2 Sources
[1]
Meta signs major AI deal to train Llama on Oracle Cloud
Oracle Cloud will look to enhance the training and deployment of Llama LLM family Oracle has signed a deal with Meta that will see the latter use Oracle Cloud to enhance the training and deployment of its Llama LLM family. "We just signed an agreement with Meta - for them to use Oracle's AI Cloud Infrastructure - and collaborate with Oracle on the development of AI Agents based on Meta's Llama models," CTO Larry Ellison confirmed. "Oracle Cloud Infrastructure trains several of the world's most important generative AI models because we are faster and less expensive than other clouds...Oracle trained AI models and AI Agents will improve the rate of scientific discovery, economic development and corporate growth throughout the world. The scale of the opportunity is unimaginable." Meta's decision to use Oracle Cloud comes as Oracle CEO Safra Catz praised the company's record-breaking cloud infrastructure growth, driven by continued demand for AI, which requires large amounts of computing power. Despite announcing a healthy 9% year-over-year increase in quarterly revenue, shares in Oracle are down, marking a lack of investor confidence. On the whole, Oracle's second-quarter results were a mixed bag. Quarterly revenue rose 9% year-over-year to $14.06 billion, however analysts had previously anticipated this figure would stand at $14.1 billion. Adjusted earnings per share also fell short of the anticipated $1.48 by one cent. Making up more than three-quarters (77%) of the company's entire revenue was its cloud business, which saw revenue rise by 12% to $10.81 billion. Catz boasted: "Record level AI demand drove Oracle Cloud Infrastructure revenue up 52% in Q2, a much higher growth rate than any of our hyperscale cloud infrastructure competitors." She added that GPU consumption was up 336% in the three-month period. However, the outlook remains precarious, with the company forecasting revenue growth of 7% to 9% in the current quarter, falling short once again of analysts' expectations. TechRadar Pro has asked Oracle and Meta for more details on the deal, but neither company immediately responded to our email.
[2]
Oracle Powers Meta's Llama Models
On Monday, Oracle announced its fiscal 2025 Q2 results, reporting a 9% year-over-year increase in total revenue, reaching $14.1 billion. Cloud revenue totalled $5.9 billion, up 24% in both USD and constant currency. Interestingly, Larry Elison, CTO at Oracle, revealed during the earnings call that Oracle had signed an agreement with Meta to use the company's AI Cloud infrastructure and collaborate on the development of AI Agents based on Meta's Llama models. "Record-level AI demand drove Oracle Cloud Infrastructure revenue up 52% in Q2, a much higher growth rate than any of our hyperscale cloud infrastructure competitors," said Oracle CEO Safra Catz. She added that Oracle expects Cloud revenue to top $25 billion this fiscal year. "We delivered the world's largest and fastest AI SuperComputer scaling up to 65,000 NVIDIA H200 GPUs," said Catz commenting on how GPU consumption rose by 336% during the quarter, reflecting rapid growth in AI-related workloads. However, when compared to industry leaders, Oracle's cloud revenue scale still lags. In the previous quarter, Amazon Web Services (AWS) earned $27.5 billion in revenue, a 19% increase from the same quarter a year ago. Microsoft's Intelligent Cloud revenue reached $24.1 billion, up 20%. Meanwhile, Google Cloud's revenue surged 35%, reaching $11.4 billion. The company shares fell 9% on Tuesday, signalling investor concerns over fierce competition in the cloud business amid higher demand from AI service providers, Reuters reported. "Oracle Cloud Infrastructure trains several of the world's most important generative AI models because we are faster and less expensive than other clouds," said Ellison. Ellison further emphasised the scope of Oracle's AI capabilities, saying, "Oracle trains dozens of specialised AI models and embeds hundreds of AI Agents in cloud applications," including automating drug design, detecting fraud, and analyzing satellite images to improve agricultural output. Oracle has officially entered the multi-cloud era. Following its partnership with Microsoft Azure and Google Cloud, earlier this year, the company also teamed up with AWS to launch Oracle Database@AWS. This service will allow customers to use Oracle Autonomous Database and Exadata Database Service within AWS' data centres. "We're at the very beginning of multi-cloud. It's going to exceed well over $100 million in its first year," said Ellison. He further explained that the first year began when Oracle got AWS and had all three, including Microsoft Azure and Google Cloud. "It will be a multibillion-dollar business, and it'll be a combination of AWS, Google, and Azure, plus all of these Cloud@Customer, Dedicated Region Cloud@Customer. We're going to have hundreds of those regions," he added. Ellison said that the demand from individual customers, large banks, and telecoms is so great that they are building nearly half a dozen data centres.
