The company would thereby make it easier for the OEM partners to achieve the goals set in the contracts.
Back in the spring, we wrote that AMD would cover a wide range of performance in the mobile market as part of a Halo-branded Strix Point design, which will be based on three or four models with different parameters. In terms of the top configuration, you can count on 16 Zen 5 cores and an IGP hiding 20 multiprocessors, while the memory capacity can be 128 GB, of which the IGP itself can receive 96 GB.
AMD primarily attacks Apple’s Max designs with development, but it is a problem in the PC market that there are certain contracts with OEM partners regarding how many dedicated GPUs must be ordered in order to obtain various discounts. This is also used by Intel and NVIDIA, so such a high-performance IGP fundamentally shakes up the market structure, as it will cover such a large speed range that the use of a dedicated GPU in many machines becomes redundant.
The entire segment practically faces a problem that is difficult to deal with, because the Apple Max series hides very high-performance designs, and these can only be effectively combated with a similar, integrated solution. So the change is essential from this point of view, but it is a problem for the current contracts. AMD handles this in-house in a very simple way, as the Ryzen AI Max series is considered a dedicated GPU on paper. In other words, the OEM partner who orders one does not have to worry about losing out on discounts for dedicated GPUs, while it is not mandatory to use such a hardware element in order to be able to compete more effectively against similar systems from Apple.
In the longer term, of course, the support of the partners will probably be adjusted, and a system can be prepared that is much better adapted to the needs of the modern market, so the current method is only considered a temporary solution.
Source: prohardver.hu
