Elon Musk is on a mission to build new supercomputers. As the CEO of Tesla and his new artificial intelligence startup xAI, the tech titan has big plans for how artificial intelligence can help to supercharge his businesses.

In January, he wrote on X that Tesla should be viewed as an AI/robotics company rather than a car company. Tesla’s custom-built supercomputer named Dojo is key to this transformation. Tesla has said it plans to spend $500 million to build the supercomputer in Buffalo, New York. Tesla is also building another supercomputer cluster, called Cortex, at the company’s headquarters in Austin, Texas.

Dojo will process and train AI models using the large amounts of video and data captured by Tesla cars. The goal is to improve Tesla’s suite of driver assistance features, which the company calls Autopilot, and its more robust Full Self-Driving or FSD system. Subscriptions to Tesla’s FSD features cost $99 a month and include automatic lane changes, automatic parking and automatic stopping for traffic lights and stop signs.

“They’ve sold what is it, 5 million plus cars. Each one of those cars typically has eight cameras plus in it. And if you think then that those cars are driving around, let’s just say 10,000 miles a year on average, they’re streaming all of that video back to Tesla,” says Steven Dickens, chief technology advisor at the Futurum Group. “So what can they do with that training set? Obviously they can develop Full Self-Driving and they’re getting close to that.”

Despite their names, neither Autopilot nor FSD make Tesla vehicles autonomous and require active driver supervision, as Tesla states on its website. In the past, the company has garnered scrutiny from regulators who say that Tesla falsely advertised the capabilities of its Autopilot and FSD systems. But reaching full autonomy is critical for Tesla, whose sky-high valuation is largely dependent on bringing robotaxis to market, some analysts say.

The company reported lackluster results in its latest earnings report and has fallen behind other automakers working on autonomous vehicle technology. These include Alphabet-owned Waymo, which is already commercially operating fully autonomous taxis in several U.S. cities, GM’s Cruise and Amazon’s Zoox. In China, competitors include Didi and Baidu.

Tesla hopes Dojo, which Musk says has been running tasks for Tesla since 2023, will change that. A Tesla robotaxi event originally scheduled for August is now expected to occur in early October.

Dojo can also be useful for training Tesla’s humanoid robot, Optimus, which the company plans to use in its factories starting next year. Musk has said that Tesla plans to spend $10 billion this year on AI.

Musk is also betting on supercomputers to run his new AI venture xAI. Musk launched xAI in 2023 to develop large language models and AI products, like its chatbot Grok, as an alternative to AI tools created by OpenAI, Microsoft and Google.

Despite being one of its founders, Elon Musk left OpenAI in 2018 and has since become one of the company’s harshest critics. In June, it was announced that xAI would build a supercomputer in Memphis, Tennessee to train Grok. In early September, Musk revealed that a portion of the Memphis supercomputer, called Colossus, was already online.

To learn more about Elon Musk’s supercomputer plans, watch the video.

Alibaba on Thursday released more than 100 open-source artificial intelligence models and boosted the capabilities of its proprietary technology as it looks to ramp up competition with rivals.

The newly-released models, known as Qwen 2.5, are designed for use in applications and sectors ranging from automobiles to gaming and science research, Alibaba said. They have more advanced capabilities in math and coding, it added.

The Hangzhou-headquartered firm is looking to increase competition with domestic rivals such as Baidu and Huawei, as well as U.S. titans like Microsoft and OpenAI.

AI models are trained on huge amounts of data. Alibaba says its models have the ability to understand prompts and generate texts and images.

Open-source means that anyone — including researchers, academics and companies — around the world can use the models to create their own generative AI apps without needing to train their own systems, saving time and expense. By open sourcing the models, Alibaba hopes more users will use its AI.

The Chinese e-commerce giant first launched its Tongyi Qianwen, or Qwen, model last year. Since then, it has released improved versions and says that, to date, its open source models have been downloaded 40 million times.

The company also said that it upgraded its proprietary flagship model called Qwen-Max, which is not open-source. Instead, Alibaba sells its capabilities through its cloud computing products to businesses. Alibaba said that Qwen Max 2.5-Max surpassed rivals such as Meta‘s Llama and OpenAI’s GPT4 in several areas including reasoning and language comprehension.

Alibaba also launched a new text-to-video tool based on its AI models. This allows users to input a prompt and the AI will create a video based on it. This is similar to OpenAI’s Sora.

“Alibaba Cloud is investing, with unprecedented intensity, in the research and development of AI technology and the building of its global infrastructure,” Eddie Wu, CEO of Alibaba, said in a statement.

Wu, who took over the role of CEO at Alibaba last year amid a historic reshuffle, has been trying to reinvigorate growth at the tech giant, as it faces headwinds including rising competition and a sluggish Chinese consumer.

Alibaba is one of the biggest cloud computing players in China, but internationally, it trails the likes of Amazon and Microsoft. The company is hoping that its latest AI offerings may tempt customers inside and outside of China to sign up to its cloud services, boosting a division which has been sluggish but showed early sign of an acceleration in the June quarter.

