CoinDesk
Google plans to launch two specialized AI chips next year, one targeting confidential computing and another designed for data-heavy applications such as coding and drug discovery. Bloomberg Intelligence estimates TPU sales, including related hardware and software, could exceed $135 billion by 2028, up from an estimated $12 billion this year, with the chips costing about 30% less than Nvidia GPUs. Nvidia separately lost more than $250 billion in market value after its board announced an investigation into reports that CEO Jensen Huang had been targeted by foreign intelligence.
Why it matters
Google is broadening its TPU portfolio into workloads that require large data volumes and the protection of sensitive information. The estimated 30% cost advantage gives Google a direct pricing lever against Nvidia’s GPUs while giving developers a credible alternative for specialized AI infrastructure.
The projected expansion in TPU sales is bullish for a broader AI ecosystem that increasingly includes chips designed for specific workloads rather than one general-purpose accelerator. It also strengthens Google’s position across cloud computing, AI services and enterprise applications.
Market impact
Nvidia’s leadership position now faces pressure from two directions: a stronger in-house chip program at Google and uncertainty surrounding Huang. Shares fell 4.3% on Monday after losing more than $250 billion in market value, while Google gained 3.4% as the Nasdaq rose more than 2%.
Nvidia said it had no reason to believe national security or company secrets had been compromised. Independent directors and outside lawyers will oversee the inquiry, while the board backed Huang and said he would cooperate. The near-term market focus is whether lower-cost TPUs begin eroding Nvidia’s pricing power as Google expands specialized AI deployments.
Frequently asked questions
-
What two AI chips does Google plan to launch next year?
One TPU will target confidential computing for sensitive workloads, while the other will support data-heavy applications such as coding and drug discovery.
-
How much cheaper are Google’s TPUs than Nvidia GPUs?
Bloomberg Intelligence estimates that the TPUs cost about 30% less, creating a potential pricing advantage for specialized AI deployments.
-
How large could Google’s TPU business become by 2028?
TPU sales, including related hardware and software, could exceed $135 billion by 2028, compared with an estimated $12 billion this year.
-
Why did Nvidia lose more than $250 billion in market value?
Nvidia’s board announced an investigation into reports that CEO Jensen Huang had been targeted by foreign intelligence, adding uncertainty to the company’s leadership.
-
What did Nvidia say about the leadership investigation?
Nvidia said it had no reason to believe national security or company secrets had been compromised. Independent directors and outside lawyers will oversee the inquiry.
Lookonchain
WuBlockchain
WatcherGuru
CoinTelegraph