How AI has given a boost to the chip design ecosystem

How AI has given a boost to the chip design ecosystem

Many companies and startups are building chips for specific AI use cases

Sujit.John@timesgroup.com

01.12.2021

Speech recognition had an error rate of 16% around the time Apple’s Siri was launched early last decade. Which means it wouldn’t understand many of the words/sentences we spoke to her, and so it provided no answers, or wrong answers. But as we spoke to her more, she learnt from it. Today, speech recognition systems have significantly lower error rates, they can even understand accents. But it has taken years to get there.

If you need to build great AI systems quickly, you need to throw a lot of data and compute power into it. More and more use cases are emerging where the AI system needs to instantaneously understand what’s going on to be able to respond to it. Braking by autonomous cars is a classic one.

Chips with AI acceleration, and chips that are designed for AI are coming in to deal with this. Semiconductor companies, startups, and even those like Google, Apple, Amazon and Facebook, for whom AI is central to what they do, are all developing such chips. A lot of this work is happening in India, one of the world’s foremost chip design hubs.

Srikanth Velamakanni, cofounder & CEO of analytics company Fractal Analytics, says such chips are essential to deal with the massive volumes of data that many systems now generate. “One flight of an aircraft generates more data than Google generates in a day. Because it’s got so many sensors and such high velocity data coming through,” he says. It’s similar in factories and industrial equipment. “You have to comb through all this data in real-time to see what may be failing. Human beings are not capable of that. It also needs hybrid computing, a combination of edge and server. We are looking at Intel’s new processors with AI acceleration to see how much of a performance boost we can create in these kinds of applications,” he says.

Fractal is also looking at these chips for a solution they call Customer Genomics, which mines massive customer data, like in banks, and recommends the next best action for the customer.

Ruchir Dixit, India country manager at semiconductor tools maker Siemens EDA, says such analytics is possible to do in software, but it won’t be fast enough. Many have used GPUs, because they are designed for heavy-duty graphics processing, but even those fall short for emerging requirements. “A machine learning algorithm implemented on hardware is always orders of magnitude faster. When I launch a software programme on my laptop, it has to find time from the CPU, even as the CPU deals with other computations it may be involved in, like an on-going video call. But if you put it in hardware, it doesn’t care what else you are doing, it will do it immediately because that’s what it is designed to do,” he says.

Different AI functions may require their own different chips. Understanding images in a car, with its limited power and limited ability to absorb heat, may require a different AI chip from that in a factory, which has AC power coming in and where how much heat the hardware dissipates may not be a concern.

Alok Jain, VP of R&D at semiconductor tools company Cadence Design Systems, says there are very specific chips for speech recognition, for face recognition. “It all depends on the level of complexity, the level of cores that are required on the size of data available to you. It depends on how dependent the variables it is dealing with are – if they are dependent, the communication between the cores becomes important,” he says.

Given the variety of chips needed, even startups, he says, have found a great opportunity to serve various niches. SambaNova, Groq, Cerebras in the US are among them. In India, there are those like AlphaICs and QpiAI.

IBM recently announced the Telum processor, its first processor to contain on-chip acceleration for AI inferencing and targeted at the financial services industry. Google is known to be building a large team in India for its chip design.

Prakash Mallya, MD of Intel India, says the choice of whether to use a chip designed for a specific AI application, or a general purpose chip with AI acceleration will depend partly on software capabilities within the organisation. The former, he says, requires more software capabilities to program and to build the IT stack.

Never dilute standards for recognition at workplace

Never dilute standards for recognition at workplace

By Prabir Jha

01.12.2021

At Diwali, we all received numerous messages. Many were the usual forwarded types, some were mass messages on egroups, some clear visual re-pastes, and still fewer were typed or customised for you, mentioning your name.

How did your response vary? Did some mean more to you than the others. Not that getting a cut-paste greeting had an indifferent intent, but it just did not connect. Worse, we were both recipients and transmitters, and responders. And we did also respond differently.

This is the issue with much of the recognition efforts in organisations. They either become de rigueur, a ritual that just stayed a checkbox. The Diwali moments got me to cull some learnings from experience for most firms and their leaders…

Clarity over intent: Corporations often get their intent muddled. What do you want to recognise? Outcome? Behaviour? Then, who do you want to recognise? Individuals or cross-functional teams? When do you want to recognise? Now or later? Who should recognise? Someone really big or someone closer to the awardee? The simple answer would be a medley. But think well on these questions. All of these will make or break your recognition culture.

