Google is pushing for some of its business cloud customers to offload their “most demanding workloads” onto pricey Intel virtual machines.
In a post on the Google Cloud blog, the company has announced the launch of the Intel Software Center of Excellence after a successful pilot scheme in 2021.
Businesses with use-cases such as video encoding, web transactions, databases, analytics, and AI inference are invited to apply for the free scheme, which will involve Google and Intel collaborating with accepted applicants to access and optimize their workflow performance.
Personalized cloud collaboration
The opportunity for Intel to personally review a business’ internal resources in order to deliver a personalized report on how it can implement specific optimisations to a workflow running on Google Cloud could seem enticing.
However, while the scheme is free, Intel VMs will cost businesses more on account of the company’s Xeon Scalable (opens in new tab) processors, which offer built-in AI acceleration via the AVX-512 set of vector instructions – first introduced in 2013 – and the more recent Deep Learning Boost (opens in new tab) feature which leverages the technology.
While it’s true that VMs fitted with Xeon Scalable processors can provide impressive speed boosts to some workflows, the software that powers them must be optimized by developers for use with the tech.
High costs and limited use cases could be the reasons why Google and Intel are only offering a hands-on, “white glove service”, and one that’s only available to certain “high-growth enterprises”, subject to application.
Companies who don’t need, can’t leverage, or simply aren’t aware of the latest Intel processor technology could be put off by an automatic change to service.
While Intel has long been the only option for businesses looking to take advantage of AI acceleration, AMD’s latest raft of Epyc server processors will ship with support for AVX-512 this autumn.