Crypto World

Nvidia (NVDA) Stock Gains as $82B Revenue Stream Emerges from AWS and China Deals

Published

4 weeks ago

25 March 2026

NewsAdmin

Quick Overview

Nvidia shares advanced 1.6% in premarket hours Wednesday, trading at $177.97
Arm Holdings unveiled the Arm AGI CPU for data centers, projecting $15B yearly revenue by 2031
The Arm chip doesn’t directly challenge Nvidia’s GPU dominance but may overlap with Nvidia’s Vera CPU lineup
Amazon Web Services committed to acquiring 1 million Nvidia GPUs for AI inference workloads, valued above $50 billion
Nvidia restarts H200 chip manufacturing and develops China-compliant Groq 3 variants, potentially adding $32B in annual sales

Nvidia shares moved higher during early Wednesday sessions, brushing aside concerns about Arm Holdings’ entry into the AI chip arena. The development coincided with two significant revenue opportunities that had escaped widespread attention.

NVDA Stock Card
NVIDIA Corporation, NVDA

Arm revealed its inaugural data center CPU on Tuesday evening—the Arm AGI CPU—identifying Meta Platforms and OpenAI among its initial clients. During an after-hours investor presentation, Arm outlined aggressive financial targets, forecasting approximately $15 billion in yearly CPU revenue by 2031 as part of a comprehensive $25 billion revenue objective.

Arm shares surged 12% in premarket activity following the disclosure.

However, industry observers were swift to clarify that this new processor doesn’t directly challenge Nvidia’s flagship GPU offerings.

Benchmark Research’s Cody Acree noted that Arm’s strategy is “less about catching up to the accelerator wave and more about inserting itself deeper into the architecture that governs how AI infrastructure actually runs.”

Jensen Huang, Nvidia’s CEO, appeared in Arm’s promotional content, characterizing their nearly twenty-year collaboration as the backbone for “one seamless platform, from cloud to edge to AI factories.”

The competitive dynamic becomes more nuanced regarding Nvidia’s recently launched Vera CPUs, introduced during last week’s developer conference. J.P. Morgan’s Harlan Sur highlighted potential overlap between Arm’s chip and that product category. He additionally noted Meta’s existing agreement with Nvidia for Arm-architecture CPUs—complicating the competitive landscape.

Amazon Web Services Makes Massive Nvidia Commitment

Separately, Amazon Web Services revealed plans to procure 1 million Nvidia GPUs dedicated to AI inference capabilities. The announcement caught many off guard—AWS had previously promoted itself as housing “the largest cluster of non-Nvidia chips in the world” following its October 2025 Indiana data center deployment.

The agreement encompasses a “broad mix” of six supplementary Nvidia chip variants, including the recently announced Groq 3 inference processors, alongside Nvidia networking equipment. Industry estimates place the complete package well beyond $50 billion, with completion targeted by late 2027.

This singular agreement accounts for approximately 25% of Nvidia’s total 2025 annual revenue.

Chinese Market Revenue Resumes

CEO Jensen Huang confirmed last week that Nvidia is resuming manufacturing of its H200 processor—engineered to meet U.S. export control requirements—specifically for Chinese customers. Industry sources suggest a China-compliant Groq 3 variant is also under development.

Nvidia had incorporated zero Chinese data center sales into its Q1 projections. Throughout 2025, those revenues approximated $8 billion quarterly—roughly $32 billion on an annualized basis, representing about 15% of total 2025 revenue.

Together, the AWS contract and China market reentry represent over $82 billion in revenue streams absent from Nvidia’s current financial forecasts.

Nvidia shares traded 1.6% higher at $177.97 during premarket hours Wednesday, rebounding from a 0.3% decline in the previous session.

Source link

Crypto World

LayerZero Says Kelp Setup Caused Exploit, as Aave Loss Questions Mount

Published

5 minutes ago

20 April 2026

NewsAdmin

LayerZero Says Kelp Setup Caused Exploit, as Aave Loss Questions Mount

Interoperability protocol LayerZero claims that an inadequate setup tied to Kelp’s decentralized verifier network (DVN) enabled malicious actors to steal $290 million from Kelp DAO, adding that preliminary signs point to North Korea-linked threat actors.

An attacker drained about 116,500 Restaked ETH (rsETH), worth as much as $293 million at the time, from Kelp DAO’s LayerZero-powered rsETH bridge on Saturday.

LayerZero said Monday that the exploit stemmed from a single point of failure in Kelp’s setup, which relied on a single LayerZero DVN as the only verified path, despite LayerZero previously advising them against this.

“LayerZero and other external parties previously communicated best practices around DVN diversification to KelpDAO. Despite these recommendations, KelpDAO chose to utilize a 1/1 DVN configuration.”

In practice, that meant Kelp relied on a single verification path for cross-chain messages rather than requiring multiple independent checks.

The exploit quickly shifted attention from the technical cause to the question of who should absorb the losses, while the fallout spread into Aave, where the attacker used rsETH as collateral to borrow real liquidity.

Aave’s total value locked (TVL) had fallen by about $8.9 billion to $17.5 billion at the time of writing after the exploiter used the stolen funds to borrow on Aave, leaving about $195 million in “bad debt,” triggering withdrawals on the lending protocol.

LayerZero said Kelp’s rsETH bridge relied solely on the LayerZero Labs DVN, and argued that the incident reflected an unsafe application configuration rather than a compromise of LayerZero itself. The company said it is now urging all applications using 1/1 DVN setups to migrate to multi-DVN configurations and will stop signing or attesting messages for apps that retain the single verifier design.

Losses spark blame fight after $290 million Kelp exploit

With no recovery or compensation plan yet announced, users and market observers spent Monday debating whether losses should sit with Kelp DAO, LayerZero, Aave or rsETH holders themselves.

Yishi Wang, founder and CEO of open-source hardware wallet OneKey, said that the best path forward was to negotiate with the hacker, offer a 10% to 15% bounty, and get the bulk of the funds back.

“If negotiations fail, LayerZero’s ecosystem fund should foot the bulk of the bill—it’s got the deepest pockets and the most long-term skin in the game,” wrote the founder in a Monday X post, adding that Kelp DAO is “broke” and could make it up with tokens and future revenue, or consider selling the project.

Analytics platform DeFiLlama’s pseudonymous founder, 0xngmi, outlined three solutions, including the option to “socialize” losses among all users, “rug rsETH holders on L2s,” or try to return holder balances to a pre-hack snapshot, which would be “very hard to do,” he wrote in a Monday X post.

Cointelegraph reached out to Aave for comment, but had not received a response by publication.

Exploit raises Aave liquidation risks

Investor concerns about the Kelp exploit have significantly reduced Ether (ETH) liquidity on Aave, the lending protocol’s core collateral asset.

This low liquidity presents a “critical safety risk where liquidations of ETH collateral cannot take place while markets are at 100% utilization,” said MoneySupply, the pseudonymous head of strategy at Aave competitor lending protocol Spark, in a Saturday X post.

“With current illiquidity conditions on Aave, a 15-20% ETHUSD price drop could cause significant bad debt accumulation (on top of any potential issues attributable to the direct rsETH exploit),” he said.