NVIDIA A100 Tensor Core GPU now made available on Google Cloud
Category: #technology  By Saipriya Iyer  Date: 2020-07-09
  • share
  • Twitter
  • Facebook
  • LinkedIn

NVIDIA A100 Tensor Core GPU now made available on Google Cloud

California-based technology major Nvidia Corporation has reportedly announced that its NVIDIA A100 Tensor Core GPU has been made available on Google Cloud.

A100 has been developed based on the lately introduced NVIDIA Ampere architecture and marks NVIDIA’s greatest generational leap ever. The product is highly efficient to improve training and inference computing performance by 20 times, further delivering enormous workload speedups to power the AI revolution.

Sources state that A100’s breakthrough performance on Google Compute Engine will accelerate the discoveries in rapidly growing critical industries. A100 accelerates unpredictable and complex workloads of all sizes in areas comprising scientific computing, AI training, real-time conversational AI, and inference applications.

Read More: NVIDIA brings AI solution to minimize InfiniBand data center downtime

According to Manish Sainani, Google Cloud’s Director of Product Management, the company’s customers are often looking for the latest hardware and software services that would help them impel innovation on AI and scientific computing workloads. He has further stated that the company feels proud to be the first major cloud provider to initiate the marketing of NVIDIA A100 GPUs, similar to the way it had been with NVIDIA T4 GPUs. The company is also thrilled to witness how its customers will use these new capabilities, added Sainani.

Reliable sources state that NVIDIA A100 has been made available in alpha on Google Chrome Engine merely a month following its introduction and has landed on the cloud sooner than any other NVIDIA GPU. With this move, Google has become the first cloud service provider to furnish the newly launched NVIDIA GPU.

The new A2 VM instances have been equipped to ensure different levels of performance for efficient acceleration of workloads across CUDA-enabled machine learning inference and training, high performance computing, and data analytics. A100 is especially known for effectively powering various compute-intensive applications comprising AI training and inference, scientific computing, genomics, data analytics, 5G services, edge video analytics, and more.

 

Source Credits:

https://blogs.nvidia.com/blog/2020/07/07/nvidia-ampere-gpus-google-cloud/

  • share
  • Twitter
  • Facebook
  • LinkedIn


About Author

Saipriya Iyer    

Saipriya Iyer

A content developer by choice, Saipriya Iyer holds a rich experience portfolio of more than five years in the content creation domain. Equipped with substantial expertise across the business, technology, and finance domains, Saipriya currently pens down insightful art...

Read More

More News By Saipriya Iyer

Ericsson, POST Luxembourg sign deal to deploy 5G RAN in Luxembourg
Ericsson, POST Luxembourg sign deal to deploy 5G RAN in Luxembourg
By Saipriya Iyer

POST Luxembourg and Ericsson have reportedly entered into a multiyear agreement for the deployment of 5G RAN and 5G Core technologies in Luxembourg. POST is reported to have gone live via Ericsson’s 5G on the 16th of October with technology dep...

Google Cloud obtains Secure Cloud Framework Agreement from Canada
Google Cloud obtains Secure Cloud Framework Agreement from Canada
By Saipriya Iyer

Google Cloud has reportedly acquired a Secure Cloud Framework Agreement from the federal government of Canada. The agreement will allow Google Cloud to go for direct sale of collaboration technologies and innovative cloud platform to federal agencies...

MFE Rentals teams up with Boston Dynamics to provide SPOT robot
MFE Rentals teams up with Boston Dynamics to provide SPOT robot
By Saipriya Iyer

MFE Rentals has reportedly joined hands with globally leading mobile robotics services provider, Boston Dynamics. Through the partnership, MFE Rentals’ clientele will gain access to the Boston Dynamics’ SPOT robot which has been designed ...