Enterprises of the previous decade have transformed from transactional to digital. Today, digital enterprises use machine learning pipelines with humans-in-loop. However, the enterprises of tomorrow will be aiming for end-to-end AI-driven core business solutions, or intelligent enterprises. To address these demands, Google this week announced TensorFlow Enterprise at the ongoing TensorFlow World conference.
TensorFlow, one of the most popular machine learning frameworks, was open sourced by Google in 2015. Although it caters to every user, those deploying it on Google Cloud can benefit from enterprise-grade support and performance from the creators of TensorFlow. That’s why they launched TensorFlow Enterprise — for AI-enabled businesses on Google Cloud.
The difference between the two versions is, TensorFlow Enterprise services and products are unique to Google Cloud, but are built using TensorFlow. TensorFlow Enterprise adds enterprise and cloud-specific features. TensorFlow Enterprise currently includes Deep Learning VMs, Deep Learning Containers, or AI Platform Training and Prediction.
What Does It Offer
TensorFlow Enterprise is aimed at incorporating the following into the solutions that it provides to its users:
- Enterprise-grade support
- Cloud-scale performance
- Managed services
To meet these criteria, TensorFlow Enterprise introduces some improvements in the way TensorFlow Dataset reads data from Cloud Storage.
When the Google Cloud team ran the same code on a Compute Engine VM first with TensorFlow 1.14, then with TensorFlow Enterprise to compare the average number of examples per second.
And they found that there was a significant improvement with the use of TensorFlow Enterprise.
Through the power of Google Cloud’s TensorFlow Enterprise, we can quickly test, build and scale our Machine Learning models at massive scale, allowing us to serve up the most relevant ads and drive revenue for developers.
For certain versions of TensorFlow, the team promises to provide security patches and select bug fixes for up to 3 years. These versions will be supported on Google Cloud, and all patches and bug fixes will be available in the mainline TensorFlow code repository.
“Combined with our deep expertise in AI and machine learning, this makes TensorFlow Enterprise the best way to run TensorFlow,” the company wrote in their announcement.
The other key features can be summarised as follows:
- As part of their long term support service, the security patches and select bug fixes are taken care of up to 3 years.
- White-glove service for clients(read below).
- Enterprise-ready and performance-tuned TensorFlow through containers and virtual machines.
- Automatic provisioning and optimising of resources across CPUs, GPUs, and Cloud TPUs.
For AI businesses especially, TensorFlow Enterprise also offers white-glove service to help cutting-edge use cases and their AI challenges. It also includes engineer-to-engineer assistance from both Google Cloud and TensorFlow teams at Google. In order to qualify for this service, one must have spent $500,000 annually or be willing to commit to a spend of $500,000 annually on Deep Learning VMs, Deep Learning Containers, and/or AI Platform Training and Prediction.
According to NASSCOM, the cloud market in India is likely to soar to $7.1 billion by 2022 with the developmental leaps in Big Data analytics, Artificial Intelligence and Machine Learning and Internet of Things (IoT).
This is a crucial time to be developing new cloud technologies for solving problems. Cloud is now being used for navigating ships carrying cargo, connecting online and offline retail markets, machine learning in fintech for fraud detection and many more.
Back in April, CEO Thomas Kurian described Google Cloud’s new services to target different industries that include media and entertainment, healthcare, retail, financial services, public sector. Now with the introduction of TensorFlow Enterprise, GCP is poised to become bigger than ever.
