#AnthropicComputeRace

The AI compute race is accelerating. Anthropic has secured access to more than 220,000 NVIDIA GPUs and over 300 MW of compute capacity through its partnership with SpaceX’s Colossus infrastructure. Reports also suggest Anthropic is exploring additional compute channels, including Microsoft’s Maia AI chips. 220K+ NVIDIA GPUs 300 MW compute capacity Multi-cloud strategy across AWS, Google, Microsoft and SpaceX Massive expansion to support future AI models The battle for AI leadership is no longer just about better models — it’s about who can secure the most compute. Will compute power become the ultimate competitive advantage in the AI era? #AnthropicComputeRace

🚀 The AI compute race is accelerating. Anthropic has secured access to more than 220,000 NVIDIA GPUs and over 300 MW of compute capacity through its partnership with SpaceX’s Colossus infrastructure. Reports also suggest Anthropic is exploring additional compute channels, including Microsoft’s Maia AI chips. 🔹 220K+ NVIDIA GPUs 🔹 300 MW compute capacity 🔹 Multi-cloud strategy across AWS, Google, Microsoft and SpaceX 🔹 Massive expansion to support future AI models The battle for AI leadership is no longer just about better models — it’s about who can secure the most compute. 👀 Will compute power become the ultimate competitive advantage in the AI era? $NVDA #AnthropicComputeRace

🚀 The AI compute race is accelerating. Anthropic has secured access to more than 220,000 NVIDIA GPUs and over 300 MW of compute capacity through its partnership with SpaceX’s Colossus infrastructure. Reports also suggest Anthropic is exploring additional compute channels, including Microsoft’s Maia AI chips. 🔹 220K+ NVIDIA GPUs 🔹 300 MW compute capacity 🔹 Multi-cloud strategy across AWS, Google, Microsoft and SpaceX 🔹 Massive expansion to support future AI models The battle for AI leadership is no longer just about better models — it’s about who can secure the most compute. 👀 Will compute power become the ultimate competitive advantage in the AI era? $NVDA $ANTHROPIC $OPENAI #AnthropicComputeRace #OKXPizzaDay #TrillionDollarIPOs

SpaceX has submitted an S-1 filing, revealing its possession of 18,712 Bitcoin, valued at approximately $1.29 billion as of March 31, 2026. According to NS3.AI, the filing indicates that a potential listing could value SpaceX at around $1.75 trillion. Additionally, the filing disclosed a binding agreement with Anthropic for compute services, valued at nearly $45 billion over the next three years. #SpaceXBitcoinHoard #TrillionDollarIPOs #CryptoMomExitsSEC

$SPACEX IPO ISN’T JUST A LISTING… IT’S A FUTURE BET 🌍🔥 At first glance, the filing looked like another major tech IPO. But the deeper investors looked. the clearer it became this is something entirely different. 👀 This isn’t only about rockets anymore. It’s Elon Musk putting his long-term vision for humanity, AI, and global connectivity into one massive public empire. Some of the revelations inside the filing stunned the market: 💰 Musk could reportedly earn a historic $1 TRILLION compensation package if SpaceX reaches a $7.5 trillion valuation and achieves a self-sustaining Mars population of 1 million people. 🤖 Reports suggest Anthropic one of Grok’s biggest competitors entered a massive ~$45B partnership connected to SpaceX’s COLOSSUS computing network. 📡 Buying SpaceX may also mean indirect exposure to: Starlink, xAI, X, and Grok — all operating within the same expanding ecosystem. 📊 The company generated nearly $18.7B in revenue while continuing to spend aggressively on growth and infrastructure. 🛰️ Despite operating thousands of satellites, reports claim the fleet remains uninsured#NvidiaBeatsButDrops #CoinMoveAlert #OKXPizzaDay

