While still technically in pre-release, the Mac-optimised TensorFlow fork supports native hardware acceleration on Mac devices with M1 or Intel chips through Apple’s ML Compute framework.

The new TensorFlow release boasts of an over 10x speed improvement for common training tasks. While impressive, it has to be taken in the context that the GPU was not previously used for training tasks. 

A look at the benchmarks still indicates a substantial gap between the Intel and M1-based Macs across various machine learning models:

In a blog post, Pankaj Kanwar, Tensor Processing Units Technical Program Manager at Google, and Fred Alcober, TensorFlow Product Marketing Lead at Google, wrote:

“These improvements, combined with the ability of Apple developers being able to execute TensorFlow on iOS through TensorFlow Lite, continue to showcase TensorFlow’s breadth and depth in supporting high-performance ML execution on Apple hardware.”

We can only hope that running these workloads doesn’t turn MacBooks into expensive frying pans—but the remarkable efficiency they’ve displayed so far gives little cause for concern.

The post TensorFlow is now available for those shiny new ARM-based Macs appeared first on AI News.

The cloud-based solution aims to help accelerate the delivery of software in a rapidly-changing world while maintaining a high bar of quality.

Gareth Smith, CTO of Eggplant, said:

“The launch of our cloud platform is a significant milestone in our mission to rid the world of bad software. In our new normal, delivering speed and agility at scale has never been more critical.

Every business can easily tap into Eggplants’ AI-powered automation platform to accelerate the pace of delivery while ensuring a high-quality digital experience.” 

Enterprises have accelerated their shift to the cloud due to the pandemic and resulting increases in things such as home working.

Recent research from Centrify found that 51 percent of businesses which embraced a cloud-first model were able to handle the challenges presented by COVID-19 far more effectively.

Eggplant’s Digital Automation Intelligence (DAI) Platform features:

• Cloud-based end-to-end automation: The scalable fusion engine provides frictionless and efficient continuous and parallel end-to-end testing in the cloud, for any apps and websites, and on any target platforms. 
• Monitoring insights: The addition of advanced user experience (UX) data points and metrics, enables customers to benchmark their applications UX performance. These insights, added to the UX behaviour helps improve SEO. 
• Fully automated self-healing test assets: The use of AI identifies the tests needed and builds and runs them automatically, under full user control. These tests are self-healing, and automatically adapt as the system-under-test evolves.   

The solution helps to support the “citizen developer” movement—using AI to enable no-code/low-code development for people with minimal programming knowledge.

Both cloud and AI ranked highly in a recent study (PDF) by Deloitte of the most relevant technologies “to operate in the new normal”. Cloud and cybersecurity were joint first with 80 percent of respondents, followed by cognitive and AI tools (73%) and the IoT (65%).

Eggplant’s combination of AI and cloud technologies should help businesses to deal with COVID-19’s unique challenges and beyond.

(Photo by CHUTTERSNAP on Unsplash)

The post Eggplant launches AI-powered software testing in the cloud appeared first on AI News.

The new video courses assume the developer already has a basic standard of Python skills. If you don’t, I’m afraid you’ll need to brush up on those first. Fortunately, Microsoft released a 44-part “Python for Beginners” series last autumn (or “fall” for our American friends.)

For those with the Python skills, or if you’ve just consumed all 44-parts of Microsoft’s course in record time, the new courses are around three hours each. 

The first course, More Python for Beginners, features 20 videos and covers areas such as lambdas, inheritance, and asynchronous operations.

The second course, Even More Python for Beginners: Data Tools, consists of 31 videos and really dives into using Python for machine learning and data science. Students are taught how to use popular Python libraries for the aforementioned topics; along with using the Jupyter Notebooks browser-based development environment.

Each of the courses are still led by Christopher Harrison, senior program manager at Microsoft, and Susan Ibach, business development manager from Microsoft AI Gaming. 

“While we’re not going to get into conversations about choosing algorithms or building models, we are going to introduce what you’ll use when you begin the journey. We’ll highlight Jupyter Notebooks, the favorite tool of data scientists,” Harrison and Ibach wrote in a blog. 

As of writing, the first part in Microsoft’s Python course series has been viewed over 1.75 million times.

The course’s popularity is of little surprise given the huge interest in Python as AI talent becomes more in-demand; with six-figure salaries not unheard of. In last year’s GitHub Octoverse report, Python overtook Java to become the second most popular language on the world’s largest repository host.

The post Microsoft releases two Python video courses which help aspiring AI developers appeared first on AI News.

Edison is an AI platform launched last year to help with leveraging data from imaging devices.

Kieran Murphy, President and CEO at GE Healthcare, says:

“We introduced Edison just one year ago at RSNA to help health providers take advantage of data in new and significant ways.

With the introduction of the Edison Developer Program, and a suite of new intelligent applications and smart devices powered by Edison, we are building on that promise as we continue to work with partners to realize our collective goal of advancing the future of health.”

While there’s significant interest in using AI for healthcare, the time it takes for implementing new innovations is “cumbersome and complex,” according to GE Healthcare.

GE Healthcare aims to serve clinicians’ need for a single solution to assist with integrating AI algorithms into existing workflows to help them, and therefore their patients, benefit from the potential of these new technologies much faster.

