Intel has launched its Iris Xe Max discrete graphics processor, aimed at thin-and-light notebooks powered by 11th Gen Core Tiger Lake processors.
The Iris Xe Max is based on the Xe LP graphics architecture and it’s targeted at the ‘thinner the better’ market, i.e. consumer graphics in ultra-slim laptops. The GPU is built on Intel’s 10nm SuperFin silicon fabrication process. It features an LPDDR4X dedicated memory interface with 4GB of memory at 68 GB/s of bandwidth and uses PCIe 4.0 x4. It also has 96 Xe execution units translating to 768 programmable shaders, 96 TMUs and 24 ROPs. The GPU is clocked at 1.65GHz. That should mean a superior experience than what we’ve seen before in ultra-thin laptops.
Alongside those specs, the Iris Xe Max also uses Deep Link, a new technology from Intel that’s previously been referred to as GameDev Boost. Besides helping with 3D graphics rendering, it also provides augmentation of the Xe Media Multi-Format Encoders of the iGPU and DGPU which will help scale video encoding performance. Intel reckons the combination will offer more than double the encoding performance of NVENC on a GeForce RTX 2080 graphics card. That’s meant to be because of a common software framework that connects the media encoding capabilities of the Tiger Lake CPU along with the new GPU so it works better and smarter together.
Continuing the intelligent trend, Deep Link also scales up AI deep-learning performance between the Tiger Lake CPUs and the dGPU.
Keen to game with the Iris Xe Max? That’s an option too. Intel reckons it’ll offer over 30 FPS playable in many AAA games at 1080p. It compares it favourably to the GeForce MX450 as the GPU of choice for ultra-slim laptops.
Exclusively available with thin-and-light notebooks based on 11th Gen Core Tiger Lake processors, Intel also has plans for desktop add-in cards sometime in the first half of 2021. For now, Dell, Acer and Asus are all launch partners with the CPU/GPU combination with the chip debuting on the Inspiron 15 7000, Swift 3x and VivoBook TP470 respectively.