CIA-funded Cerebra Wafer Scale Engine AI chip


Hello and welcome to Protocol Enterprise! Today: why a new chip from Cerebras has a lot of exciting potential, what developers really think about the tools of their trade, and how Mandiant expects to evolve under Google Cloud.

CIA wants cheap AI training

AI training is difficult and expensive enough that the CIA found a way to make it cheaper.

Through its non-profit investment operation called In-Q-Tel – which is also funded by several other members of the US intelligence community – the agency has bet that a Silicon Valley startup called Cerebras can make a plate-sized chip that can dramatically reduce the cost of AI model training.

  • AI models are methods of organizing mathematical calculations by breaking them down into steps and then defining rules for how the steps talk to each other.
  • But to improve AI models, they get bigger, and then, of course, more expensive to train because they need a lot more chips to do it effectively.

Cerebras decided to change that. On Wednesday, it announced it had hit a major road marker on the trip: CEO Andrew Feldman said the company’s engineers had successfully trained a 20 billion-parameter AI model on its size of a plate.

  • Installing an AI model on a single chip offers distinct performance benefits for training models that can reduce the time needed from weeks to days, which also makes it cheaper.
  • Cerebras calls superchips the Wafer Scale Engine, and the idea behind fitting a model of 20 billion parameters onto a single one is to make the cost of training AI cheap enough that anyone can do it.
  • The cheapest option seems to be $25,000 a week for a Cerebras system powered by the WSE-2, although customers can rent it for a few hours from a cloud provider if that’s what they need. need.

Cerebras is backed by the CIA – sort of. Feldman told Protocol that In-Q-Tel is one of the company’s backers, but likely received the bulk of regular VC funding.

  • In-Q-Tel invests on behalf of several spy agencies, but usually doesn’t drop a ton of money on every bet.
  • Cerebras has raised $735 million, putting it at a valuation of $4.1 billion.
  • VCs became interested in AI chip startups when it became clear that general-purpose computing offered by Nvidia, AMD, and Intel might need help with specialized chips.

Read the full story here.

— Max A. Cherney (email | Twitter)


The competitive advantage of digital solutions

For the past 50 years, SAP has worked closely with our customers to solve some of the world’s most complex problems. We have also seen and participated in rapid accelerations of technology in response. Across all industries, certain pathways have emerged to help companies navigate the unexpected challenges of recent years.

Click here to learn more about SAP

The developers speak

From programming languages ​​to cloud platforms, one of the best places to track enterprise technology trends is the annual report. Stack Overflow Developer Surveywhich received more than 70,000 responses this year.

As it has been for nine years, the most commonly used programming language is still JavaScript. The decades-old language managed to beat its competitors by a significant margin: 65% versus 55% for HTML/CSS and just over 48% for Python and SQL. For new developers, however, HTML/CSS and Python are just as popular as JavaScript.

Overall, Python and TypeScript are two programming languages ​​that seem to be making serious inroads with developers. Python and TypeScript are two of the most common languages ​​that developers want to use, if they aren’t already. Rust is still the most popular among developers, with 86% of developers planning to continue using the language.

In the world of cloud development, AWS is the most widely used platform, with 51% of developers doing significant work on it. Microsoft Azure and Google Cloud are constantly battling for second place in the hearts and minds of developers, and this year Azure came out on top with 29% of developers relying on the platform. For new developers, however, the story was different: Google Cloud edged out AWS and Azure, but still lost the top spot to Heroku.

It’s a key trend revealed by the survey: although legacy programming languages ​​and cloud platforms have a hold on professional developers, other vendors still have room to gain new developers.

— Aisha counts (E-mail | Twitter)

Mandiant’s Mandate

Google’s $5.4 billion deal to acquire cybersecurity powerhouse Mandiant could shake up the industry. At the RSA conference this month in San Francisco, I sat down with Mandiant’s chief technical officer, Marshall Heilman, to discuss what the merger – which is expected to close later in 2022 – might change for the company and for Google Cloud.

Assuming the acquisition is completed, how will Google Cloud customers benefit?

When we sell an integrated solution, with someone else’s technology [plus] our intelligence, it doesn’t matter whether it’s a Google-specific customer or a Microsoft-specific customer. They are [all] benefiting from it. Now, Google customers might benefit a little more over time as we’ll natively gain more experience in GCP. [Google Cloud Platform].

What is important and what will change for both parties?

I think we’re going to try to take the best of the Google culture and what made them successful, and try to bring that to Mandiant – while taking Mandiant’s approach to how we run businesses and which we work in the security world, and bring that to Google. What I hope, personally, is that we mix the two environments.

What other benefits could Mandiant get from joining Google?

If you look at our information, Google has a lot more data than we do. It would be fantastic [to get access to]. I’m sure there’s things we can learn from [Google’s] engineering. I think from a consulting perspective, I know that Google doesn’t really have consulting services like we do, in the security aspects. It is therefore an expertise that we can bring to them.

—Kyle Alspach (E-mail | Twitter)


The competitive advantage of digital solutions

When companies invest in maintaining their “green ledger” with the same commitment they have to their financial ledgers, they will be able to connect their environmental, social and financial data holistically so they can steer their business towards sustainability. Ultimately, what gets measured gets managed.

Click here to learn more about SAP

Thanks for reading – see you tomorrow!

Source link


Comments are closed.