Groq to offer on-demand services from Saudi Arabia by year end

20% to 40% of Groq's token-as-a-service traffic will be provided from Saudi

#Saudi #infrastructure - US artificial intelligence infrastructure specialist Groq will provide 20% to 40% of its on-demand token-as-a-service for AI inference by the year-end from a new inferencing data centre soon to open in Saudi Arabia. In a joint announcement made at the Global AI Summit (GAIN) with Aramco Digital CEO Tareq Amin, Groq founder and CEO Jonathan Ross confirmed that Groq has registered a regional headquarters in Saudi Arabia and will be live with an inferencing data centre service by the end of 2024. Future plans developed with Aramco could see Groq expand its Saudi operations to provide one billion tokens-per-second of compute compacity.

Key details of the announcement are as follows:

• AI infrastructure specialist Groq has registered a subsidiary in Saudi Arabia announced plans to open a new services operation in the Kingdom by the end of 2024. The announcement was made by Jonathan Ross CEO Groq accompanied by Aramco Digital CEO Tareq Amin at GAIN 2024, the Global AI Summit in Riyadh.

• According to Ross, Groq is the second largest semiconductor company in AI after NVIDIA, with a community of 444,200 developers.

• Ross announced that Groq incorporated its new regional headquarters in Saudi Arabia last week.

• Groq will deploy at least 25 million tokens-per-second of compute by the end of Q1 2025. According to Ross, the company has the supply chain to build systems to delivery one billion tokens-per-second from Saudi Arabia in the future, which would require about 600 Megawatts of power, which could be provided locally.

• 20% of Groq’s 25 million tokens-per-second capacity will be provided via its new Saudi operations primarily for local consumption in 2024. The company plans to expand this investment to support 25 million tokens, with the ability to serve the Middle East, and parts of Asia, Europe and Africa.

• Groq’s ultimate goal of deploying one billion tokens-per-second of capacity in Saudi Arabia would allow 4 billion people globally to connect with the service.

• Ross also announced that this week Groq has set a new AI inference speed record, processing 544 tokens-per-second.

• The Global AI Summit announcement makes good on a commitment announced during LEAP24, to deploy ‘an enormous amount of compute’ in Saudi Arabia.

Read about Aramco’s large language model METABRAIN:

• World's largest industrial LLM revealed! (Middle East AI News)

VIDEO

• Watch the Groq announcement made at GAIN (YouTube)