Intel partners with Dell to deliver RAG-based solution powered by Gaudi 3 AI and Xeon 6
Earlier this week, Intel launched the Xeon 6 with P-cores and Gaudi 3 AI accelerators to encourage AI adoption. It also announced that it’s partnering with OEMs, including Dell, to co-engineer and deliver production-ready retrieval-augmented generation (RAG) solutions.
Although Intel didn’t explicitly name the Dell PowerEdge XE9680 server, an article published eariler this year by Dell suggests that this is the solution in question. “The incorporation of the Intel Gaudi 3 accelerator into this ecosystem further amplifies its capabilities,” wrote Deania Davidson, Director of Product Planning and Management for Dell’s AI-optimised compute portfolio.
She added this this offers “clients a choice to tailor their systems to specific computational needs, particularly associated with GenAI workloads.”
RAG enables enterprises to customise large language models (LLMs) without fine-tuning, cutting costs and accelerating time to deployment. Dell is currently co-engineering RAG-based solutions with Intel that leverage Gaudi 3 and Xeon 6.
The Xeon 6 with performance cores boasts twice the performance of its predecessor for AI and HPC workloads. It features increased core count, double the memory bandwidth and AI acceleration capabilities embedded in every core.
Meanwhile, Intel said the Gaudi 3 AI accelerators have been designed specifically for large-scale generative AI, and feature 64 Tensor processor cores and eight matrix multiplication engines.
It also includes 128 gigabytes of HBM2e memory for training and interference, and 24 200 gigabit ethernet ports for scalable networking. Intel also claims the Gaudi AI accelerators have 20% more throughput.
“Demand for AI is leading to a massive transformation in the data centre, and the industry is asking for choice in hardware, software and developer tools,” said Intel Data Centre and Artificial Intelligence Group Executive Vice President and General Manager Justin Hotard.
“With our launch of Xeon 6 with P-cores and Gaudi 3 AI accelerators, Intel is enabling an open ecosystem that allows our customers to implement all of their workloads with greater performance, efficiency and security.”
This latest announcement by the giant chip maker comes as attempts to grab a larger slice of the AI chip market, which Nvidia continues to have a stronghold over.
It also follows news that Qualcomm approached Intel about a potential takeover deal. Despite being one of the most valuable chipmaker in the world, Intel over recent years has been trying to play catchup with other major chip players. In early August, Intel shares sank 26%, the worst day for the company since 1974.
Related stories
NEXT UP
Fabio De Bernardi, VP of Business Development at Adverity: “Having the right partner, whether it’s an agency or a reliable data stack, is essential”
We interview Fabio De Bernardi, VP of Business Development at Adverity where he is responsible for partnerships globally.
International Repair Day 2024: “Everyone that wants to keep things in use for longer is welcome to get involved”
We interview Fiona Dear, Co-Director of The Restart Project, about International Repair Day 2024 and why it matters to both consumers and businesses
Abby Lim, Founder and Director of Absolution Private: “AI has helped quicken and improve the processes of translating our thoughts”
We interview Abby Lim, the Founder and Director of Absolution Private Limited, an integrated marketing agency based in Singapore.