Machine Learning: Google Is Right Away Offering Gpu's Equally A Service (Goog; Nvda)

From Google's Cloud Platform blog:

GPUs every bit a service amongst Kubernetes Engine are right away by as well as large available 

[Editor's note: This is 1 of many posts on venture features enabled past times Kubernetes Engine 1.10. For the amount coverage, follow along here.]
Today, we’re excited to denote the full general availability of GPUs inwards Google Kubernetes Engine, which accept expire 1 of the platform’s fastest growing features since they entered beta before this year, amongst core-hours soaring past times 10X since the terminate of 2017. Together amongst the GA of Kubernetes Engine 1.10, GPUs brand Kubernetes Engine a slap-up check for venture motorcar learning (ML) workloads. By using GPUs inwards Kubernetes Engine for your CUDA workloads, you lot produce goodness from the massive processing might of GPUs whenever you lot need, without having to contend hardware or fifty-fifty VMs.

We late introduced the latest as well as the fastest NVIDIA Tesla V100 to the portfolio, as well as the P100 is by as well as large available. Last but non least, nosotros too offering the entry-level K80, which is largely responsible for the popularity of GPUs. All our GPU models are available every bit Preemptible GPUs, every bit a means to bring down costs spell benefiting from GPUs inwards Google Cloud. Check out the latest prices for GPUs here. As the increase inwards GPU core-hours indicates, our users are excited almost GPUs inwards Kubernetes Engine. Ocado, the world’s largest online-only grocery retailer, is ever looking to apply state-of-the-art motorcar learning models for Ocado.com customers as well as Ocado Smart Platform retail partners, as well as runs the models on preemptible, GPU-accelerated instances on Kubernetes Engine.

“GPU-attached nodes combined amongst Kubernetes render a powerful, cost-effective as well as flexible environs for enterprise-grade motorcar learning. Ocado chose Kubernetes for its scalability, portability, potent ecosystem as well as huge community support. It’s lighter, to a greater extent than flexible as well as easier to keep compared to a cluster of traditional VMs. It too has slap-up ease-of-use as well as the might to attach hardware accelerators such every bit GPUs as well as TPUs, providing a huge boost over traditional CPUs.” — Martin Nikolov, Research Software Engineer, Ocado

 ...MUCH MORE

See too the "Getting Started amongst Google Kubernetes Engine" at Coursera if you lot desire to showtime your DIY ML co.

As Wikipedia says almost Kubernetes:

The master codename for Kubernetes inside Google was Project Seven, a reference to Star Trek grapheme Seven of Nine that is a 'friendlier' Borg.^[9] The vii spokes on the bicycle of the Kubernetes logo is a nod to that codename.