There was a stark study in contrasts at the top of the chip industry this week. Nvidia, which recently overtook Intel to become the world’s most valuable chipmaker, reported blowout earnings. Jensen Huang, its founder and chief executive, used the moment to lay out his vision for what comes next.
At the very moment that he was talking, Intel tried to revive its flagging support on Wall Street by accelerating the repurchase of $10bn worth of shares. As a gesture of confidence in its ability to keep churning out cash, it was certainly powerful: the buyback matches the total free cash flow that Nvidia has generated over the past three years combined. But it couldn’t hide the fact that Intel’s manufacturing process technology has hit a wall.
The failure has left a big question over where Intel’s future lies, and whether it even has a place at the leading edge of chip manufacturing. But as it flexes its muscles and surveys the landscape before it, what will Nvidia, the new king of the heap, become?
Over the past 21 years, the Californian company has taken its graphical processing units, or GPUs, from their original market in gaming PCs to data centres, where their parallel processing capabilities have made them the main engines for the data-intensive task of training AI systems. With its attempt to buy Arm, the SoftBank-owned chip design firm, it is now trying to consolidate that position, while also getting its first toehold in some of the biggest current and future markets for silicon.
Data centres, according to Mr Huang, are the new “computing unit”. Rather than write programs that run on a single processor or server, coders will one day write software designed to run on an entire data centre: what happens behind the scenes, as data shuttles between machines and large-scale internet services are parcelled out in the most efficient way, will be the concern of a company like Nvidia, not the programmer. Nvidia already has the GPUs and networking technology to fulfil much of this: what it lacks is a base in CPUs, or central processing units, the core of Intel’s business.
Years of design work by Arm to match Intel in data centre CPUs have finally been paying off. One sign of this: Amazon, the biggest provider of cloud computing infrastructure, is now designing its own Arm-based server chips, under the suitably weighty name Graviton.
Owning all the intellectual property for the silicon that powers giant data centres would give Nvidia a tidy position that even Intel — which makes more than half its profits from this market — could not match. But it isn’t clear an acquisition is necessary: Nvidia can (and already does) licence IP from Arm. And there is a good chance that the pursuit of an acquisition, which it has yet to confirm, will antagonise others in the industry and spark unpredictable competitive responses.
A deal would leave Nvidia in the position of not only selling its own silicon, but also licensing foundational IP that other companies — some of them Nvidia’s competitors — need to create their own chips.
The chip industry already has a company like this. Qualcomm, which invented mobile communications, has long come under fire from some of its biggest customers and rivals over its techniques for generating the maximum income from selling both IP licences and its own chips. A US appeals court overturned a government antitrust case against Qualcomm last week, ruling that its business practices, though “hyper-competitive”, didn’t break the law.
The pursuit of Arm may be motivated mainly by Nvidia’s ambitions in data centres, but it also opens up a broader landscape. Nvidia is already active in other AI markets, like robotics and self-driving cars. And Arm would give Nvidia its first entry into the “internet of things” and — by far Arm’s biggest current market — smartphones.
Removing the neutral ownership of SoftBank would have unpredictable consequences. Apple — which fought a long and wearying battle against Qualcomm through the courts before reaching a settlement last year — is one giant Arm customer that Nvidia would need to find common ground with. Qualcomm itself uses Arm technology in the CPUs embedded in its all-in-one chipsets that run high-end smartphones.
These companies don’t compete directly with Nvidia at the moment, so being forced to turn to it for a technology licence may not cause immediate problems. But if Nvidia ends up paying more than $32bn for Arm — the price that has been discussed — it will have strong incentives to maximise its licence revenue to recoup its outlay.
And in the background, there will always be a lingering question about where a newly emboldened Nvidia will turn next as it looks for new silicon worlds to conquer.
"chips" - Google News
August 21, 2020 at 10:01AM
https://ift.tt/3iYb8ub
Nvidia flexes its muscles as the new king of chips - Financial Times
"chips" - Google News
https://ift.tt/2RGyUAH
https://ift.tt/3feFffJ
Bagikan Berita Ini
0 Response to "Nvidia flexes its muscles as the new king of chips - Financial Times"
Post a Comment