Intel Xeon6+ to enable demand for more compute

The Intel Xeon 6+ processors extend the Xeon 6 family with a focus on performance density, power efficiency, and operational scale for cloud‑native, agentic AI‑driven, and network‑intensive workloads.

With the world witnessing a significant demand for denser CPU to run agentic AI applications, Intel’s new XEON 6+ processors that were unveiled at Computex 2026, is expected to enhance the capabilities of the data center for the modern architecture.

The Intel Xeon 6+ processors, an expanded 800 Series Ethernet portfolio, features the Intel Ethernet E835 controllers and network adapters. These developments highlight a clear industry shift where AI becomes more agentic, and the CPU is re‑emerging at the center of modern AI infrastructure.

According to Kevork Kechichian, executive vice president and general manager of Intel Data Center Group, with Xeon serving as the control plane, Intel is taking a systems‑level approach to performance and efficiency at scale, with Intel serving as the control panel.

“We are seeing this significant demand for more CPUs. The landscape is transforming and we still have a need for acceleration. And from our perspective, we changed our narrative a bit more into workload focus. We want to make sure that we are creating that narrative and those solutions for our partners. We have lots of assets within Intel, from IP to technology to the foundry and different opportunities on scale-up, scale-out networks, different accelerators. And we are here today and the coming weeks and months to start delivering those solutions to our partners. And again, this is very exciting from a Xeon 6 Plus perspective,” he said during a briefing at Intel House in Computex.

The Intel Xeon 6+ processors extend the Xeon 6 family with a focus on performance density, power efficiency, and operational scale for cloud‑native, agentic AI‑driven, and network‑intensive workloads. Capable of addressing orchestration, concurrency and data movement demands of emerging agentic AI, Xeon 6+ emphasizes scale‑out performance. The chip’s optimization makes room for new AI workloads without requiring disruptive data center redesigns.

“As AI becomes more agentic, the constraints shift to orchestration, concurrency, and data movement. That shift reinforces a core reality: the CPU remains the control plane for the modern AI infrastructure. With Xeon 6+ and Ethernet E835, we’re tightly coupling compute and networking to reduce bottlenecks and enable efficient, secure scaling of real‑world agentic workflows,” Kechichian said.

Intel Xeon 6+ processors are already being tested within telecom network infrastructures and configured into data center systems with platforms available across the ecosystem. These include servers, networking and integrated solutions from and used by ASUS, Dell Technologies, Ericsson, GIGABYTE, HPE, Lenovo, Supermicro – and others developing on Xeon 6+.

Need for more compute power

In a panel discussion at Intel House in Taiwan during Computex, Magnus Ewerbring, CTO for APAC at Ericsson, shared that the Xeon 6+ is enhancing performance for Ericsson and customers.

“With the Xeon 6+, we see a 60% increase in performance per watt. That's a significant improvement that immediately is attractive for our customers. We know the network load will increase in the next five years by up to three times. For that, we need an architecture that is stable and robust, where we have a level of freedom also how we distribute our compute on different parts of the hardware layer and draw advantages of that. That's what we feel we have now with the Xeon 6+, where we have hardware support, accelerators for certain loads, and then a more general compute for certain loads. That gives us flexibility to adapt our application over time as load traffic patterns change as well,” Ewerbring said.

For Benny Shieh from Giga Computing, a subsidiary of GIGABYTE and a leader in accelerated computing and infrastructure solution, the platform design based on the Intel Xeon 6+, combined with technology like the Intel AEP, gives more insight.

“Instead of only seeing the total server power consumption, the operator can understand which applications or services are actually consuming power,” he said.

Giga Computing announced support for Intel Xeon 6+ processors across GIGABYTE high performance server platforms. The new line of Efficient-core processors delivers major gains in compute density, throughput, and power efficiency for hyperscale, cloud, and telco deployments.

Managing the supply chain

When it comes to the supply chain of the new chips, Kechichian highlighted that all the data center products are within the Intel foundry.

“Within that framework, we're optimizing the throughput of our products. If you look at these chips, even if we say one technology as a headliner, these things have multiple dyes. There's lots of trade-offs that we do. This is something we discuss, especially given the supply constraint situation. Almost every 48 hours we see how to optimize. What's the ratio of one product slowing down versus how many down the other product? Although it's not particularly interesting, it's actually quite disruptive to our partners and customers. It creates this opportunity for us to be more optimized from an R&D perspective and harvest more wafers. That's what the team has been doing for the last six to nine months,” he said.

According to Kechichian, Intel has supply from Samsung, SK Hynix and Micron, with a significant quantity to support them, although the price is getting higher compared to the total cost of the GPU server, and customers still can’t accept it.

“It's a problem, but not big for us. From our perspective, we're not an OEM. We buy our machines from OEMs and we put them in our own data centers. Memory is a very serious problem for us. But I would say it's rather for our customers because it affects the ready for service date. And it also affects pricing, which is something we have very little to do. We hope they can be resolved someday. But at this point, the only thing we can do is just have a good relationship with OEMs and have a good relationship with all the suppliers,” he explained.

This includes making sure Intel has enough allocation beforehand, which they normally do.