Generative AI and large-scale language models (LLM) show no signs of slowing down. Organizations are eager to take action and leverage this innovative technology to gain an edge over their competitors. However, developing AI and ML models requires significant amounts of computing power, preferably with dedicated infrastructure for AI development and applications.
And that’s exactly what Google Cloud brings to market with the general release of A3 instances, available for purchase this month. Although A3 was announced earlier this year, Google Cloud chose the recent Next conference as the venue to announce news about when the technology will be available to the general public.
What is an A3 GPU supercomputer?
Google Cloud’s A3 supercomputer is NVIDIA H100 Tensor Core GPUenables unprecedented performance and makes it faster and easier to develop and deploy AI apps on the Google Cloud platform. The combination of the NVIDIA H100 Tensor Core GPU and Google’s networking advances makes the A3 VM a powerful tool. Let’s take a look at the details.
- 10x more network bandwidth than A2 VMs on Google Cloud
- 26 exaflops performance significantly reduces model training time and associated costs.
- 4800 MHz DDR5 DIMM provides 2TB of host memory
When it comes to developing ML models, speed is everything. A3 VM accelerates training so organizations can build more complex ML models faster and lay the foundation for LLM development.
winning partnership
The other story here is that consumers benefit when the tech giants partner up. The announcement of the general release of A3 instances comes after NVIDIA was named Google Cloud’s Generative AI Partner of the Year. This represents a critical technology partnership at a time when the best candidates for generative AI infrastructure are not yet fully determined.
There are other elements to this partnership, including the general availability of H100 GPUs on Google Cloud’s VertexAI platform in the coming weeks. However, the ultimate goal is to allow customers to easily build and deploy their AI applications on Google Cloud.
“Accelerated computing and generative AI will work together to create unprecedented We are at a tipping point where we are accelerating innovation at an unprecedented pace.” “Our expanded collaboration with Google Cloud will help developers accelerate the development of infrastructure, software, and services that improve energy efficiency and reduce costs.”
“Google Cloud has a long history of innovation in AI to help our customers innovate and accelerate,” said Kurian. “Many of Google’s products are built and delivered on NVIDIA GPUs, and many of our customers seek NVIDIA-accelerated computing to power efficient development of LLMs that advance generative AI. .”
As the demand for AI and ML modeling grows at lightning speed, the business world demands purpose-built products. Google Cloud’s A3 VMs ensure faster development and deployment times.