AMD and KT, a Korean telecom company, invest $22M in Series B funding for AI software developer Moreh

AMD and KT Invest in Moreh’s AI Software for Optimizing and Creating AI Models

Advanced Micro Devices (AMD) and Korean telco KT are part of the investors supporting Moreh, a startup that develops AI software tools for optimizing and creating AI models. In its latest funding round, Moreh raised a total of $22 million in Series B funding, bringing its total raised capital to $30 million. The company’s flagship AI software, MoAI, is designed to be compatible with existing machine learning frameworks such as PyTorch and TensorFlow, enabling applications and AI models to run on various platforms.

The Need for Advanced AI Software

According to a recent report, the existing AI software is suitable for smaller-scale AI models that use only a few GPUs. However, it falls short when it comes to more massive AI infrastructure. As AI continues to mature at an enterprise scale, companies are facing challenges with their IT infrastructure and data architectures, which are often deemed “unfit” for training AI models.

Moreh’s AI solutions address this challenge by offering users the ability to build more flexible AI infrastructure. This is particularly crucial given the global shortage of GPUs. The startup’s AI software allows GPUs and other AI chips, such as NPUs, to operate AI models without the need for any code changes. This includes large language models like GPT-3 and T5.

Successful Partnerships and Impressive Performance

KT has been collaborating with Moreh since 2021 to develop a cost-effective and scalable AI infrastructure powered by AMD GPUs and MoAI software. KT has found that Moreh’s offering outperforms Nvidia’s DGX in terms of both performance speed and GPU memory capacity. Moreh claims that its platform, in combination with AMD’s MI250 Instinct accelerator, showed 116% higher GPU throughput compared to Nvidia’s A100. Furthermore, AI developers using Moreh’s software can reduce the time required to initiate training for large AI models by 10%.

Open Source Language Model and Revenue Goals

Moreh recently completed the training of a Korean-language-based language model with 211 billion parameters, which will be released as open source later this year. The startup also started generating revenue in 2021 and aims to reach approximately $30 million by the end of 2023.

“The AI software ecosystem supporting AMD AI hardware continues to grow, providing choice for data scientists and other users of AI as they build the AI models and solutions that will drive the continued growth of this industry,” said Brad McCredie, corporate vice president of data center GPU and accelerated processing at AMD.

Future Plans and Funding Allocation

With the new funding, Moreh plans to allocate resources to research and development, product expansion, and hiring additional staff. The startup currently has 70 employees. South Korean VC firms, Smilegate Investment and Forest Partners, also participated in the Series B funding round.

Editor Notes

In the rapidly evolving field of AI, startups like Moreh are pushing the boundaries of what is possible. The collaboration between AMD, KT, and Moreh demonstrates the importance of innovative AI software tools in enabling the development of advanced AI models and solutions. As the demand for AI continues to grow, it is crucial to have accessible and efficient platforms for AI infrastructure. Moreh’s success in surpassing Nvidia’s DGX in performance speed and memory capacity highlights the potential of its MoAI software. It will be interesting to see how Moreh’s open source language model further contributes to the AI community and what future developments the startup will bring to the industry.

For more AI news and updates, visit GPT News Room.

Source link


Related articles

Los Creadores de Contenido en Google

Title: Google Empowers Web Editors with New Feature Introduction: Google has...

Interview: Lenovo’s Role in Democratizing AI

Leveraging Generative AI: Lenovo's Journey Towards Accessibility and Security Generative...