Beken also make Risc-v chips, which wouldn't use the Ethos IP, but could use a risc-v accelerator.Having given more thought to the recently announced Arm–Beken partnership, I believe there are several implications worth considering for BrainChip.
Firstly, Arm’s decision to partner with Beken on ultra-low power edge AI suggests that, for now, Arm is prioritizing conventional AI architectures-such as Cortex-M and Ethos microNPUs - over neuromorphic approaches like Akida. If Arm believes it can meet its low-power goals through its existing IP and Beken, it may see less urgency to license or invest in external neuromorphic solutions in the near term.
This could indicate that Arm does not yet view neuromorphic computing as commercially mature or strategically necessary, which in turn would reduce the likelihood of Arm becoming a cornerstone investor in BrainChip - something that had been speculated or hoped for in the past.
That said, the fact that Beken is a Chinese company may introduce limitations for deeper collaboration with Arm, particularly given ongoing US - China tech tensions and export controls. If so, BrainChip might offer Arm a more secure and geopolitically stable pathway to future IP development, especially in sensitive industries like defense, healthcare, and industrial automation.
Then again, Arm might choose to hedge its bets. While Beken focuses on traditional digital MCUs, BrainChip’s Akida architecture provides event-driven compute, sparsity exploitation, and on-chip learning which Beken's technology doesn't offer.
Given all of these considerations, it may be worth raising this partnership and its implications at the AGM to clarify BrainChip’s position and to understand how the Board views evolving opportunities with Arm.
I think most of our hopes lie with Frontgrade and risc-v. If Frontgrade can make their GRAIN chip fly, then others, like Beken should/might follow.