THE AI WAR JUST ENTERED ITS MOST DANGEROUS PHASE #OpenAIvsAnthropic Two AI giants. Two completely different strategies. And now… the entire market is forced to choose a side. OpenAI dominates consumer attention with over 900 million users and a rumored Q4 IPO targeting an eye-watering $852B valuation. Meanwhile, Anthropic is quietly becoming the monster hiding behind enterprise AI. 32% enterprise market share. 3x revenue growth. And whispers of a potential $900B raise. This is no longer just a competition between AI models. It’s a battle between two visions of the future. Bet on OpenAI if you believe scale is everything. Hundreds of millions of users. Global brand dominance. Consumer mindshare so massive it becomes impossible to replace. But bet on Anthropic if you believe the real money lives inside enterprise contracts, sticky clients, and premium AI infrastructure quietly powering the corporate world. And then came the moment that changed everything: Microsoft officially ended its model exclusivity with OpenAI in 2026. That single move shattered the illusion that the AI race already had a winner. Now the battlefield is wide open again. Capital is rotating. Narratives are colliding. And traders are no longer just watching the AI war… They’re trading it. Both OpenAI and Anthropic now have Pre-IPO perpetual contracts available on OKX, allowing traders to speculate on valuation swings without owning a single share. No venture capital. No private equity access. Just pure market conviction. This isn’t just another tech rivalry anymore. It’s becoming the financialization of the AI future itself. And the scariest part? The winner of this war may end up controlling the next generation of the internet. $ANTHROPIC $OPENAI

But this time, the narrative is bigger than normal AI tokens. OKX bringing attention to $OPENAI, $ANTHROPIC and $SPACEX changed the psychology. Now traders are not only chasing crypto AI. They are chasing the idea that private tech, pre-IPO speculation and on-chain markets are merging. That is why attention is spreading across: $OPENAI $ANTHROPIC $SPACEX $VIRTUAL $AIXBT $AI $FET $RENDER $TAO $NEAR $ICP $IO This is a powerful basket. But powerful does not mean safe. AI narratives move fast because they combine three things: Future hype. Retail imagination. High volatility. That is why the upside can be explosive, but the exits can be brutal. The healthier AI-related names are the ones with broader infrastructure stories: $RENDER for compute. $TAO for decentralized intelligence. $FET for agents. $NEAR for AI applications. $ICP for on-chain compute. $LINK and $PYTH for data. The more emotional zone is: $OPENAI $ANTHROPIC $SPACEX $VIRTUAL $AIXBT $AI These can attract massive attention, but when a trade becomes too crowded, the market starts hunting late buyers. That is the real risk. AI is not dead. AI is not guaranteed. AI is a liquidity magnet. And in this market, liquidity magnets can pump hard and reverse faster than expected. The play is not to blindly chase every AI candle. The play is to separate narrative strength from emotional FOMO. Because the market may love AI. But the market loves liquidations even more. #USTreasuryHits19YrHigh #SamsungStrikeBegins

𝗢𝗞𝗫 𝗜𝘀 𝗧𝘂𝗿𝗻𝗶𝗻𝗴 𝗪𝗮𝗹𝗹 𝗦𝘁𝗿𝗲𝗲𝘁 𝗜𝗻𝘁𝗼 𝗮 𝟮𝟰/𝟳 𝗖𝗿𝘆𝗽𝘁𝗼 𝗠𝗮𝗿𝗸𝗲𝘁 This is bigger than tokenized stocks. OKX is building a new layer where stocks, commodities, AI, chips, crypto equities and pre-IPO names trade inside one crypto-native market. That is why #StocksGoOnChain matters. Gold and silver are already represented through $XAU and $XAG . Oil exposure is live through $CL , $BZ and $USO . Broader market risk appears through $SPY and $QQQ , while Korea and chip-cycle exposure show up through $EWY . But the real battlefield is AI hardware. $NVDA is the face of the AI chip trade. $AMD is the challenger. $TSM is the manufacturing backbone. $ARM powers the architecture layer. $MU , $DRAM , $WDC and $SNDK sit inside the memory and storage cycle. $INTC , $QCOM , $MRVL , $AVGO and $COHR show how deep the semiconductor chain goes. Then comes the software and cloud layer: $AAPL , $MSFT , $GOOGL , $AMZN , $META , $ORCL and $PLTR . The crypto-linked equity layer is even more interesting: $MSTR is the Bitcoin treasury bet. $COIN is exchange infrastructure. $HOOD is retail trading access. $CRCL is the stablecoin infrastructure trade. $BMNR brings mining exposure back into the picture. And now OKX has the explosive pre-IPO layer too: $OPENAI , $ANTHROPIC and $SPACEX . This is where things get serious. Retail used to wait for Wall Street to open the IPO door. Now traders can price AI, space, chips, gold, oil, Bitcoin equities and mega-cap tech in one environment. That changes market psychology. Crypto is no longer only trading crypto. It is absorbing TradFi. #StocksGoOnChain is not just a hashtag. It is the early version of one global liquidity layer. #StocksGoOnChain #TradeAIStocksOnOKX

