The key to whether Nvidia's acquisition of Arm will succeed lies with the government, not the company

This article is translated from: DesignNews

The market began to generally worry that Nvidia would not be able to acquire Arm. For example, the UK questioned the deal because SoftBank had reneged on its promises when it acquired Arm. In addition, China also needs to approve the deal, and the current Sino-US relations have cast doubt on whether the deal can be realized.

Nvidia's acquisition of Arm is challenging as it requires regulatory approval and comes amid growing global trade hostility and geopolitical tensions between major players, namely China, the United States and the United Kingdom.

There is a precedent for this concern. Between 2017 and 2018, Qualcomm attempted to acquire NXP. Ultimately, Qualcomm raised its offer for NXP to $44 billion, higher than Nvidia's current offer for Arm. At the time, Qualcomm hoped to expand its advantages in mobile communications to the automotive and industrial sectors.

Qualcomm's failure to obtain Chinese regulatory approval was largely due to the tense trade relations between China and the United States. Now, under the current White House administration, relations have further deteriorated. Recently, in addition to Huawei, the US government has also targeted China's most modern semiconductor wafer fab, Semiconductor Manufacturing International Corporation (SMIC).

In addition to the story between Qualcomm and NXP, can we use MIPS as an analogy to see what the future fate of Arm will be?

Similar to Arm, MIPS's IP changed hands several times. Finally, it was finally controlled by a Chinese company.

To better understand these issues, I reached out to Chris Rowen, CEO of BabbleLabs, former CEO of tensilica, and co-founder of MIPS, to ask if MIPS could offer any lessons or cautionary tales for Nvidia’s acquisition of Arm?

"I think the analogy is not perfect. The MIPS architecture was very prominent in the late 80s and early 90s, and was widely licensed. But then they were bought by SGI, specifically for the purpose of developing high-end technology for use in SGI systems. Some general purpose processors continued to work, but almost all embedded technology licenses were closed, so MIPS effectively disappeared from the market in the 1992-1996 period. Most notably, MIPS did not launch any embedded processor core products (although NEC had a project to support Nintendo). This opened the door for ARM to quickly capture the 32-bit embedded core market, most notably with Nokia and Texas Instruments leading the way in mobile phones, and then in the broader space."

Rowen continued: “When MIPS was spun off again in 1996, it took them another few years to actually have a core product, by which time ARM had become a widely used system-on-chip standard. The same could have happened if Nvidia had licensed Arm primarily for internal purposes, but Nvidia seems really committed to ARM being primarily a licensing business. However, you have to ask, what changes would Nvidia make to ARM to get a return on their $40 billion investment? If they made noticeable changes, would others move to RISC V or MIPS, even though they lag far behind in market presence and product breadth?”