Updated: Apr 6
If you asked most people to name the most powerful forces in the technology industry, it is doubtful that anyone will consider Arm, the chip designer founded in Cambridge 30 years ago.
Despite this, its technology can be used in almost every smartphone and many of the sensors that are ushering in the "internet of things."
The majority of recent Arm news has focused on regulatory concerns regarding Nvidia's proposed acquisition of the company. But, as if to emphasize its commitment to innovation, the company announced its first new chip design in a decade this week.
According to Arm's CEO, Simon Segars, this is a symbol of the company's aspirations in a computing future where all is in the cloud.
"We're anticipating that pretty soon every piece of data that is shared touches an Arm [processor] along the way," he says.
"We've been developing capabilities in our processor that allow more and more complex AI algorithms to be run on the processor itself," he explains - and the new chips will also have a focus on security.
Dedicated AI chips are the next big thing in the semiconductor industry, with specialist companies like the UK's Graphcore already making an impact.
Arm is hoping that Nvidia's technology and financial firepower will give it an edge.
Google, IBM, and Microsoft have all invested heavily in quantum science. However, some of their claims about progress toward a working quantum computer that can solve large real-world problems later turned out to be exaggerated.
Nonetheless, this week, two small British start-ups each claimed a breakthrough.
Quantum Motion, a company founded by Oxford and London academics, claims to have discovered a way to use old-fashioned silicon chip technology to speed up the development of qubits, the fundamental building blocks of quantum computers.
"There are lots of weird and wonderful ways that people are trying to build quantum computers using exotic things like superconducting circuits or trapped ions in a vacuum," explains Prof John Morton, co-founder of Quantum Motion.
"What we're trying to do is to take the same kind of technology which is used to build the silicon chip in your smartphone... in order to build quantum computers that can really scale up to the level needed to solve the really big problems."
Virginia Cipriano Tejel, a quantum engineer, and his doctoral student describes the thrill she felt in the lab when she realized that one electron in a silicon transistor was exhibiting quantum properties.
"You're like, wow, I've measured something really, really small. That's fundamentally something from physics and from nature."
If that is a breakthrough in the production of quantum computers, Cambridge Quantum Computing, a British company, believes it has shown just how path-breaking such a device could be.
This week, it revealed what it coined "ground-breaking proofs that reveal quantum computers can learn to reason under conditions of partial information and uncertainty."
Dr. Mattia Fiorentini, one of the researchers behind the study, says that until now, machines have struggled to think in this way, which comes naturally to humans.
"Classical computers in particular, are very good at executing procedural tasks, they're not good at modelling probability, modelling uncertainty," he explains.
However, he argues that quantum computers would be able to cope with a wide variety of probabilities by their very design - "so there seems to be a sort of natural match here."
The hope is that this new type of computer will be able to work well in places where there is a lot of uncertainty, such as diagnosing medical problems based on scans and forecasting where financial markets will go.
A skeptic would argue that a working quantum computer capable of such tasks is always around five years away. However, researchers claim that progress is now speeding up.
"We can measure it and it's happening," according to Mattia Fiorentini.
So maybe we should brace ourselves for a time when a machine can diagnose any illness or trade the stock market better than any person.
To help their work, Newsmusk allows writers to use primary sources. White papers, government data, initial reporting, and interviews with industry experts are only a few examples. Where relevant, we also cite original research from other respected publishers.
Source - Tech Tent, BBC News