Google Undercuts Nvidia and GPU Cloud Rivals With Custom Chip
0h ago|4 min readStandard
Fazen Markets Editorial Desk
Collective editorial team · methodology
google-googlnvidia-nvda
Fazen Markets Editorial Desk
Collective editorial team · methodology
Trades XAUUSD 24/5 on autopilot. Verified Myfxbook performance. Free forever.
Risk warning: CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. The majority of retail investor accounts lose money when trading CFDs. Vortex HFT is informational software — not investment advice. Past performance does not guarantee future results.
Fazen Markets — Google announced the deployment of its new, fully custom Axion AI chip on 24 May 2026, signaling a direct challenge to Nvidia’s accelerator dominance and a strategic pivot with severe implications for GPU cloud renters like CoreWeave and Nebius. The announcement contributed to a down day for AI hardware stocks. Nvidia stock closed at $215.33, down 3.64%, while Google’s parent Alphabet traded at $382.97, down 1.53%, as of 21:50 UTC today. Reporting on the announcement first appeared on finance.yahoo.com.
The AI compute market has been dominated by Nvidia's H-series GPUs, which power the vast majority of training and inference workloads. This reliance created a supply bottleneck and high pricing power for Nvidia, translating into soaring revenues and a market capitalization exceeding the GDP of many nations. Hyperscalers like Google, Amazon, and Microsoft have long sought to reduce this dependency through custom silicon, aiming to lower costs and optimize performance for internal workloads.
What changed now is the transition from research and limited deployment to full-scale, production-grade adoption. Google’s Axion chip is not a test-bed project but a fully supported, commercially available product for its cloud customers. The catalyst is the maturation of internal chip design teams over a multi-year period, combined with the escalating cost of continuing to purchase Nvidia hardware at scale. Google previously used its Tensor Processing Units (TPUs) for specific tasks, but Axion represents a broader, general-purpose AI accelerator designed to compete directly with Nvidia’s core offerings.
This move follows Amazon's successful Graviton series for general compute and its Trainium and Inferentia chips for AI. Microsoft is also developing its own Athena AI chips. The collective shift marks an inflection from a single-vendor ecosystem to a multi-supplier market, fundamentally altering the economics of cloud AI services.
The immediate market reaction provided a clear assessment of perceived winners and losers. Nvidia's 3.64% single-day decline to $215.33 represented a steeper drop than the broader tech sector. Alphabet’s more modest 1.53% slide to $382.97 likely reflects the significant capital expenditure required for this strategic shift, offsetting the long-term cost-saving benefits. The day's trading ranges showed heightened volatility, with NVDA oscillating between $214.86 and $221.01.
| Metric | Google (GOOGL) | Nvidia (NVDA) |
|---|---|---|
| Price | $382.97 | $215.33 |
| Daily Change | -1.53% | -3.64% |
| Intraday Range | $381.77 - $388.74 | $214.86 - $221.01 |
This announcement directly pressures the business models of pure-play GPU cloud providers. CoreWeave and Nebius, which built their services on renting Nvidia hardware, now face a formidable competitor whose underlying cost structure could be significantly lower. For context, a single Nvidia H100 GPU can cost tens of thousands of dollars, and cloud providers mark up this hardware for rental. Google’s in-house silicon eliminates that markup, allowing it to potentially offer AI compute at far more aggressive price points.
The second-order effects ripple across several sectors. Direct losers include Nvidia and its reseller/cloud partners. Nvidia faces not just a loss of a major customer but the creation of a powerful competitor in its largest end-market. Chip equipment suppliers like ASML and Applied Materials may see a neutral to positive impact as they supply tools to all chipmakers, including Google’s manufacturing partners TSMC or Samsung. Server OEMs like Dell and Hewlett Packard Enterprise could see pressure as hyperscalers design more of their own hardware, reducing purchases of off-the-shelf servers.
The primary risk to the bearish Nvidia thesis is execution. Custom silicon is notoriously difficult to design, scale, and support with a strong software ecosystem. Nvidia’s CUDA platform remains the industry standard, and a mass developer exodus to Google’s new stack is not guaranteed. A failed or delayed Axion rollout would reinforce Nvidia’s moat.
Positioning data from recent weeks shows institutional investors rotating out of pure-play AI hardware names and into diversified tech giants with in-house chip capabilities. Flow has also increased into semiconductor design software firms like Cadence and Synopsys, which benefit from the proliferation of chip design projects regardless of the final manufacturer.
Markets will watch for two immediate catalysts: Google’s Q2 2026 earnings call in late July for detailed commentary on Axion adoption and capex guidance, and Nvidia’s next earnings report for any indication of order cancellations or slowdowns from cloud giants. The performance of CoreWeave’s rumored IPO, should it proceed, will serve as a live referendum on investor appetite for the GPU-cloud model in this new environment.
Key levels to monitor include Nvidia’s 200-day moving average, currently near $205, as a test of long-term trend support. For Alphabet, resistance sits at its recent high near $395. A sustained break above that level would signal investor confidence that the long-term benefits of vertical integration outweigh the near-term capex burden.
The next major industry event is the TSMC Technology Symposium in early June, where details on next-generation process nodes may reveal the manufacturing roadmap for both Nvidia’s and Google’s future chips.
For AI startups, Google's Axion chip could lower the cost of training and running large models, improving margins and reducing burn rates. However, it also creates a new lock-in risk, migrating dependency from Nvidia’s CUDA ecosystem to Google’s cloud platform and proprietary software stack. Startups must weigh lower compute costs against reduced portability between cloud providers.
The parallel to Apple's transition from Intel to its M-series chips is strong. Both are vertical integration moves by cash-rich tech giants to control their core technology stack, reduce costs, and differentiate their products. The key difference is market: Apple controls its own closed hardware ecosystem, while Google is selling compute as a service in a competitive, open market against established incumbents.
The impact on consumer GeForce GPUs is likely minimal in the short term. The data center and consumer markets are segmented, with different chips and economic drivers. However, if the data center business slows, Nvidia may allocate more production capacity and R&D focus to the consumer segment, potentially increasing supply or innovation for gaming and creator GPUs over a multi-year horizon.
Vortex HFT is our free MT4/MT5 Expert Advisor. Verified Myfxbook performance. No subscription. No fees. Trades 24/5.
Position yourself for the macro moves discussed above
Start TradingSponsored
Open a demo account in 30 seconds. No deposit required.
CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. You should consider whether you understand how CFDs work and whether you can afford to take the high risk of losing your money.