Intel has been working on its own graphics chips for some time. After the graphics architecture Xe was already integrated as an integrated graphics unit of the Tiger Lake processors (alias 11. Core -i generation) debuted, now the first stand-alone graphics chip follows. It’s called Iris Xe Max and is based on the DG1 chip that has been distributed to interested developers since the beginning of the year.

Iris Xe Max is initially only intended as an additional GPU for notebooks. With a Thermal Design Power (TDP) of 15 watt, the chip plays in a league with low-end -GPUs like Nvidia’s GeForce MX 350. It also ends up roughly in this range with regard to 3D performance – that’s not enough for detailed 3D game worlds.

Intel promises a 3D performance for the Iris Xe Max, which in roughly corresponds to that of Nvidia’s low-end GPU GeForce MX 350.

(Image: Intel)

LPDDR4X and no GDDR RAM Instead, Intel rather, that the GPU lends a helping hand to creative users: Iris Xe Max contains the same functional units and an identical number of shader cores (96 Execution Units, 768 Shader) like the integrated Iris Xe, but has its own memory (4 GByte LPDDR4X with 68 GByte / s – yes, no GDDR5 / 6) and a higher clock rate of 1, 65 GHz.

Even more: AI routines or video programs can Distribute the processing load on both GPUs, which together upscale images and convert videos. At the start, the latter only works when converting several videos at the same time, but Intel promises that common single-stream encoding is already running across all media units in the laboratory. This function is to come in the first half of the year 25 for everyone.

Iris Xe Max – technical data

(Image: Intel)

3D here or there At 3D graphics, on the other hand, only ever use one GPU – not surprising, because AMD’s efforts a few years ago to have an integrated GPU run together with an additional chip also failed due to the complexity. The Iris Xe Max is the standard setting for games, but that’s not nailed down: Some games like the MOBA Dota 2 are heavily dependent on latencies between the CPU and GPU, so that higher frame rates can be achieved there if the integrated Iris Xe is used instead – which is then also set via the driver profile.

Advantage: When the Iris Xe Max is put to sleep, the processor has that The notebook’s cooling system is available alone, which potentially enables higher clock frequencies. In general, Intel promises that the waste heat budget provided by the cooling system will be dynamically distributed between the processor and the additional GPU, depending on the load situation. This makes sense and is even common in the industry for flat notebooks: Even with notebooks with Nvidia’s Max-Q graphics chips, the cooling system is designed for an average shared load that is lower than the sum of the separate TDPs of the CPU and GPU.

Notebook debut The first notebooks with Iris Xe Max come from Acer, Asus and Dell .

(Image: Intel)

The first notebooks with Iris Xe Max should be in stores this year. Acer unveiled its Swift 3x a few days ago, and the same applies to the VivoBook TP 470 from Asus. The third in the league is the Dell Inspiron 15 7000 2-in-1, for which no information about local availability was available at the time of going to press.

Beginning 2021 Iris Xe Max will then also come as a graphics card for desktop PCs. However, Intel does not want to sell them individually, but only to deliver them to PC manufacturers who then build them into comparatively inexpensive complete computers. It is to be expected that more than the Watt TDP of the mobile version are provided for the desktop graphics card .

(mue)