Think of effect, not ease: Very often, firms adopt cookie-cutter thinking while designing their recognition programmes. What is easy to do? What can be scaled up? What can get system-generated? And they kill the spirit of recognition. If you do recognise, do it well. How will this be experienced by the recipient? Work backwards from there.

Harmonise but differentiate: Much of recognition can be a timely thank you, but an authentic one. Different levels of impact or behaviour must be expected across the hierarchy. Don’t dilute your recognition standards. Differentiate. You may have a broad framework that enables recognition at various levels of business. This helps interconnect but always differentiate, otherwise it will become another misplaced example of tokenism.

Heart over mind: While one makes a thoughtful judgment on what to recognise, the recognition philosophy is essentially a heart exercise. It must make the person feel appreciated, recognised and wanted. It must trigger a sense of pride and inspiration, after seeing the recognition to that person. It must make lunch table conversations positive, not toxic. So, think well about your recognition . Don’t distribute recognition like alms. You thought you bought peace? You just trivialised your recognition. Recognition does not breed complacency.

Period: I have had this conversation with many CEOs. Many are hesitant in appreciating, as they believe recipients will become lazy. And those upset will drag their feet. To me, both are untenable arguments. As I told one CEO once, “If you are waiting for them to reach Mount Everest before you appreciate, they won’t go beyond the first base camp.” Every recognition does not have to be an Oscar. But every positive word allows you to raise the bar.

The writer is the founder & CEO of Prabir Jha People Advisory

5,000 apply for ex gratia, many more sans Covid death docus

5,000 apply for ex gratia, many more sans Covid death docus

Amrita.Didyala@timesgroup.com

Hyderabad:01.12.2021

Around 5,000 people have so far applied for Covid compensation in Telangana, as per health authorities. With many more returning from Meeseva centres each day, hoping to get proper certificates so they can apply for the compensation.

The compensation for family members of Covid victims was announced by the Centre a month ago. The Telangana government released guidelines for payment of the ex gratia recently.

So far, 6.75 lakh people in Telangana have been infected by Covid-19, of which 3,992 succumebd to the virus, as per official figures. However, these figures have taken into account only deaths of patients who did not have any other ailment and deaths recorded in hospitals. Authorities say, deaths in patients having other complications were not labelled as Covid deaths, making it tough for their kin to seek compensation.

Elaborating on other situations where the kin would not have required certificates, an official from the health department, said, “several patients who died on the way to hospitals, in ambulances, during the first and second wave might have gone unrecorded. Many died in remote villages with or without treatment. Some underwent Covid tests, but there were no death certificates. Others do not have reports of supportive tests like CT Scan, RTPCR etc.”

“Besides this, in case of hospitalised patients, some families did not even take the bodies home during the first wave due to the associated stigma. The government itself conducted the cremation. Such family members have no evidence that the patient died due to Covid-19,” added the official.

Despite the fact that the state has formed a district-level ‘Covid-19 death ascertaining’ committee (CDAC) a fortnight ago, many are unable to get the necessary documents. “There are many people turning up saying that the patient had tested positive, but had died at home while undergoing treatment or had suddenly deteriorated and died. As a result, the families do not have the required reports. Their death certificates do not mention Covid as cause of death and they are now unable to apply for compensation,” said a meeseva employee from Ameenpur area.

Jr docs to boycott outpatient services

Jr docs to boycott outpatient services

TIMES NEWS NETWORK

Hyderabad:01.12.2021

The Telangana Junior Doctors Association (TJUDA) will boycott outpatient services and hold protests from December 1 to December 3 over repeated delays in NEET counselling.

The doctors have maintained the delay has not only caused loss of pay for 1.6 lakh doctors across the country, but has also put and additional burden on the remaining batches. Moreover, in view of a possible third wave keeping an entire batch of the healthcare system could have devastating impact.

“T-JUDA stands in accordance with consensus decision of Federation of All India Medical Association (FAIMA) and other states resident doctors associations (RDAs) to hold protest and boycott OPD services from December 1 to December 3 regarding expedition of NEET PG-21 counselling process, provided the Supreme Court hearing happens by the stipulated date of 03-12-2021, beyond which we will have to escalate the protest by boycotting all elective services throughout the state,” the TJUDA said in an official statement. The association said repercussions of this will have to be borne by the government.

“Our primary demand is to take necessary measures for expediting NEET-PG counselling as well as admission process and to fast-track the court proceedings on an urgent basis by Union Government and Supreme court of India,” said Dr D Sagar, president TJUDA. “With the possibility of an imminent upcoming Covid wave, it is essential that the counselling process is started at the earliest to prevent collapse of the healthcare system...”

