AMD and Korean telco KT back AI software developer Moreh in $22M Series B TechCrunch

Image credit: Jonathan’s Kitchen/Getty Images

Advanced Micro Devices (AMD) and Korean telecom company KT are among the investors in More, which makes an artificial intelligence software tool that optimizes and builds artificial intelligence models. The Santa Clara and Seoul-based startup announced that it has raised a $22 million Series B round, bringing its total revenue to $30 million.

Moreh says its flagship AI software, MoAI, is similar to Nvidia’s CUDA, but compatible with existing machine learning frameworks like Meta’s PyTorch, Google’s TensorFlow, and AI programs and models that previously only ran on Nvidia.

The startup believes that existing AI software is suitable for small-scale AI models that use only a few graphics processing units (GPUs), but inadequate for larger AI infrastructures. According to a recent report, as enterprise-scale artificial intelligence is growing rapidly, companies’ IT infrastructure and data architecture are not suitable for training artificial intelligence models.

Moreh’s AI solutions allow users, including AI developers, data center operators and AI chip makers, to build flexible AI infrastructures, Moreh co-founder and CEO Gangwon Jo told TechCrunch. to create more acceptance, which is important given the general shortage of GPUs worldwide.

The three-year-old startup’s AI software enables GPUs and other AI chips, such as neural processing units (NPUs), to run AI models, including large language models (LLMs) like GPT-3 and T5, without any code changes. explained. (The NPU is a specialized chip designed to speed up machine learning algorithms, while the GPU is used for cloud computing, artificial intelligence, machine learning, and large-scale data analysis where numerical precision is required. .)

KT will work with Moreh from 2021 to build a cost-effective and scalable AI infrastructure with AMD GPUs and MoAI software.

KT claims that Morehs offering has proven that its technology works and that its performance in developing LLM and other AI models has surpassed Nvidias DGX in terms of processing speed and GPU memory capacity. AMD’s MI250 Instinct accelerator powered by the MoAI platform offers 116 percent more graphics processing power than Nvidia’s A100, according to Moreh, who also said AI developers can cut a tenth of the total time required to start training large AI models from Reduce through your software.

Recently, the startup completed a Korean-based LLM tutorial with 211 billion parameters and will release the model as open source later this year.

Moreh started generating revenue in 2021 and aims to reach around $30 million by the end of 2023, Joe told TechCrunch.

The AI ​​software ecosystem that powers AMD’s AI hardware continues to grow, said Brad McCurdy, VP of Data Center GPU. And processing is accelerated at AMD, it said in a statement.

Moreh will use the money for research and development, product development and hiring additional staff, which now has 70 employees. South Korean VC firms Smilegate Investment and Forest Partners, Morehs’ current investor, also participated in the Series B round.



#AMD #Korean #telco #software #developer #Moreh #22M #Series #TechCrunch
Image Source : techcrunch.com

Leave a Comment