Microsoft promotes cloud momentum from Y Combinator startups

EzDubs, a developer of language translation technology, started out the way many tech startups do. Launched into the public clouds by Amazon and Google.

However, after EzDubs went through the Y Combinator startup program last year, the company quickly pivoted, adding of Microsoft cloud in the mix. This is because the founders of EzDubs found out about one partnership that allowed Y Combinator companies to receive $350,000 worth of Microsoft Azure credits.

It was a “godsend,” EzDubs co-founder Padmanabhan Krishnamurthy told CNBC. The credits were especially helpful, Krishnamurthy said, because Microsoft has been at the forefront of the artificial intelligence boom, investing in OpenAI and hosting many projects that use the company’s large language models (LLM).

On Azure, EzDubs was able to access the advanced graphics processing units (GPUs) required for a new wave of AI model training that no other cloud provider could match.

“At that point, it was a no-brainer,” said Krishnamurthy, who founded his company in 2022 as genetic AI was taking off. “It was the exact setup we needed,” with GPU availability that “literally no one else had.”

The story of EzDubs can be heard in various forms from startups across the AI ​​landscape. While Amazon Web Services maintains the market lead in cloud infrastructure and Google remains a popular choice for companies using multiple clouds, Microsoft’s perceived strength in artificial intelligence gives the company an edge, at least as far as startups are concerned.

In Amazon’s second-quarter earnings report Thursday, the company said AWS revenue rose 19% from a year earlier, trailing Microsoft’s 29% growth for the latest period, though that includes other services as well. cloud other than Azure.

AWS was the first of the cloud providers to give credit to new companies, hoping they would stick around until the credits ran out and eventually turn into big spenders. AWS’ Program activation launched in 2013, following the launch of its core EC2 (compute) and S3 (storage) services in 2006, it helped strengthen Amazon’s dominance of the public cloud.

Microsoft’s access to heavy GPU clusters combined with its long history as an enterprise technology company ubiquitous in IT departments changes the narrative. And, of course, money matters.

In November, Microsoft formed a partnership with Y Combinator — known for its contribution to replication Dropbox, Airbnb, Stripe and other companies — which gave $350,000 in loans to startups entering the accelerator. Startups in some other programs, such as the Alchemist Accelerator and Alt Capital’s Generate, are also eligible.

Amazon followed in April, announces $500,000 in credits to Y Combinator companies, including $200,000 in cloud credits and $300,000 in proof-of-concept credits using the cloud provider’s Trainium and Inferentia chips for AI, an AWS spokesperson said in an email. The current bid includes $350,000 in AWS credits, plus $300,000 earmarked for use of the custom silicon, the spokesperson said.

Annie Pearl, Microsoft’s corporate vice president, told CNBC that before the Y Combinator partnership, only about 5 percent of companies in the program were building on Azure. By May, more than 50% were using Azure, he said. A spokesperson later said that 58% of Y Combinator startups had accepted Microsoft’s credit offer, a figure that did not reflect actual Azure usage.

AWS said it sees a different dynamic play.

“That claim just doesn’t ring true to us,” an AWS spokesperson said in an email, referring to Pearl’s statement that more than half of Y Combinator startups were using Azure. “In their early stages, startups may accept advertising credits from different cloud providers, but when they mature and have to make a decision about who to trust with the future of their organization, they largely turn to the provider with the best security, reliability and scalability’.

Amazon said in April suspension that over 80% of startups in Y Combinator’s 2022 and 2023 batches were running on AWS.

Microsoft and Amazon are competing for startups far beyond accelerator programs. Last month, AWS doubled to $200,000 the maximum amount of credit a startup can use if it raised a Series A round of funding in the past year; CNBC reported. In the Microsoft for Startups Founders Hub program, companies can receive $150,000 in Azure credits.

Looking at the overall cloud market, industry data shows that Microsoft has narrowed Amazon’s lead. AWS’s share in the first quarter of this year was 31%, and Azure was in second place with 25%, according to research firm Canalys. Three years earlier, AWS controlled 32% of the market, while Microsoft was at 19%, Canalys is appreciated.

Microsoft CEO Satya Nadella said on an earnings call in October that more emerging businesses are turning to Azure because of demand for OpenAI’s models.

“We’re expanding our reach with companies that come from digital,” he said. “Leading AI startups are using OpenAI to power their AI solutions, thus making them Azure customers as well.”

InKeep, whose technology allows companies to search internal documents using chatbots, chose to use Azure while participating in Y Combinator in early 2023, shortly after OpenAI launched ChatGPT. OpenAI’s LLM subjects were not available on other clouds.

“Especially when I started, OpenAI had state-of-the-art models,” Nick Gomez, co-founder and CEO of InKeep, said in an interview. InKeep also started using of Google Cloud Platform for specific workloads.

Gomez said Azure has less downtime than other clouds and is fast even when dealing with compute-intensive AI models. He said data privacy is very important to customers when it comes to AI training. OpenAI initially trained models with customer data, but later discontinued the practice, CEO Sam Altman told CNBC’s Andrew Ross Sorkin last year.

“People were asking all the time, ‘Are you training with our data?’ Gomez said. “Being able to say, ‘Hey, no, we don’t, we use Azure, they don’t maintain it, they don’t train on it,’ something like that, definitely put a lot of people at ease.”

Cloud infrastructure has proven not to be a winner-takes-all market. Amazon, Microsoft and Google are steadily increasing their revenue in a business that Canalys is waiting to expand by 20% this year to nearly $350 billion.

This is partly because large companies are increasingly using multiple clouds to ensure they are not overly dependent on a single vendor and to take advantage of the different services and technologies from different providers. For startups that rely on venture funding to fuel their operations, accepting credit from multiple vendors allows them to control their expenses, which is especially important given the high cost of running AI workloads.

Accepting credits is “almost like raising money,” said Prady Modukuru, co-founder and CEO of Sync Labs, a developer of lip-synching technology.

“Nobody can spend $20,000 to $30,000 a month on infrastructure costs,” said Modukuru, a former Microsoft product manager.

Modukuru said Sync Labs has used Amazon, Google and Microsoft, but started with Azure earlier this year while at Y Combinator. It’s the only place the company could find GPUs, he said.

“We would just request and within an hour and a half we would have access to them in Azure,” Modukuru said. “That’s what we needed as a startup.”

Earlier this year, Sync Labs learned how to run high-performance code on multiple GPUs by talking to Microsoft engineers during office hours, Modukuru said. AWS also has its experts at Y Combinator’s founders, a spokesperson said.

AWS has other ways to counter Microsoft and its close partnership with OpenAI. For example, Amazon poured billions of dollars into Anthropic, which develops its own LLMs. Anthropic has released a model that is at least as good as OpenAI’s GPT-4, said Daksh Gupta, CEO of Greptile, a startup that helps developers work with source code.

Because Anthropic’s model is available on AWS, Greptile plans to end its use of the Azure OpenAI service and switch to AWS’s competing Bedrock tool, Gupta said.

“For the quality of the experience, it doesn’t make sense to pinch pennies on it,” he said. “We spend what we need to spend.”

However, OpenAI gives Microsoft a big lead in AI and forces AWS into the unfamiliar position of trying to play catch-up. Kareem Nassar, who co-founded EzDubs with Krishnamurthy, said OpenAI’s rapid market penetration helped Microsoft address complex AI infrastructure maintenance issues.

“I know it’s battle-tested,” Nassar said. “I wasn’t hitting huge bugs. You could just tell he’s got some mileage on him.”