Intel Nova Lake CPUs will come with multiple SKUs with 12 Xe3P iGPUs, while entry-level models will utilize the internal 18A-P node.
Several Intel Nova Lake CPUs Will Feature An Upgraded Xe3P iGPU With Up To 12 Cores
The Xe iGPU has seen immense upgrades since the release of Xe2. Lunar Lake turned out to be a great showcase for Xe2, followed by the Arc B-series discrete lineup, and since then, Xe3 has been showcasing its strengths across mobile platforms such as laptops and handhelds. Looking ahead, Intel will be introducing its Xe3P architecture, internally codenamed Celestial. The new iGPU will offer increased performance and enhanced features, and Nova Lake will be the first family to use this architecture.
Now, new details have emerged as to which SKUs will feature the Xe3P iGPU architecture. This is going to be a high-end iGPU for a range of SKUs in the Nova Lake lineup. As per Jaykihn, the Nova Lake-H (Mobile) family will feature a 12 Xe3P segment, comprising multiple SKUs. These will be the replacements for Panther Lake 12 Xe3 SKUs. The rest of the 4 Xe and 2 Xe configurations will be based on the existing Xe3 architecture.
We know from previous reports that Intel is also working on a Nova Lake Desktop chip with the same 12 Xe3P cores. This is going to be an 8-core processor with just 8 E-cores. This chip is designed for the edge market, but since it is a desktop design, it will be really interesting to see how the iGPU performs on a desktop platform against AMD's Ryzen APUs that offer up to 8 RDNA compute units.
Based on recent Xe3 vs RDNA 3.5 benchmarks, Intel has a much bigger lead over the mainstream Ryzen parts. In some cases, the lead is almost double, so Intel could offer its own Nova Lake CPUs with beefier iGPUs to tackle the competition. Intel is also working on its bigger "Halo" SKUs and a custom-silicon design with NVIDIA that will house RTX tiles. So Intel has a range of GPU-Heavy chips coming out in the coming years.
AMD Olympic Ridge vs Intel Nova Lake-S:
| CPUs | Intel Core Ultra 400 | AMD Ryzen 10000? |
|---|---|---|
| Family | Nova Lake-S | Olympic Ridge |
| Architecture | Coyote Cove (P-Core) Arctic Wolf (E/LP Core) | Zen 6 |
| CPU Process | TSMC N2P | TSMC N2P |
| Core Count (Max) | 52 | 24 |
| Thread Count (Max) | 52 | 48 |
| Max P-Cores | 16 | 24 |
| Max E-Cores | 32 | N/A |
| Max LP-E Cores | 4 | N/A |
| Max Cache (L2+L3) | 160-320 MB | 96 MB L3 |
| Max bLLC Cache | 144-288 MB | 64 MB per stack? |
| DDR5 (1DPC 1R) | 8000 MT/s CUDIMM - Yes | 7200 MT/s? CUDIMM - Yes |
| PCIe 5.0 Lanes (Max) | 36 | TBD |
| PCIe 4.0 Lanes (Max) | 16 | TBD |
| Socket Support | LGA 1954 | AM5 |
| Max TDP (PL1) | 125-175W | 125W+ |
| Max Power | ~700W (Dual) ~350W (Single) | TBD |
| Launch | 2027 | 2027 |
18A-P is used for entry-level Nova Lake Chips, Rest Built On TSMC N2P
The other interesting detail mentioned by Jaykihn is that Intel will be building some of its entry-level chips using its own internal 18A-P process technology. These chips will feature a 4+0 core configuration, while the rest of the chips in the mainstream/high-end category will utilize the TSMC N2P process.
Intel's Panther Lake currently has its compute tile built on the 18A process technology. The 18A-P technology is primarily for external customers, but Intel has recently reconsidered using both 14A and 18A-P for its internal products. 18A-P, just like 14A, has a lot to prove, & with risk production already begun, we will soon see the first showcase of this node as early as 1H 2027.
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