Share
Copy Link
Meta signs a major deal with Oracle to use its cloud infrastructure for training and deploying the Llama language model family, marking a significant collaboration in the AI industry.
In a significant move that underscores the growing importance of cloud infrastructure in AI development, Meta has signed a major deal with Oracle to leverage its cloud services for enhancing the training and deployment of Meta's Llama language model family 1. This collaboration marks a pivotal moment in the AI industry, bringing together a leading social media giant and a renowned cloud services provider.
Oracle's Chief Technology Officer, Larry Ellison, announced the partnership during the company's fiscal 2025 Q2 earnings call. Ellison emphasized the superiority of Oracle's AI Cloud Infrastructure, stating, "Oracle Cloud Infrastructure trains several of the world's most important generative AI models because we are faster and less expensive than other clouds" 2. This claim is supported by Oracle's impressive 52% revenue growth in its Cloud Infrastructure segment, outpacing its competitors in the hyperscale cloud market.
Oracle's CEO, Safra Catz, reported a staggering 336% increase in GPU consumption during the quarter, reflecting the rapid growth in AI-related workloads 2. This surge in demand has propelled Oracle's cloud business to new heights, with cloud revenue now accounting for 77% of the company's total revenue 1. Oracle expects its cloud revenue to exceed $25 billion this fiscal year, showcasing the lucrative nature of AI-driven cloud services.
The partnership is expected to significantly boost the capabilities of Meta's Llama language models. By leveraging Oracle's advanced cloud infrastructure, Meta aims to enhance both the training and deployment processes of its AI models 1. This collaboration could potentially accelerate the development of more sophisticated and efficient AI agents based on the Llama architecture.
Oracle's partnership with Meta is part of a broader multi-cloud strategy that includes collaborations with other tech giants such as Microsoft Azure, Google Cloud, and Amazon Web Services 2. Larry Ellison projects that this multi-cloud approach will evolve into a multibillion-dollar business, with Oracle planning to build numerous data centers to meet the growing demand from large enterprises.
Despite the positive news surrounding the Meta partnership and strong cloud growth, Oracle's stock experienced a 9% drop following the earnings announcement 2. This decline reflects investor concerns about fierce competition in the cloud business and Oracle's ability to maintain its growth trajectory in the face of established players like AWS, Microsoft, and Google Cloud.
As the AI industry continues to evolve rapidly, partnerships like the one between Meta and Oracle are likely to play a crucial role in shaping the future of AI development and deployment. The collaboration not only highlights the critical importance of robust cloud infrastructure in AI advancement but also signals a new era of strategic alliances in the tech world.
Summarized by
Navi
[2]
Google has launched its new Pixel 10 series, featuring improved AI capabilities, camera upgrades, and the new Tensor G5 chip. The lineup includes the Pixel 10, Pixel 10 Pro, and Pixel 10 Pro XL, with prices starting at $799.
60 Sources
Technology
11 hrs ago
60 Sources
Technology
11 hrs ago
Google launches its new Pixel 10 smartphone series, showcasing advanced AI capabilities powered by Gemini, aiming to compete with Apple in the premium handset market.
22 Sources
Technology
11 hrs ago
22 Sources
Technology
11 hrs ago
NASA and IBM have developed Surya, an open-source AI model that can predict solar flares and space weather with improved accuracy, potentially helping to protect Earth's infrastructure from solar storm damage.
6 Sources
Technology
19 hrs ago
6 Sources
Technology
19 hrs ago
Google's latest smartwatch, the Pixel Watch 4, introduces significant upgrades including a curved display, AI-powered features, and satellite communication capabilities, positioning it as a strong competitor in the smartwatch market.
18 Sources
Technology
11 hrs ago
18 Sources
Technology
11 hrs ago
FieldAI, a robotics startup, has raised $405 million to develop "foundational embodied AI models" for various robot types. The company's innovative approach integrates physics principles into AI, enabling safer and more adaptable robot operations across diverse environments.
7 Sources
Technology
11 hrs ago
7 Sources
Technology
11 hrs ago