Arm – Balancing Act Pt. II

Welcome rebrand and more chip questions.

• While the rebrand of Arm’s portfolio brings much better understanding to its offering, the real question doing the rounds is whether Arm is going to start making its own chips.
• While this will greatly increase revenues, it risks really upsetting the majority of its current customers which is never a good idea for any business.
• Arm will rebrand its compute platforms with a specific name for each market vertical such as Arm Neoverse for infrastructure, Niva for PCs, Lumex for mobile and so on.
• Within each vertical, each generation will have a single version number and a designation to indicate which performance tier it belongs to, such as Pico for the lowest and Ultra for the highest.
• Mali will remain the GPU brand.
• This is a welcome change as all of the confusing Cortex numbering will be gone, and as long as one knows the names (not too difficult), it will be much easier to understand Arm’s portfolio and its proposition.
• The question, of course, is whether this is a prelude to Arm moving to become more vertically integrated and designing its own chips.
• Although nothing has been confirmed, there are hints everywhere.

• • First 2024 20-F: where Arm states on page 24: “We have been, and may in the future be, engaged to advise on or design chips for certain existing customers and other third parties, including affiliates of SoftBank Group, across a variety of use cases and end markets.”
  • This is a change from the 2023 20-F where it stated on page 13: “In the future, we may be engaged to supply chip designs for certain existing customers and other third parties, including affiliates of SoftBank Group, across a variety of use cases and end markets”.
  • This implies that during fiscal 2024, a customer engaged Arm to do more than it usually has done, and if the rumour machine is right, this could have been Meta asking it to design and/or build a data centre chip for AI.
  • This is very far from a confirmation, but it is an indication of how Arm may be planning to change its business model.
  • Second, 2024 litigation with Qualcomm: where Arm said it had given the possibility of making complete silicon chips serious consideration.
  • Because Arm does not currently make any chips, Arm was able to truthfully state that it does not make chips under cross-examination, and there the matter largely rested.
  • Given the pressure that Arm is under to show long-term sustainable high growth from SoftBank, it would be churlish to think that it had not at least considered the possibility of taking this road.
  • Third, Mr Kevork Kechichian: who is currently the EVP of Solutions Engineering at Arm.
  • Mr Kechichian spent 12 years at Qualcomm, where between 2016 and 2019, he ran the Snapdragon team.
  • After a 4-year stint at NXP, Mr Kechichian joined Arm to run Solutions Engineering.
  • Solutions Engineering does not necessarily mean building chips, but according to the trial transcripts, Mr Kechichian attempted to hire Gerard Williams (Oryon CPU core designer and Nuvia founder), which is yet another hint at a move towards making chips.
• Taken together, these hints are an indication that Arm does intend to become a chipmaker in its own right which is something that Arm will need to approach with great delicacy.
• This is because the foundation of the colossal Arm ecosystem is Arm’s independence, meaning that chipmakers can base their processors on the Arm architecture safe in the knowledge that they will be fairly treated.
• If Arm starts making chips, this independence is called into question and already, even just the rumour of Arm making chips has many of Arm’s customers feeling very uneasy.
• There is a way to square the circle which would be for Arm to limit itself to making fully customised chipsets for specific customers in a relationship that looks more like outsourced R&D as opposed to making chips to sell to anyone and everyone.
• The rumoured relationship with Meta where Arm makes a fully custom data centre chip just for Meta could be an example of this.
• Even this looks like a high-risk strategy to me, as I suspect that Arm’s customers could easily still see this as business that they could have engaged in.
• Any move into making chips will also be another shot in the arm for RISC-V increasing its ability to compete with Arm.
• If Arm is going to go down this road (which seems likely given the number of sources), it needs to go on a charm offensive and make the case to its customers that its independence will be maintained.
• Failure to succeed will see customers accelerate their RISC-V programs and give the overall RISC-V ecosystem far more air to breathe than it currently has.
• If RISC-V hits critical mass and standards for higher chip functions begin to form and an ecosystem is created, this will greatly increase RISC-V’s ability to compete with Arm.
• Arm currently has complete dominance in many verticals and is growing share, but for this to continue with Arm as a chipmaker, it needs to strike a delicate balance between maximising its revenue opportunity, keeping its customers happy and ensuring that RISC-V stays in its box.
• Making its own chips is a high-reward but high-risk proposition.