Intel Unveils Lunar Lake Architecture: New P and E cores, Xe2-LPG Graphics, New NPU 4 Brings More AI Performance

New Graphics: Intel Xe2, 2nd Gen Arc Xe Core For Mobile

Along with Lunar Lake, Intel has just unveiled its Xe2 graphics architecture for mobile, supported by the 2nd Generation Arc Xe Core. On paper, it offers an extraordinary bump in performance and efficiency. Aside from gaming, which we don't think the 4P+4E part would cut it, we've opted to focus on the critical takeaways of Intel's presentation for graphics, including the media engine within.

Intel's introduction of the Xe2 architecture significantly improves computational capabilities by providing up to 67 TOPs and offering increased ray tracing units compared to Xe-LPG on Meteor Lake. According to Intel, the 2nd gen Xe-cores offer 1.5x faster graphics performance than Meteor Lake, which is helped and achieved by the new XMX engines. Enhanced XeSS kernels deliver improved graphics and compute performance.

One element that Intel looks to have changed from Meteor Lake is that it offers more flexible and higher-quality display outputs. Within the Display engine, the streams in the dual-pixel pipeline can be combined for multi-stream transport. With this architecture, ports will be available in four locations, which will be flexible for connectivity. An eDP port is also provided in Intel's configuration, which will augment the display to set high resolutions and refresh rates for the output on high-end, premium, and capable displays.

Intel's eDisplayPort 1.5 includes the panel replay feature integrated with adaptive sync and selective update mechanisms. This helps decrease power consumption by refreshing only the parts of the screen that change instead of the entire display. These innovations save not only energy but also improve visual experiences by reducing display lag and increasing sync precision.

Portraying the pixel processing pipeline is one of the fundamental bases on which Intel's display engine sits, enabling six planes per pipeline for advanced color conversions and compositions. In addition, it integrates hardware support for color enhancement, display scaling, pixel tuning, and HDR perceptual quantization, ensuring that the graphics on the screen is vibrant and accurate. The design is quite flexible, highly power-efficient, and performance-engineered to support various input and output formats, at least on paper. Intel hasn't provided any quantifiable power metrics, TDPs, or other power elements so far.

When considering compression and encoding, the architecture Xe2 extends up to 3:1 display stream compression visually losslessly, including transport encoding for HDMI and DisplayPort protocols. These chip features further reduce the data load and maintain high resolution at the output without losing visual quality.

Intel's adoption of the VVC codec is a big deal for video compression technology improvement. This codec offers up to a 10% reduction in file size compared to AV1, supported by adaptive resolution streaming and advanced content coding for 360-degree and panoramic videos. This will ensure lower bitrates for streaming without losing quality—an essential aspect for modern multimedia applications.

The Windows GPU software stack is robust, from top to bottom, with the support of D3D, Vulkan, and Intel VPL APIs and frameworks. This means that combining these qualities provides comprehensive support for the varied runtimes and drivers in the market, thus increasing its overall efficiency and compatibility in different software climates.

Intel's Xe2 and second-generation Arc Xe Core improve performance, efficiency, and flexibility significantly. These innovations strengthen Intel's position in the competitive landscape of solutions for mobile graphics, with reinforced capabilities across display, media, and compute operations.