According to the Financial Times, Arm has established a new “solutions engineering” team to create prototype chips for laptops, smartphones, and other consumer electronics and show off the capabilities of its technologies. However, sources close to Arm have denied plans to sell or licence the product and insist that it is only working on a prototype. There are concerns in the industry that Arm plans to either sell chips itself or licence such designs, competing with its licensees.
Arm typically offers its clients licences for its instruction set architecture, logical designs of its CPUs or GPUs, silicon-tested physical designs of its CPUs or GPUs, and various other IP blocks. But this time, a source in the industry told FT that the chip being developed by the solutions engineering team, which is being led by industry veteran Kevork Kechichian, is reportedly more advanced than ever.
Because of the project’s complexity, some experts in the field believe that Arm may develop its own SoCs or, at the very least, licence reference designs rather than IP.
FT’s sources close to Arm claim that this is not the case. According to reports, the company only wants to create a prototype chip or chips to demonstrate the power and performance of its intellectual property. Complex system-on-chip development is also incredibly expensive. A fairly complex 5nm SoC design may cost as much as $540 million (including software), whereas a complex 3nm SoC may cost as much as $1.5 billion to develop, including software.
Arm has not responded to the situation directly, so we are left to infer what the solutions engineering team at the company is focused on. There may be a justification for investing in the company’s chip designs given the rising costs of chip design.
The company might, for instance, be creating customizable silicon-proven reference designs with IP that are ensured to function flawlessly when used with specific process technologies. Few businesses have the funds to invest $500 million to $1.5 billion in a chip design, but they might prefer to licence a product that is certain to work.
In the upcoming years, many of its clients may decide to licence chiplets or chiplet designs instead of IP due to costs, which is another reason for Arm to develop physical implementations of its intellectual property.
But in both situations, Arm might find itself in a battle with its own clients, including Qualcomm, MediaTek, NXP, and other companies that provide chips to the industry. This will undoubtedly increase their propensity to use different instruction set architectures, like RISC-V, and is, of course, a strategic threat to Arm.
Meanwhile, smaller businesses will either go out of business or switch to open-source RISC-V designs, which is also a strategic threat, if they are unable to licence their most recent technology to remain competitive with larger ones.
Also Read:
