Tuesday 29th March 2016
Self-Driving Car Startup Fights to Beat Tesla and Google
We want to ship a product by the end of the year that people will be able to install in their own cars and it will give them more self-driving capability than the Tesla today. – George Hotz
George Hotz’s pitch is that he can build self-driving car algorithms faster and better than any carmaker or even Google.
“Google is going to ship by the end of 2020? We’re actually making this stuff work,” said Hotz, who’s wearing jeans and a black hoodie with a large white comma on the front for his new company, Comma.ai.
Since he revealed his ambitions in a Bloomberg Businessweek article published last December, Hotz has attracted plenty of attention. The CEOs of Delphi, a major auto parts supplier, and Nvidia, maker of graphics processing units, have paid visits to his basement office at the “Crypto Castle,” a three-story house located in San Francisco’s Potrero Hill neighborhood and occupied by some of the city’s Bitcoin entrepreneurs.
He’s generated enough excitement to score an unannounced seed investment from venture capital firm Andreessen Horowitz that values Hotz’s tiny, fledgling company at $20 million, according to sources.
Hotz began Comma last October and he’s well past the lone-hacker-in-the-basement stage. Yunus Saatchi, who has a PhD from the University of Cambridge in artificial intelligence, has joined as chief machine learning officer. Saatchi was a colleague of Hotz’s at Vicarious, a San Francisco-based AI startup with $72 million in financing from investors like Musk and Amazon’s Jeff Bezos.
Jake Smith, a roommate of Hotz’s in the Crypto Castle who is involved in the Bitcoin community, is head of operations. And Elizabeth Stark, another prominent fixture in the Bitcoin startup world, is Comma’s legal advisor. (They’re all wearing Comma.ai shirts when I meet them.) Hotz plans to hire around eight people total in the coming three months. He’s looking for people in machine learning and consumer hardware.
Hotz is also starting work on what will become the company’s first product — a self-driving kit that car owners will be able to purchase directly from Comma to equip their vehicles with autonomous driving capabilities. He hasn’t come close to working out the details of what this product will ultimately look like, but he said it might be a dash cam that plugs into the on-board diagnostics 2 port, which gives access to the car’s internal systems and is found in most cars made after 1996. It will provide cars with ADAS features, like lane-keeping assistance and emergency breaking.
“We believe our killer app is traffic,” Hotz said. “Humans are bad at traffic. We can make something that drives super-humanly smooth through traffic.”
Hotz said he won’t be able to turn every car into a semi-autonomous vehicle. At a minimum, the car will have to have anti-locking brakes and power steering. He’s hoping Comma’s product will work most with the five top-selling cars in the United States. – Aaron Tilley
Sunday 24th April 2016
A $2 Billion Chip to Accelerate Artificial Intelligence
Two years ago we were talking to 100 companies interested in using deep learning. This year we’re supporting 3,500. In two years there has been 35X growth. – Jen-Hsun Huang, CEO of Nvidia
The field of artificial intelligence has experienced a striking spurt of progress in recent years, with software becoming much better at understanding images, speech, and new tasks such as how to play games. Now the company whose hardware has underpinned much of that progress has created a chip to keep it going.
Nvidia announced a new chip called the Tesla P100 that’s designed to put more power behind a technique called deep learning. This technique has produced recent major advances such as the Google software AlphaGo that defeated the world’s top Go player last month.
Deep learning involves passing data through large collections of crudely simulated neurons. The P100 could help deliver more breakthroughs by making it possible for computer scientists to feed more data to their artificial neural networks or to create larger collections of virtual neurons.
Artificial neural networks have been around for decades, but deep learning only became relevant in the last five years, after researchers figured out that chips originally designed to handle video-game graphics made the technique much more powerful. Graphics processors remain crucial for deep learning, but Nvidia CEO Jen-Hsun Huang says that it is now time to make chips customized for this use case.
At a company event in San Jose, he said, “For the first time we designed a [graphics-processing] architecture dedicated to accelerating AI and to accelerating deep learning.” Nvidia spent more than $2 billion on R&D to produce the new chip, said Huang.
It has a total of 15 billion transistors, roughly three times as many as Nvidia’s previous chips. Huang said an artificial neural network powered by the new chip could learn from incoming data 12 times as fast as was possible using Nvidia’s previous best chip.
Deep-learning researchers from Facebook, Microsoft, and other companies that Nvidia granted early access to the new chip said they expect it to accelerate their progress by allowing them to work with larger collections of neurons.
“I think we’re going to be able to go quite a bit larger than we have been able to in the past, like 30 times bigger,” said Bryan Catanzero, who works on deep learning at the Chinese search company Baidu. Increasing the size of neural networks has previously enabled major jumps in the smartness of software. For example, last year Microsoft managed to make software that beats humans at recognizing objects in photos by creating a much larger neural network.
Huang of Nvidia said that the new chip is already in production and that he expects cloud-computing companies to start using it this year. IBM, Dell, and HP are expected to sell it inside servers starting next year. – Tom Simonite
Monday 23rd May 2016
Nvidia Smashes Q1 Expectations On ‘Sweeping’ AI Adoption
Nvidia (NVDA) rocketed after the maker of graphics chips beat Q1 sales expectations and topped earnings views, led by faster adoption of artificial intelligence technology that utilizes Nvidia graphics chips.
CEO Jen-Hsun Huang credited accelerated growth of deep-learning, or AI, technology for the Q1 beat.
“Accelerating our growth is deep learning, a new computing model that uses the GPU’s (graphics processing unit) massive computing power to learn artificial intelligence algorithms,” he said in the company’s earnings release. “Its adoption is sweeping one industry after another, driving demand for our GPUs.”
Nvidia’s soon-to-be-released Pascal chip will continue that drive, he said.
“Our new Pascal GPU (graphics processing unit) architecture will give a giant boost to deep learning, gaming and VR (virtual reality),” he said. “Pascal processors are in full production and will be available later this month.” – Allison Gatlin
Monday 23rd May 2016
Nvidia Bringing Artificial Intelligence to America’s Best Hospital
In order to advance healthcare by applying the latest artificial intelligence techniques to improve the detection, diagnosis, treatment and management of diseases, NVIDIA announced that it is a founding technology partner of the Massachusetts General Hospital (MGH) Clinical Data Science Center.
MGH — which conducts the largest hospital-based research program in the United States, and is the top-ranked hospital on this year’s US News and World Report “Best Hospitals” list — recently established the MGH Clinical Data Science Center in Boston.
The center will train a deep neural network using Mass General’s vast stores of phenotypic, genetics and imaging data. The hospital has a database containing some 10 billion medical images.
To process this massive amount of data, the center will deploy the NVIDIA DGX-1 — a server designed for AI applications, launched recently at the GPU Technology Conference — and deep learning algorithms created by NVIDIA engineers and Mass General data scientists.
“Deep learning is revolutionizing a wide range of scientific fields,” said Jen-Hsun Huang, CEO and co-founder, NVIDIA. “There could be no more important application of this new capability than improving patient care. This work will one day benefit millions of people by extending the capabilities of physicians with an incredibly powerful new tool.”
Using AI, physicians can compare a patient’s symptoms, tests and history with insight from a vast population of other patients. Initially, the MGH Clinical Data Science Center will focus on the fields of radiology and pathology — which are particularly rich in images and data — and then expand into genomics and electronic health records. – Nvidia
Thursday 18th August 2016
The AI Gold Rush
Companies are lining up to supply shovels to participants in the AI gold rush.
The name that comes up most frequently is NVIDIA (NASDAQ: NVDA), says Chris Dixon of Andreessen Horowitz; every AI startup seems to be using its GPU chips to train neural networks.
IBM (NYSE: IBM) and Google, meanwhile, are devising new chips specifically built to run AI software more quickly and efficiently.
And Google, Microsoft and IBM are making AI services such as speech recognition, sentence parsing and image analysis freely available online, allowing startups to combine such building blocks to form new AI products and services.
More than 300 companies from a range of industries have already built AI-powered apps using IBM’s Watson platform, says Guru Banavar of IBM, doing everything from filtering job candidates to picking wines. – The Economist
Friday 30th September 2016
Chris Dixon: In 2 years Everyone Will Use Driverless Cars on Highways
Within ten years, roads will be full of driverless cars.
Maybe within two, depending on where you’re driving.
That’s what Chris Dixon, a partner at prestigious Silicon Valley investment firm Andreessen Horowitz believes.
Dixon has written extensively about the future of autonomous vehicles and invested in a number of startups in the space, from self-flying delivery drones to, a company founded by a young man who built a self-driving car in his garage.
“All of the trends we’ve been observing over the last decade — from cloud computing to cheaper processing — have hit a tipping point,” Dixon says. “This is the core that’s getting people excited about AI, and specifically around autonomous vehicles and autonomous cars.”
It’s also cheaper than ever to build a smart car. Dixon says many driverless car companies use tiny chips made by a publicly-traded company, NVIDIA. NVIDIA’s chips only cost a couple hundred dollars.
“For $200, you could get what 10 years ago was a supercomputer on a little board and put it in your car, and it can run one of these sophisticated deep learning systems,” he says.
Additionally, a lot of the AI for autonomous vehicles is open-sourced, like Google’s product TensorFlow. This allows everyone in the space to create more accurate technology faster, because they can learn from each other’s data sets and build off the findings.
” I bet in two years, it will be the norm that on the highway, you’re not driving half the time or you’ll be using driver assistants heavily,” he says.
“It’s easier on highways and in suburbs,” says Dixon. “So you can imagine pushing a button on your Uber or Lyft app, and depending on the situation and location, an autonomous car comes or a person comes.”
He adds, “When will an Uber roll up without a person in it in New York City? That’s farther away. But I think that’s more like five years away, not 20.”
Dixon likens the promise of self-driving to Henry Ford’s Model T, which was like the iPhone of the time — a real technology game changer. At first, consumer cars seemed impossible — roads weren’t paved and no one knew how to drive cars. But the product was a hit, and everything changed to make way for them. – Alyson Shontell
Wednesday 30th November 2016
The world realizing Nvidia’s potential in AI. – Stephen Cole
Wednesday 30th November 2016
Moore’s Law is Now in NVIDIA’s Hands
- We’ll soon see the power of computing increase way more than any other period. There will be trillions of products with tiny neural networks inside. – Jen-Hsun Huang, CEO of NVIDIA
Intel has long ago ceded leadership for Moore’s Law. And so, understandably, they have trumpeted the end of Moore’s Law for many years. To me, it sounds a lot like Larry Ellison’s OpEd declaring the end of innovation in enterprise software, just before cloud computing and SaaS took off. In both cases, the giants missed the organic innovation bubbling up all around them.
For the past seven years, it has not been Intel but NVIDIA that has pushed the frontier of Moore’s processor performance/price curve.
For a 2016 data point, consider the NVIDIA Titan GTX. It offers 10^13 FLOPS per $1K (11 trillion calculations per second for $1,200 list price), and is the workhorse for deep learning and scientific supercomputing today. And they are sampling much more powerful systems that should be shipping soon. The fine-grained parallel compute architecture of a GPU maps better to the needs of deep learning than a CPU.
There is a poetic beauty to the computational similarity of a processor optimized for graphics processing and the computational needs of a sensory cortex, as commonly seen in neural networks today.
I was going to update the Kurzweil Curve (the meaningful version of Moore’s Law) to include the latest data points, and found that he was doing the same thing.
Here is the preliminary version. The 7 most recent data points are all NVIDIA, with CPU architectures dominating the prior 30 years:
It’s what gives us hope for the future and I think it’s the most important thing ever graphed.
Here is the prior version
Moore’s Law is now in NVIDIA’s hands. Consider the GTX Titan X for 2016. 11 TFLOPS for $1,200. That would be on the order of 10^13 for the far right side of the graph, perfectly on the line.
Wednesday 30th November 2016
Building an AI Portfolio
The following stocks offer exposure to Artificial Intelligence. – Lee Banfield
Google (NASDAQ: GOOGL)
Stock Price: $776
Market Cap: $531 billion
Healthcare Images – Google Deepmind
Machine Learning – GoogleML
Autonomous Systems – Google Self-driving Car
Hardware – GoogleTPU
Open Source Library – TensorFlow
IBM (NYSE: IBM)
Stock Price: $162
Market Cap: $154 billion
Enterprise Intelligence – IBM Watson
Healthcare – IBM Watson Health
Amazon (NASDAQ: AMZN)
Stock Price: $752
Market Cap: $355 billion
Personal Assistant – Amazon Alexa
Open Source Library – DSSTNE
Microsoft (NASDAQ: MSFT)
Stock Price: $60
Market Cap: $473 billion
Personal Assistant – Cortana
Open Source Libraries – CNTK, AzureML, DMTK
Nvidia (NASDAQ: NVDA)
Stock Price: $94
Market Cap: $50 billion
Stock Price: $1,250
Market Cap: $176 billion
Personal Assistant – Viv
Qualcomm (NASDAQ: QCOM)
Stock Price: $68
Market Cap: $100 billion
Tesla (NASDAQ: TSLA)
Stock Price: $188
Market Cap: $29 billion
Illumina (NASDAQ: ILMN)
Stock Price: $133
Market Cap: $20 billion
Healthcare, Cancer Detection – Grail
Mobileye (NYSE: MBLY)
Stock Price: $37
Market Cap: $8 billion
Wednesday 14th December 2016
Full Self-Driving Hardware Becoming Available on All Tesla Cars
This is huge news. It was just a few years ago that the sensors/cameras used on the Google cars were over $100,000 to achieve level 3 autonomy.
To have the hardware component installed on all Tesla cars (including the $35k Model 3) moving forward happened years ahead of when I feel most of us that follow autonomous vehicle tech would have imagined. From a tech perspective, this is mind-blowing news. – Nathan Wright
Musk announced that all Tesla cars being produced as of today, including the Model 3, will have everything they need onboard to achieve full Level 5 self-driving in the future.
The biggest change might be the new onboard computer that provides over 40 times the processing power of the existing Tesla hardware, which actually runs the in-house neural net the car maker has developed in order to handle processing of data inbound from the vision, sonar and radar systems.
Musk said on call discussing the most recent update to the existing driver assistance Autopilot software that it basically stretched computing power to the limit, which is why the upgraded CPU is required for full Level 5 autonomy. The new GPU is the Nvidia Titan, Musk said on the call, though it was a “tight call” between Nvidia and AMD.
The validation required for full autonomy will still take some more time, but Musk said on a call that it’s actually already looking like it’ll be at least two times as safe as human driving based on existing testing. – Darrell Etherington
Saturday 24th December 2016
Marc Andreessen: If We Were a Hedge Fund We’d Put All Our Money Into Nvidia
Nvidia’s dominance of the GPU sector–it has more than a 70% share–and its expansion into new markets have sent its stock soaring. Its shares are up almost 200% in the past 12 months
There are an estimated 3,000 AI startups worldwide, and many of them are building on Nvidia’s platform. They’re using Nvidia’s GPUs to put AI into apps for trading stocks, shopping online and navigating drones.
“We’ve been investing in a lot of startups applying deep learning to many areas, and every single one effectively comes in building on Nvidia’s platform,” says Marc Andreessen of venture capital firm Andreessen Horowitz. “It’s like when people were all building on Windows in the ’90s or all building on the iPhone in the late 2000s.
“For fun,” adds Andreessen, “our firm has an internal game of what public companies we’d invest in if we were a hedge fund. We’d put all our money into Nvidia.” – Aaron Tilley
Saturday 31st December 2016
The World Realizing Nvidia’s Potential