It's easy to forget that Silicon Valley starts with'silicon', and that there would be no technology innovation without innovation at the silicon level. And Graphcore is well aware of that as the Bristol-based company is designing its own dedicated AI chipset. That's why I'm glad to announce that Graphcore co-founder and CEO Nigel Toon is joining us at TechCrunch Disrupt Berlin. Graphcore has managed to attract a ton of attention from day one. Originally founded in 2016, the startup has raised more than $300 million from top investors, such as Sequoia Capital, BMW, Microsoft, Samsung and a ton of others. The company last raised a $200 million Series D round led by Atomico and Sofina.

The report titled, The Shift to Predictive Engagement, recommends using predictive, real-time and reactive methods in conjunction. That's because without historical data to compare against it is impossible to retrain algorithms. Conducted by the Which-50 Digital Intelligence Unit in partnership with Genesys the report cites the use of analytics by full-service media buying agency, Audience Group, which implemented predictive engagement techniques last year. In the Investigates report, director James McDonald told authors that only a year ago the agency was manually allocating its spend between advertising, SEO, premium video, data use, remarketing banners, and mobile. Each channel would have its own budget approval, and then the team would execute slavishly for those channels.

The rise of deep-learning (DL) has been fueled by the improvements in accelerators. Accelerators allow DL models to crunch a large amount of data, which is vital for them to achieve high accuracy. In fact, AlexNet, the famous winner of the ILSVRC 2012 competition, was trained on GPUs. GPU continues to remain the most widely used accelerator for DL applications, due to several of its features, such as high performance, continued improvements in its architecture and software-stack, ease of programming using high-level languages such as CUDA and availability of GPUs in cloud. "Accelerating DL models" is chasing a moving target As DL models are becoming more pervasive and accurate, their compute and memory requirements are growing tremendously.

Intermap Technologies ("Intermap"), a global leader in geospatial intelligence solutions, today announced the launch of its Lido Surface Data NEXTView ("NEXTView") data solution, co-developed with Lufthansa Systems, for the Unmanned Aircraft Systems (UAS) market. NEXTView is a high-accuracy, global 3D elevation dataset customized for aviation applications. It is continually refreshed to ensure currency and compliance with regulatory update requirements. The UAS market is composed of Unmanned Aerial Vehicles (UAVs, or drones) and the control systems that fly them. It is a critical time for UAS technology as it expands in many commercial, government and military applications around the world.

Barrister and human rights advocate Fiona McLeod, who delivered the Solomon Lecture titled Accountability in the age of the artificial, said accountability was under threat in Australia and internationally. Ms McLeod called for a robust national integrity commission. "We have settled for a'trust us or vote us out' model of democracy and a veneer of transparency resulting in a piecemeal and un-strategic approach to accountability," she said. "For example, the federal government committed to bring in a Commonwealth integrity commission after its hand was forced by independents in the last Parliament. "The original preferred model of a closed hearings, with no power of the commission to initiate inquiries, appeared to me more like a benign hall monitor issuing'don't run' notices than a body capable of balancing competing public and private interests, of driving a culture of anti-corruption throughout government, private organisations and the community."

The 2008-09 financial collapse led to a Federal Reserve directive that banks with consolidated assets over $50 billion have additional risk assessment frameworks and budgetary oversight in place. To assess a bank's financial foundation, the Federal Reserve oversees a number of scenarios (company-run stress tests). Referred to as the Comprehensive Capital Analysis and Review (CCAR) process, these tests are meant to measure the sources and use of capital under baseline as well as stressed economic and financial conditions to ensure capital adequacy in all market environments. As Ayasdi reports, Citi consistently struggled to pass its annual stress test, failing two of the first three stress tests. The bank was in need of a way to rapidly create accurate, defensible models that would prove to the Federal Reserve that they could adequately forecast revenues and the capital reserve required to absorb losses under stressed economic conditions.

In 1951, Marvin Minsky asked an imaginary mouse to navigate an imaginary maze. Together with Dean Edmonds, Marvin carefully connected three hundred vacuum tubes together with an assembly of motors and light bulbs, applying ideas about the wiring of neurons in human and animal brains. Minsky and Edmonds watched the virtual mouse's progress on a bank of lights and offered rewards when it moved toward its goal. Through repeated tries, the mouse learned to escape the maze. When researchers coined the term "artificial intelligence" (AI) five years later, they hoped to prove in one summer that every feature of learning and intelligence could be conducted by a machine.

Variation and Disease: Translating genetic findings into therapeutics remains an unsolved challenge, partly because in 93% of cases, disease-associated common variants do not disrupt proteins directly, but instead alter their genomic control elements. Our group develops and uses epigenomic maps of regulatory elements, and cellular circuits linking them to their regulators and target genes, in order to understand how human genetic variation contributes to disease and cancer. We have developed resources and methods for studying how genetic variation impacts gene expression, regualtory region activity, cellular phenotypes, and ultimately human disease. We have applied these methods to obesity, Alzheimer's disease, cardiovascular traits, psychiatric disorders, and cancer, resulting in multiple insights. In addition to dissecting these circuits, we have used gene manilations and genome editing to reverse the phenotypic signatures of disease from risk and non-risk individuals, paving the way for genomics-based therapeutics.

There is one aspect of Tesla where it is miles ahead of the competition. And that is in its use of data to build what might just be the world's most sophisticated, cutting-edge neural network anywhere. Silicon Valley loves buzz words that metaphorically underscore the "next big thing." Data, for instance, was "the new oil." It was just waiting there, an unrefined asset, ready to be tapped, refined and harnessed to drive competitive advantage. But the hype and hoopla over big data has been eclipsed by the nitty gritty reality of the technical challenges in actually converting that structured, unstructured and semi-structured stuff into something truly valuable.

Get a look at the future of virtualization with NVIDIA's latest demos -- from accelerated AI to RTX graphics -- at VMworld, this week in San Francisco. Thousands of technical professionals, software architects, data scientists and more will be at the annual conference to gain insight into the most recent advancements in virtualization. NVIDIA vComputeServer brings GPU acceleration to server workloads like AI, deep learning and high performance computing. Businesses can run GPU workloads in virtualized environments for improved security, utilization and manageability. And with NVIDIA NGC GPU-optimized software, users can access models, training scripts and workflows that can further accelerate AI.