Narendra Sen, CEO of NeevCloud, is an ambitious man. He envisions building an AI cloud infrastructure built for Indian clients with 40,000 graphics processing units (GPUs) by 2026. The infrastructure aims to support Indian enterprises with training, estimation, and other AI workloads.

Utah, another prominent AI cloud provider, recently drew attention for its recognition as the inaugural NVIDIA Partner Network (NPN) cloud partner in India, earning Elite Partner status globally. .

Yotta’s ambitious plan is to establish an AI cloud infrastructure with 32,768 GPUs by the end of 2025. NeevCloud wants to do better.

Additionally, NeevCloud will soon launch an AI inference platform that will provide open source models such as the Llama 3 series, Mistral models and DBRX by Databricks.

“Later this month, we plan to introduce DBRX as we see a huge demand for the model in this country. After that, we will launch text-to-image models from Stability AI.

“We are focusing on a phased approach to ensure smooth operation. User experience is paramount, and if everything goes as planned, we can expand the range. Meanwhile, We are also expanding our capabilities along with these developments,” Sen said the aim.

Assessment of AI

Inference involves applying a pre-trained model to new input data to make predictions or decisions. So far, nearly 3,500 developers have signed up to use NeevCloud’s inference platform.

The company plans to launch a beta version of the platform this month and provide free tokens to developers to come and test NeevCloud’s inference platform.

For the inference platform, the company plans to leverage nearly 100 NVIDIA GPUs, including the NVIDIA H100s, A100, and L40 GPUs. Furthermore, NeevCloud plans to introduce AMD’s M1300X into the mix.

AMD’s M1300X offers cost advantages over the rest. Indians don’t care about brand. All they care about is that the API should work, the latency should be fast, and they should get instant tokens — that’s it,” Sen said.

Neev may also be the first company to bring Cloud. GrokLanguage Processing Units (LPU) of India. Recently, Sean posted a photo of himself on LinkedIn, posing with Jonathan Ross at Grok’s headquarters in San Francisco.

AD 4nXcccXKj6A1R2Y81J6IlZQ fW7IiwFWSyBK9kZnvKnPosSMW6RPnoEMC1N7rKMtx MUeFmciTKVfGKP5YWE eeMRR2ss npQX jY xjqwAqzylPCF0IHEHtTFx0aao9 WZTkTaMnOuqwUd4nwju4nS8lOQW

(Source: LinkedIn)

Sen, however, refrained from revealing more in this regard as things are not yet finalised. “While we all know the big names like NVIDIA, there are other players in the market that we are trying to onboard like Simba Nova,” he said.

AI Infrastructure as a Service

For AI Cloud, Sen revealed that he has placed an order with HP Enterprise (HPE) for 8,000 NVIDIA GPUs, which he expects to receive in the second half of this year.

NeevCloud will compete directly with Yotta in the AI-as-a-infrastructure space. Specifically, another company that already has Making GPUs more accessible in India is an E2E network.

The NSE-listed company offers NVIDIA’s H100 GPU and NVIDIA A100 Tensor Core GPUs at competitive prices compared to major hyperscalers.

Recently, Ola Cabs founder Bhavesh Agarwal announced his decision to make the offer. Krutrim AI Cloud To Indian developers.

Similarly, Tata Communications also partnered with NVIDIA to build A large-scale AI cloud infrastructure For our customers in the private and public sector.

While Krutrim and Tata Communications have not revealed the number of GPUs they plan to deploy, NeevCloud plans to deploy another 12,000-15,000 GPUs by 2025. “Then, by 2026 we will deploy the rest to achieve the target of 40,000 GPUs,” Sen said.

How will NeevCloud fund the 40,000 GPU acquisition?

However, deploying a GPU cluster of around 40,000 GPUs would cost billions of dollars. According to Sen, it will cost about $1.5 billion.

While Yotta is backed by the Hiranandani Group, NeevCloud is banking on its data center partners to help not only procure GPUs but also deploy them.

So far, NeevCloud has partnered with three major data center companies in India, two in Chennai and one in Mumbai. One of them is one of India’s largest data center operators, Sen said.

“What we have is a revenue-sharing model. Even though they already have the data center infrastructure, they need to deploy the GPUs on our behalf, and NeevCloud will bring customers, who have their own AI cloud. will have access to both the capacity (to be deployed) and our own data center servers,” Sen said.

Sen founded NeevCloud in 2023. However, Sen has been running a data center company called Rackbank Datacenters in Indore for many years.

Located in Crystal IT Park, Indore, the data center is 35,000 square feet and has a capacity of 32,000+ servers.

The company has innovated a liquid immersion cooling technology called Verona, to efficiently cool high-density computing hardware used for AI, machine learning, and high-performance computing (HPC) tasks.

The method involves immersing servers and other IT equipment in a dielectric, nonconductive coolant liquid, which facilitates direct heat transfer from the components to the liquid.

Source link