Kerala girl found dead in US with bullet wounds

Kerala girl found dead in US with bullet wounds

01.12.2021

A 19-year-old girl from Kerala was found dead with bullet injuries at her apartment in Montgomery, the capital of US state of Alabama, on Monday. Mariyam Soosan Mathew, a native of Niranam near Thiruvalla, was found dead in her room, as per information received by her relatives in India. According to the metropolitan of the Ahmedabad diocese Geevarghese Mar Yulios, who spoke to Mariyam’s father, a bullet fired from the apartment on the next floor had pierced through the ceiling and hit Mariyam. He added that celebrations were being held in the apartment above as part of the Thanksgiving weekend. Mariyam’s father, Boban Mathew, was a council member of the Ahmedabad diocese of the Malankara Orthodox Syrian Church while the family was in Muscat. They had relocated to Montgomery only few months ago after Mariyam’s mother, who is a nurse, got a job there. The family is trying to take the mortal remains to Kerala for the funeral. TNN

Prepare for common entrance for UG, PG seats from 2022-23: UGC

NTA UMBRELLA

Prepare for common entrance for UG, PG seats from 2022-23: UGC

Manash.Gohain@timesgroup.com

New Delhi:01.12.2021

The University Grants Commission (UGC) has written to central universities to “take appropriate measures” for conducting the Common Entrance Test (CET) for admission in undergraduate and postgraduate courses from the 2022-23 academic session onwards. It also stated that willing state and private universities too can adopt this computer-based test, which will be conducted by the National Testing Agency (NTA) in 13 languages.

TOI was first to report that the education ministry was finalising the procedural details of CET to be conducted in 45 central universities from 2022 onwards and that the details would be announced in December.

The UGC has also said that for PhD admissions, NET scores shall be considered, wherever feasible.

The letter sent to the vice chancellors of the central universities stated, “After detailed deliberations, it was resolved that the CET for UG and PG courses may be conducted from the 2022-23 academic session through NTA.” The plan was put on hold this year due the ongoing Covid pandemic.

The exams, envisaged in the National Education Policy 2020, are likely to be conducted twice a year. According to the NEP 2020 document: “The NTA will facilitate a single entrance exam for admissions to universities across the country. It will offer a ‘high quality common aptitude test’ like the SAT (Scholastic Aptitude Test, conducted in US & Canada), as well as specialised common subject examinations, at least twice every year.”

The CET for admissions to central universities, which includes the likes of JNU, BHU and DU, would be a three-hour test, divided into two sections — common aptitude test (50 questions) and domain specific tests (30 questions each) — and will eliminate the individual exams that many of the universities conduct presently.The UGC in its letter said: “Accordingly, all central universities are advised to take appropriate measures for the Common Entrance Test from the academic session 2022-23.”

Unarmed, ‘tiger’ mom snatches kid from leopard jaws

Unarmed, ‘tiger’ mom snatches kid from leopard jaws

TIMES NEWS NETWORK

Bhopal: 01.12.2021

A tribal woman took on a leopard with her bare hands and snatched her six year-old son from its jaws in a bloody fight in a village near Sanjay Gandhi National Park in MP’s Sidhi on Sunday.

The Baiga woman chased the leopard and caught up with it just when it sat down with its ‘prey’, say sources. She came out of the fight injured, but with the child in her arms. The boy also has deep claw and fang wounds, but scars — and the memory of his unarmed mother wrestling a leopard — are all that will remain.

Kiran, the feisty mother, lives in Badi Jhiriya village in the buffer zone of the national park. On Sunday evening, she was sitting with her children next to a fire outside her hut, waiting for her husband to return, unaware that a leopard was watching them. Her youngest child, a few months old, was in her lap.

In a flash, the leopard darted out of the shadows, caught six-year-old Rahul in its jaws, and ran off. Kiran was up in a flash, too. She handed the newborn to one of her other kids and sprinted after the animal. Even in the darkness, she kept up with the spotted cat and found it sitting in some bushes. Kiran says she lunged at the leopard, grabbed Rahul and pulled with all her strength. The predator seemed taken by surprise, and she could tear Rahul from its grasp.

Unwilling to give up its prey, the leopard lashed out with its claws and cut mother and child. Kiran fought back, screaming for help. By then, villagers were already running to her aid. The pounding feet and the shouts unnerved the leopard, which scampered back into the forest. The villagers took mother and child her to hospital.

Kiran with her son in Badi Jhiriya village in Madhya Pradesh