Microsoft working on 'Singularity' AI Cloud infrastructure service


New Delhi, Feb 22 (IANS): Microsoft is working on a new Artificial Intelligence (AI) infrastructure service codenamed 'Singularity' that will provide data scientists and AI researchers a way to build, scale, experiment and iterate on their models on a Microsoft Azure Cloud service built specifically for AI.

A group of researchers, mostly of Indian origin, have published a paper which provides technical details about the ‘Singularity' project.

"Lowering costs by driving high utilisation across deep learning workloads is a crucial lever for cloud providers. We present Singularity, Microsoft's globally distributed scheduling service for highly-efficient and reliable execution of deep learning training and inference workloads," the researchers wrote.

‘Singularity' is a fully managed, globally distributed infrastructure service for AI workloads at Microsoft, with support for diverse hardware accelerators.

‘Singularity' is designed from the ground up to scale across a global fleet of hundreds of thousands of GPUs and other AI accelerators.

"Singularity is built with one key goal: driving down the cost of AI by maximising the aggregate useful throughput on a given

fixed pool of capacity of accelerators at planet scale," said the researchers.

In 2019, Microsoft invested $1 billion in OpenAI, and last year, it announced that ‘Azure OpenAI Service' will give customers access to OpenAI's powerful models in addition to the security, reliability, compliance, data privacy and other enterprise-grade capabilities built into Microsoft Azure.

According to the researchers, ‘Singularity' achieves a significant breakthrough in scheduling deep learning workloads, converting niche features such as elasticity into mainstream, always-on features that the scheduler can rely on for implementing stringent service level agreements (SLAs).

"Singularity enables unprecedented levels of workload fungibility, making it possible for jobs to take advantage of spare capacity anywhere in the globally-distributed fleet, while still preserving the SLAs," they added.

Microsoft in 2020 unveiled a new powerful supercomputer in collaboration with OpenAI, making new infrastructure available in Azure to train extremely large AI models.

The supercomputer developed for OpenAI is a single system with more than 285,000 CPU cores, 10,000 GPUs and 400 gigabits per second of network connectivity for each GPU server.

 

  

Top Stories


Leave a Comment

Title: Microsoft working on 'Singularity' AI Cloud infrastructure service



You have 2000 characters left.

Disclaimer:

Please write your correct name and email address. Kindly do not post any personal, abusive, defamatory, infringing, obscene, indecent, discriminatory or unlawful or similar comments. Daijiworld.com will not be responsible for any defamatory message posted under this article.

Please note that sending false messages to insult, defame, intimidate, mislead or deceive people or to intentionally cause public disorder is punishable under law. It is obligatory on Daijiworld to provide the IP address and other details of senders of such comments, to the authority concerned upon request.

Hence, sending offensive comments using daijiworld will be purely at your own risk, and in no way will Daijiworld.com be held responsible.