Chinese companies are ramping up efforts to produce a viable alternative to Nvidia’s chips that power artificial intelligence as Beijing continues its efforts to wean itself off American technology.

U.S. sanctions slapped on China over the past few years, along with Nvidia‘s dominance in the space, have provided big challenges for Beijing’s efforts, at least in the short term, analysts told CNBC.

Nvidia’s well-documented boom has been driven by large cloud computing players buying its server products which contain its graphics processing units, or GPUs. These chips are enabling companies, such as ChatGPT maker OpenAI, to train their huge AI models on massive amounts of data.

These AI models are fundamental to applications like chatbots and other emerging AI applications.

The U.S. government has restricted the export of Nvidia’s most advanced chips to China since 2022, with restrictions tightening last year.

Such semiconductors are key to China’s ambitions to become a leading AI player.

CNBC spoke to analysts who identified some of China’s leading contenders that are looking to challenge Nvidia, including technology giants Huawei, Alibaba and Baidu and startups such as Biren Technology and Enflame.

The overarching view is that they are lagging behind Nvidia at this point.

“These companies have made notable progress in developing AI chips tailored to specific applications (ASICs),” Wei Sun, a senior analyst at Counterpoint Research, told CNBC.

“However, competing with Nvidia still presents substantial challenges in technological gaps, especially in general-purpose GPU. Matching Nvidia in short-term is unlikely.”

Chinese firms have a “lack of technology expertise”, according to Sun, highlighting one of the challenges.

However, it’s the U.S. sanctions and their knock-on effects that pose the biggest roadblocks to China’s ambitions.

Some of China’s leading Nvidia challengers have been placed on the U.S. Entity List, a blacklist which restricts their access to American technology. Meanwhile, a number of U.S. curbs have restricted key AI-related semiconductors and machinery from being exported to China.

China’s GPU players all design chips and rely on a manufacturing company to produce their chips. For a while, this would have been Taiwan Semiconductor Manufacturing Co., or TSMC. But U.S. restrictions mean many of these firms cannot access the chips made by TSMC.

They therefore have to turn to SMIC, China’s biggest chipmaker, whose technology remains generations behind TSMC. Part of the reason why it’s lagging behind, is because Washington has restricted SMIC’s access to a key piece of machinery from Dutch firm ASML, which is required to manufacture the most advance chips.

Meanwhile, Huawei has been pushing development of more advanced chips for its smartphones and AI chips, which is taking up capacity at SMIC, according to Paul Triolo, a partner at consulting firm Albright Stonebridge.

“The key bottleneck will be domestic foundry leader SMIC, which will have a complex problem of dividing limited resources for its advanced node production between Huawei, which is taking up the lion’s share currently, the GPU startups, and many other Chinese design firms which have been or may be cutoff from using global foundry leader TSMC to manufacture their advanced designs,” Triolo told CNBC.

Nvidia has found success due to its advanced semiconductors, but also with its CUDA software platform that allows developers to create applications to run on the U.S. chipmaker’s hardware. This has led to the development of a so-called ecosystem around Nvidia’s products that others might find hard to replicate.

“This is the key, it is not just about the hardware, but about the overall ecosystem, tools for developers, and the ability to continue to evolve this ecosystem going forward as the technology advances,” Triolo said.

Triolo identified Huawei as one of the leaders in China with its Ascend series of data center processors.

The firm’s current generation of chip is called the Ascend 910B, and the company is gearing up to launch the Ascend 910C, which could be on par with Nvidia’s H100 product, according to a Wall Street Journal report in August.

In its annual report earlier this year, Nvidia explicitly identified Huawei, among other companies, as a competitor in areas such as chips, software for AI and networking products.

In the area of software and building a developer community, Huawei “holds lots of advantages,” Triolo said. But it faces similar challenges to the rest of the industry in trying to compete with Nvidia.

“The GPU software support ecosystem is much more entrenched around Nvidia and to a lesser degree AMD, and Huawei faces major challenges, both in producing sufficient quantities of advanced GPUs such as part of the Ascend 910C, and continuing to innovate and improve the performance of the hardware, given U.S. export controls that are limiting the ability of SMIC to produce advanced semiconductors,” Triolo said.

The challenges facing China’s Nvidia competitors have been evident over the past two years. In 2022, Biren Technology carried out a round of layoffs, followed by Moore Threads the year after, with both companies blaming U.S. sanctions.

But startups are still holding out hope, looking to raise money to fund their goals. Bloomberg reported last week that Enflame and Biren are both looking to go public to raise money.

“Biren and the other GPU startups are staffed with experienced industry personnel from Nvidia, AMD, and other leading western semiconductor companies, but they have the additional challenge of lacking the financial depth that Huawei has,” Triolo said.

“Hence both Biren and Enflame are seeking IPOs in Hong Kong, to raise funding for additional hiring and expansion.”