Amazon Plans to Directly Compete with Nvidia Through Its Own AI Chip Sales

If Amazon Web Services follows through on its plans, the cloud services provider is set to expand its presence significantly within Nvidia’s market, potentially presenting one of the most serious challenges yet to Nvidia’s dominance in the AI chip sector.
Amazon’s head of AI, Peter DeSantis, told Bloomberg that AWS is in discussions about selling its Trainium AI chips to other companies for use in data centers. He declined to identify which firms might be interested in purchasing these chips.
According to TechCrunch, these discussions regarding chip sales are still in their early stages. They originated from Amazon CEO Andy Jassy’s annual shareholder letter issued in early April, in which he mentioned that the company’s in-house AI chips were in such high demand that he was considering selling them.
“If our chip business operated independently and sold the chips produced this year to AWS and other third parties, just as other leading chip manufacturers do, our annual revenue would be around $50 billion. The demand for our chips is so strong that it’s entirely possible we’ll end up selling large quantities of them to external clients in the future,” he stated.
Just how much of a threat could Amazon pose to Nvidia? A competitor with $50 billion in annual revenue would unlikely bring Nvidia down — which currently generates around $326 billion in annual revenue — as long as it continues to perform as well as it has recently. However, that figure is comparable to Intel’s annual revenue levels.
So far, AWS has avoided selling its AI chips for various reasons. The primary one is the revenue structure associated with these chips. While AWS does charge customers directly for the computing power provided by these chips on its cloud platform, it also earns revenue from a wide range of other services that companies need for their AI applications, including storage, security, networking, and monitoring solutions.
Equally significant is Amazon’s claim that demand for its chips exceeds production capacity. In the same April shareholder letter, Jassy noted that the current supply of Trainium chips was sold out almost immediately. He added that the same was true for the upcoming Trainium4 model, which won’t be available for more than a year. This was before AWS officially integrated OpenAI models into its service offerings.
Therefore, if AWS were to sell its chips to external parties, it would likely have to place current customers on waiting lists, unless it could produce an excess supply of chips through partners like TSMC. Even then, it would need to overcome significant competition from Nvidia to secure enough chips through TSMC, which has recently surpassed Apple as the company’s largest chip manufacturing client.
Doron Aronson, AWS spokesperson and guide during a recent private tour of AWS’s chip design facility, also confirmed that the company is open to selling these chips. “Although we have traditionally declined requests to sell chips directly, Andy mentioned that it’s quite possible we’ll sell large quantities of them to third parties in the future,” he said.
While Nvidia’s founder and CEO, Jensen Huang, recently announced that he has identified a new $200 billion market for Nvidia by selling CPUs for AI use, thereby expanding into the competitive space once held by Intel and AMD, Jassy clearly has his own ambitions for chip sales: a $50 billion market that would put him in direct competition with Nvidia.
Related article
Nvidia quietly builds multibillion-dollar behemoth to rival chip business
Nvidia CEO Jensen Huang anticipated the market by over a decade, initiating work on AI-specific chips in 2010, long before the current AI boom. A parallel strategic move in 2020—bolstering data center networking through a key acquisition—has created
Huawei to launch new AI chip in Q1 2027, challenging Nvidia
Huawei has accelerated the release of its next-generation Ascend 960DT AI processor to Q1 2027, aiming to compete directly with Nvidia, the company revealed at its Huawei Connect event on Thursday.Previously, the Chinese technology leader targeted a
Amazon launches new $1 billion FDE org following OpenAI and Anthropic
As businesses grapple with AI integration, they are turning to external experts, prompting service providers to establish specialized teams to ensure successful implementation.On Tuesday, Amazon Web Services (AWS) announced a new internal unit dedica
Related Special Topic Recommendations
Comments (0)
0/500

If Amazon Web Services follows through on its plans, the cloud services provider is set to expand its presence significantly within Nvidia’s market, potentially presenting one of the most serious challenges yet to Nvidia’s dominance in the AI chip sector.
Amazon’s head of AI, Peter DeSantis, told Bloomberg that AWS is in discussions about selling its Trainium AI chips to other companies for use in data centers. He declined to identify which firms might be interested in purchasing these chips.
According to TechCrunch, these discussions regarding chip sales are still in their early stages. They originated from Amazon CEO Andy Jassy’s annual shareholder letter issued in early April, in which he mentioned that the company’s in-house AI chips were in such high demand that he was considering selling them.
“If our chip business operated independently and sold the chips produced this year to AWS and other third parties, just as other leading chip manufacturers do, our annual revenue would be around $50 billion. The demand for our chips is so strong that it’s entirely possible we’ll end up selling large quantities of them to external clients in the future,” he stated.
Just how much of a threat could Amazon pose to Nvidia? A competitor with $50 billion in annual revenue would unlikely bring Nvidia down — which currently generates around $326 billion in annual revenue — as long as it continues to perform as well as it has recently. However, that figure is comparable to Intel’s annual revenue levels.
So far, AWS has avoided selling its AI chips for various reasons. The primary one is the revenue structure associated with these chips. While AWS does charge customers directly for the computing power provided by these chips on its cloud platform, it also earns revenue from a wide range of other services that companies need for their AI applications, including storage, security, networking, and monitoring solutions.
Equally significant is Amazon’s claim that demand for its chips exceeds production capacity. In the same April shareholder letter, Jassy noted that the current supply of Trainium chips was sold out almost immediately. He added that the same was true for the upcoming Trainium4 model, which won’t be available for more than a year. This was before AWS officially integrated OpenAI models into its service offerings.
Therefore, if AWS were to sell its chips to external parties, it would likely have to place current customers on waiting lists, unless it could produce an excess supply of chips through partners like TSMC. Even then, it would need to overcome significant competition from Nvidia to secure enough chips through TSMC, which has recently surpassed Apple as the company’s largest chip manufacturing client.
Doron Aronson, AWS spokesperson and guide during a recent private tour of AWS’s chip design facility, also confirmed that the company is open to selling these chips. “Although we have traditionally declined requests to sell chips directly, Andy mentioned that it’s quite possible we’ll sell large quantities of them to third parties in the future,” he said.
While Nvidia’s founder and CEO, Jensen Huang, recently announced that he has identified a new $200 billion market for Nvidia by selling CPUs for AI use, thereby expanding into the competitive space once held by Intel and AMD, Jassy clearly has his own ambitions for chip sales: a $50 billion market that would put him in direct competition with Nvidia.
Nvidia quietly builds multibillion-dollar behemoth to rival chip business
Nvidia CEO Jensen Huang anticipated the market by over a decade, initiating work on AI-specific chips in 2010, long before the current AI boom. A parallel strategic move in 2020—bolstering data center networking through a key acquisition—has created
Huawei to launch new AI chip in Q1 2027, challenging Nvidia
Huawei has accelerated the release of its next-generation Ascend 960DT AI processor to Q1 2027, aiming to compete directly with Nvidia, the company revealed at its Huawei Connect event on Thursday.Previously, the Chinese technology leader targeted a
Amazon launches new $1 billion FDE org following OpenAI and Anthropic
As businesses grapple with AI integration, they are turning to external experts, prompting service providers to establish specialized teams to ensure successful implementation.On Tuesday, Amazon Web Services (AWS) announced a new internal unit dedica





Home






