How ARM set itself up for a $32 billion acquisition
22 July 2016 | 0
The TV you watch perhaps has an ARM processor chip in it, as might your tablet or smart phone. So does the Amazon Echo that helps switch on the lights and air-conditioner through voice commands.
That is just a microcosm of how deep ARM goes in our daily lives. ARM’s low-power chip designs have revolutionised mobile devices and are now powering smart home devices, smart meters, weather sensors, medical devices, and industrial equipment.
ARM chips also are inside many sensor devices used in the fast-growing internet of things market. The company set itself up for growth in IoT with its dogged focus on low-power chips since the 1990s, and that vision has paid off with SoftBank announcing plans to buy ARM for a stunning $32 billion (€29 billion).
Some analysts believe SoftBank is overpaying, but the investment could eventually pay off. Some estimate 20 billion to 50 billion connected devices will be online by 2020, and those numbers represent a big growth opportunity for ARM.
“ARM will be an excellent strategic fit within the SoftBank group as we invest to capture the very significant opportunities provided by the internet of things,” Masayoshi Son, CEO and chairman of SoftBank, said in a statement.
SoftBank’s offer was compelling for shareholders and offers the chip company to grow, Simon Segars, CEO of ARM, said in a video explaining the sale.
ARM designs chips it licenses to manufacturers, which can tweak them to meet the needs of their devices. Apple uses ARM architecture in its iPhone and iPad, and other licensees include Samsung, Microsoft, Nvidia, and AMD. ARM generates revenue through licensing and royalties, unlike rival Intel, which designs and manufactures its own chips.
From its earliest days, ARM’s focus on power efficiency over performance hasn’t wavered, setting the company up for success in mobile devices, and now IoT, analysts said. An early ARM chip was used in Apple’s Newton handheld, which shipped in 1993 but wasn’t a commercial success.
“The only other guys that had a shot at [the mobile and IoT markets] were MIPS, which is now a part of Imagination Technologies,” said Nathan Brookwood, principal analyst at Insight 64. “But MIPS fell off the horse.”
In IoT, a range of devices are powered by external power sources, but a larger number rely on batteries and energy harvesting. ARM processors like Cortex-M0 are targeted at the IoT, but many older microcontrollers are still being used because of their power efficiency features.
ARM started in 1990 as a spin-off from a collaboration between Apple and Acorn Computer Group. Acorn was the brains behind the first ARM RISC chips, which appeared in a personal computer called Archimedes in 1987.
ARM’s early years were mixed, but the company started gaining attention as shipments of devices like BlackBerry started picking up in the late 1990s. Revenues started exploding in 2005, and that year, ARM shipped 1.4 billion chips for mobile devices, crossing the 1 billion threshold for the first time.
The iPhone, which launched in 2007, added to ARM’s fortunes, and more than 90 billion ARM-based chips have shipped so far. In 2015 alone, 15 billion ARM chips shipped.
ARM maintained its heritage of developing low-power processors at a time when Intel and AMD cranked up clock speeds and power draw in their PC chips. The iPhone was a big breakthrough, and the mobile explosion caught rivals like Intel off-guard. Intel then took the ARM route and started focusing on power efficiency.
IoT is again changing the shape of the chip industry. Intel in April laid off 12,000 people as it redirected focus from PCs to IoT, data centre products, and memory.
Over the decades, ARM microcontrollers were easily available from companies like Marvell and Texas Instruments to build and test devices, and that strategy helped build a large ecosystem, Brookwood said. By comparison, Intel chips weren’t as easily available, and that’s another reason ARM is built into more devices.
While ARM is growing in the IoT, it has weaknesses. ARM-based chips may not be able to handle demanding tasks in medical devices, digital advertising screens, or gambling machines, where graphics and processing are key. Those devices will better run on x86 chips, said Jim McGregor, principal analyst at Tirias Research.
Mobile and low-power IoT devices are tailor-made for ARM, and they are markets the company should continue chasing, McGregor said. For SoftBank, those markets are low-hanging fruit that will generate instant revenue.
ARM chips could also find their way into storage and networking devices, which are growing more important to the IoT, McGregor said.
The company plans to speed up processor development and double its headcount over the next five years with SoftBank’s backing. But questions remain about what ARM will do with the extra resources.
ARM has processor designs and programs in place for a range of chips, from IoT microcontrollers all the way up to beefy servers, for the next couple of years, and most semiconductor companies already have ARM licenses. So there are limited opportunities for the company to expand with its IP licensing business model, Brookwood said.
ARM’s price was depressed, and SoftBank may have felt it was a good time to acquire the company, said Richard Fichera, vice president at Forrester Research.
But ARM is still a small company, and its unique business model won’t change the way chipmakers like Intel, Nvidia, or AMD do business.
“There’s nothing [SoftBank] can do to make ARM five times their size,” Fichera said. “I don’t think it’ll transform the industry in any way.”
Though SoftBank may have overpaid, ARM is well positioned to make money with its intellectual property business, and the IoT segment is a big opportunity, Fichera said.
“ARM is well positioned to make money hand-over-fist with their IP business,” Fichera said.
IDG News Service