is often told as the story of machines getting smarter over time. What's lost is the human element in the narrative, how intelligent machines are designed, trained, and powered by human minds and bodies. In this six-part series, we explore that human history of AI--how innovators, thinkers, workers, and sometimes hucksters have created algorithms that can replicate human thought and behavior (or at least appear to). While it can be exciting to be swept up by the idea of super-intelligent computers that have no need for human input, the true history of smart machines shows that our AI is only as good as we are. On 14 February 1946, journalists gathered at the Moore School of Engineering at the University of Pennsylvania to witness a public demonstration of one of the world's first general-purpose electronic digital computers: the Electronic Numerical Integrator and Computer (ENIAC).
Deep neural networks (DNNs) have been shown to outperform conventional machine learning algorithms across a wide range of applications, e.g., image recognition, object detection, robotics, and natural language processing. However, the high computational complexity of DNNs often necessitates extremely fast and efficient hardware. The problem gets worse as the size of neural networks grows exponentially. As a result, customized hardware accelerators have been developed to accelerate DNN processing without sacrificing model accuracy. However, previous accelerator design studies have not fully considered the characteristics of the target applications, which may lead to sub-optimal architecture designs. On the other hand, new DNN models have been developed for better accuracy, but their compatibility with the underlying hardware accelerator is often overlooked. In this article, we propose an application-driven framework for architectural design space exploration of DNN accelerators. This framework is based on a hardware analytical model of individual DNN operations. It models the accelerator design task as a multi-dimensional optimization problem. We demonstrate that it can be efficaciously used in application-driven accelerator architecture design. Given a target DNN, the framework can generate efficient accelerator design solutions with optimized performance and area. Furthermore, we explore the opportunity to use the framework for accelerator configuration optimization under simultaneous diverse DNN applications. The framework is also capable of improving neural network models to best fit the underlying hardware resources.
While we take a look back at some highlight stories this week out of SXSW, there's a trial result in Apple v. Qualcomm, and a presidential candidate's past is quite relevant to our interests. Oh, and we rode in the Model Y. Almost forgot about that. That's one way to do it.Fan While you wouldn't mistake it for an official remaster, using AI Gigapixel's neural networks on old Deep Space 9 video provides a considerably cleaner, sharper look than the 480p original without introducing visual artifacts. This is just one of several cases going on between the two companies.Jury decides Apple violated three Qualcomm patents in iPhones Following a two-week trial, a jury has determined that Apple violated three Qualcomm patents in some iPhones.