Story image

High demand for Google’s machine learning processor in NZ

22 Mar 18

Strong local demand for Google’s new generation of Cloud TPUs (Tensor Processing Units) means Kiwi businesses and consumers will benefit from better services and greater productivity.  

Google’s Tensor Processing Units, released into public beta last month, are hardware accelerators optimised to speed up and scale up Machine Learning (ML) workloads.

Google Cloud New Zealand’s Brendan Bain says interest from New Zealand businesses shows there is real appetite locally to start leveraging the technology.

“The benefit to business is much faster training of Machine Learning models, in some cases reducing from several weeks to several hours. That’s like moving from basic dial-up to broadband overnight, and the productivity boost will be significant.

“Kiwi consumers can expect smarter, faster services across the board, from online shopping to old-fashioned big box retailing,” Bain says.  

A large number of Kiwi businesses are already using Machine Learning on a daily basis. 

Trade Me, which uses it to improve its search function and make recommendations to users, is one business planning to use Google’s TPUs. 

Lead data scientist Lester Litchfield says they offer a lot of opportunities to improve the site.

“TPU's will help us iterate and explore ideas quicker, and present fresher models to our users. We'll be able to deliver better experiences faster than we can with traditional CPU and GPU powered Machine Learning,” Litchfield says.

Why TPUs for machine learning?

Traditionally, writing programs for custom ASICs and supercomputers has required deeply specialised expertise.

By contrast, Cloud TPUs can be programmed with high-level TensorFlow APIs, and Google has open-sourced a set of reference high-performance Cloud TPU model implementations to help users get started:

  • ResNet-50 and other popular models for image classification
  • Transformer for machine translation and language modeling
  • RetinaNet for object detection

Google continuously test these model implementations both for performance and for convergence to the expected accuracy on standard datasets.

Over time, Google is planning to open-source additional model implementations.

Adventurous ML experts may be able to optimise other TensorFlow models for Cloud TPUs on their own using the documentation and tools provided.

52mil users affected by Google+’s second data breach
Google+ APIs will be shut down within the next 90 days, and the consumer platform will be disabled in April 2019 instead of August 2019 as originally planned.
GirlBoss wins 2018 YES Emerging Alumni of the Year Award
The people have spoken – GirlBoss CEO and founder Alexia Hilbertidou has been crowned this year’s Young Enterprise Scheme (YES) Emerging Alumni of the Year.
SingleSource scores R&D grant to explore digital identity over blockchain
Callaghan Innovation has awarded a $318,000 R&D grant to Auckland-based firm SingleSource, a company that applies risk scoring to digital identity.
IDC: Standalone VR headset shipments grow 428.6% in 3Q18
The VR headset market returned to growth in 3Q18 after four consecutive quarters of decline and now makes up 97% of the combined market.
Spark Lab launches free cybersecurity tool for SMBs
Spark Lab has launched a new tool that it hopes will help New Zealand’s small businesses understand their cybersecurity risks.
Preparing for the future of work – growing big ideas from small spaces
We’ve all seen it: our offices are changing from the traditional four walls - to no walls. A need to reduce real estate costs is a key driver, as is enabling a more diverse and agile workforce.
Bluetooth-enabled traps could spell the end for NZ's pests
A Wellington conservation tech company has come up with a way of using Bluetooth to help capture pests like rats and stoats.
CERT NZ highlights rise of unauthorised access incidents
“In one case, the attacker gained access and tracked the business’s emails for at least six months. They gathered extensive knowledge of the business’s billing cycles."