Market-ready AI applications will be deeply integrated into GE Healthcare’s vast existing solutions – on medical devices, in the cloud, or at the edge of the network.

Developers jumping aboard the new initiative will have the reach of GE Healthcare’s large userbase. GE Healthcare’s business reported $19 billion in revenue last year and spans 160 countries. On average, three patients are imaged with the company’s solutions every second.

“The Edison Developer Program exposes a number of potential capabilities of the Edison platform, including secure device connectivity, data aggregation for clinical context, advanced visualization, workflow and AI orchestration, in addition to a rich set of AI capabilities for data traceability, curation, annotation, model training and inferencing,” the company wrote in a press release.

“This set of services will reduce the complexity of developing and integrating AI and data-based healthcare applications in clinical workflows. Program members are selected and vetted based on rigorous clinical and technical evaluations as well as regulatory clearance to ensure confidence and security of solutions offered through the Edison platform.”

GE Healthcare is currently working with a range of AI and analytics companies including Arterys, iCAD, Koios Medical, MaxQ AI and Volpara.

Koios Medical, for example, built and deployed Breast Assistant – an AI-based risk assessment for breast cancer that aligns to a BI-RADS (Breast Imaging-Reporting and Data System) category.

Interested in hearing industry leaders discuss subjects like this? , , , AI &

The post GE Healthcare’s new initiative aims to boost AI adoption by health providers appeared first on AI News.

Cloud AI Platform is Google’s machine learning platform-as-a-service (ML PaaS) designed for AI developers, engineers, and data scientists. The platform is end-to-end and supports the full development cycle from preparing data, to training, all the way to building and deploying machine learning models.

Among the most noteworthy additions to the platform is support for Nvidia GPUs. As Google explains, “ML models are so complex that they only run with acceptable latency on machines with many CPUs, or with accelerators like NVIDIA GPUs. This is especially true of models processing unstructured data like images, video, or text.”

Previously, Cloud AI Platform only supported one vCPU and 2GB of RAM. You can now add GPUs, like the inference-optimised, low latency NVIDIA T4, for AI Platform Prediction. The basic tier adds support for up to four vCPUs.

AI Platform Prediction is being used by Conservation International, a Washington.-based organisation with the mission “to responsibly and sustainably care for nature, our global biodiversity, for the wellbeing of humanity,” for a collaborative project called Wildlife Insights.

“Wildlife Insights will turn millions of wildlife images into critical data points that help us better understand, protect and save wildlife populations around the world,” explains Eric H. Fegraus, Senior Director, Conservation Technology.

“Google Cloud’s AI Platform helps us reliably serve machine learning models and easily integrate their predictions with our application. Fast predictions, in a responsive and scalable GPU hardware environment, are critical for our user experience.”

Support for running custom containers in which to train models has also become generally available. Users can supply their own Docker images with an ML framework preinstalled to run on AI Platform. Developers can test container images locally before they’re deployed to the cloud.

Customers aiming to use the platform for inference – hosting a trained model that responds with predictions – can now do so. Machine learning models can be hosted using the Google Cloud AI Platform and AI Platform Prediction can be used to infer target values for obtaining new data.

Oh, and AI Platform Prediction is now built on Kubernetes which enabled Google to “build a reliable and fast serving system with all the flexibility that machine learning demands.”

Interested in hearing industry leaders discuss subjects like this? , , , AI &

The post Google Cloud AI Platform updates make it ‘faster and more flexible’ appeared first on AI News.

By providing a common framework, Acumos intends to make artificial intelligence development more accessible. Acumos differs from TensorFlow, an existing popular framework, in that it will be offer a directory for sharing AI models in addition to working on them.

“An open and connected AI platform will promote collaboration as developers and companies look to define the future of AI,” said Jim Zemlin, Executive Director at The Linux Foundation.

“Because the platform is open source, it will be accessible to anyone with an interest in AI and machine learning, and customisable to meet specific needs. We expect interest from organisations doing work with autonomous vehicles, drones, content curation and analytics, and much more.”

The founding members of Acumos are AT&T and Tech Mahindra; both of whom are contributing the initial code. The Linux Foundation will host the platform and its AI marketplace in the hope its large developer community will help to nurture an active, large ecosystem around the project.

“In The Linux Foundation, we have an able partner to increase the adoption of Acumos across the community to help enrich the platform,” comments Raman Abrol, SVP & Business Unit Head of Americas, Comms, Media & Entertainment at Tech Mahindra. “This platform jointly developed with AT&T is a testimony to our continued commitment to the open source community.”

The ultimate goal for the project is to create an industry standard which makes AI development easier and builds a community around reusable code. Whereas most existing tools are complex and designed for data scientists, Acumos will be user-centric.

“The current state of today’s AI environment is fractured, which creates a significant barrier to adoption,” said Mazin Gilbert, Vice President of Advanced Technology at AT&T Labs. “Acumos will expedite innovation and deployment of AI applications, and make them available to everyone.”

Acumos will launch early 2018 and is expected to gain further members in the coming weeks prior to release.

To find out more and how to get involved, head this way.

What are your thoughts on Acumos and the current state of AI development?

The post Acumos: Linux Foundation makes it easier to deploy AI apps appeared first on AI News.