The semiconductor giant says the new Xeon 6776P processor serving as the host CPU for Nvidia’s Blackwell Ultra-based DGX B300 system features its new Priority Core Turbo technology as well as Intel Speed Select – Turbo Frequency to enable ‘customizable CPU core frequencies’ for the purpose of maximizing GPU performance
Intel said one of its three newly revealed Xeon 6 processors designed to “boost GPU performance across demanding AI workloads” will serve as the host CPU for Nvidia’s new Blackwell Ultra-based DGX B300 system.
The Santa Clara, Calif.-based company announced the three new Xeon 6 processors with performance cores on Thursday, saying that they feature the company’s new Priority Core Turbo technology as well as Intel Speed Select – Turbo Frequency to enable “customizable CPU core frequencies” for the purpose of maximizing GPU performance.
[Related: Qualcomm’s Plan To Sell Server CPUs Includes A Partnership With Nvidia]
Intel made the announcement as CEO Lip-Bu Tan tries to chart a new path forward for the company’s AI strategy, which includes the Gaudi 3 accelerator chips that didn’t meet its modest sales goal for the product last year.
While Intel continues to push Gaudi 3 as a more affordable alternative to Nvidia’s GPUs and develops a future rack-scale platform based on its next-generation Jaguar Shores chips, the company has also made a priority to position Xeon as the best host CPU for AI servers.
“These new Xeon SKUs demonstrate the unmatched performance of Intel Xeon 6, making it the ideal CPU for next-gen GPU-accelerated AI systems,” said Karin Eibschitz Segal, interim head of Intel’s Data Center Group, in a statement.
“We’re thrilled to deepen our collaboration with Nvidia to deliver one of the industry’s highest-performing AI systems, helping accelerate AI adoption across industries,” she added.
Intel said the new Xeon 6776P processor will serve as the host CPU for Nvidia’s DGX B300 system, with each node consisting of two Xeon chips and eight Blackwell Ultra GPUs.
“With robust memory capacity and bandwidth, the Xeon 6776P supports the growing needs of AI models and datasets,” the company said.
The new Priority Core Turbo feature in these new Xeon processors “allows for dynamic prioritization of high-priority cores, enabling them to run at higher turbo frequencies,” according to Intel. “In parallel, lower-priority cores operate at base frequency, ensuring optimal distribution of CPU resources,” the company added.
“This capability is critical for AI workloads that demand sequential or serial processing, feeding GPUs faster and improving overall system efficiency,” Intel said.
Revealed at Nvidia’s GTC 2025 event in March, the DGX B300 is the x86-based complement to the AI infrastructure giant’s flagship GB300 NVL72 platform, which includes its Arm-based Grace CPUs and Blackwell Ultra GPUs to provide optimal AI performance.
With a single DGX B300 capable of 72 petaflops of 8-bit floating-point (FP8) and 144 petaflops of 4-bit float-point performance, the system serves as the building block for the B300-based DGX SuperPod cluster. This SuperPod provides 11 times faster inference on large language models, seven times more compute and four times larger memory compared to a Hopper-based DGX cluster, according to Nvidia.