AI Coins — The Trillion Dollar Bet Nobody’s Pricing While retail watches $BTC candles, the AI sector quietly built the strongest narrative of 2026. And almost nobody is positioned for what’s coming. The Setup: NVIDIA reports earnings May 20. $78.8B revenue expected. Another beat = AI thesis confirmed. Every AI token reprices instantly. But here’s what people miss: AI coins aren’t just NVIDIA proxies. They’re betting on decentralized AI infrastructure — the part Big Tech can’t fully control. The Players: 🚀 $TAO (Bittensor) — The “Bitcoin of AI.” Decentralized model training network. Real revenue, real usage, real subnets generating value. 🚀 $RENDER — Distributed GPU compute. As demand explodes, supply networks like Render matter more. 🚀 $FET Fetch.ai) — AI agents on-chain. The autonomous agent narrative is just starting. 🚀 $WLD (Worldcoin) — Sam Altman’s bet on proof-of-humanity in an AI world. Why Now: → NVIDIA earnings May 20 = catalyst → OpenAI Q4 IPO at $852B valuation → Anthropic eyeing $900B raise → Microsoft ended OpenAI exclusivity = competition opens → Every major sovereign building national AI strategies The macro story: AI infrastructure is the new oil. Whoever owns the compute and data layers wins the next decade. The Brutal Reality: Most AI tokens won’t survive. 90% are vaporware riding the narrative. But the 10% that have real product-market fit could 10-50x. The trick is filtering. Look for: ✅ Real revenue, not just promises ✅ Active developer community ✅ Compute or data utility ✅ Token has actual function, not just speculation Trade Angles: 🟢 Accumulate during fear, not euphoria 🟢 Diversify across 3-4 quality names, not all-in on one ⚠️ Watch NVDA earnings — direct correlation ⚠️ Avoid pure AI memes — they die first Bottom Line: The AI revolution is happening with or without crypto. But crypto-AI gives retail a way to participate in the infrastructure layer that VCs are locked out of. #FedMeetsNVIDIAMay20 #OpenAIvsAnthropic #StocksGoOnChain

#AnthropicComputeRace: $200 Billion Committed. 220,000 GPUs. The AI War Is Now a Real Estate Business. The AI race used to be about algorithms. In 2026, it's about megawatts. Anthropic locked in three of the biggest infrastructure deals in AI history within a single quarter. Google committed up to $40 billion — $10 billion upfront, $30 billion contingent on performance targets — with the partnership centered specifically on computing infrastructure. Anthropic then committed $200 billion to Google Cloud over five years, sending Google's stock up 10%. Then came SpaceX: SpaceXAI agreed to provide Anthropic with access to more than 300 megawatts of capacity and roughly 220,000 NVIDIA GPUs from its Colossus 1 facility. Crypto Times + 2 The result: Anthropic now has two major compute relationships — Google Cloud and SpaceXAI — which reduces single-point-of-failure risk in its infrastructure. For a company scaling to $44 billion in annualized revenue, that redundancy isn't a luxury — it's operational survival. The Crypto Basic The immediate user-facing result is straightforward: Anthropic lifted Claude usage limits and scaled up compute access for enterprise users. The strategic meaning runs deeper. Every enterprise deal Anthropic wins requires capacity to deliver on. When AI tools move from occasional chat interfaces to always-on work systems, capacity becomes destiny. Crypto NewsNewsBTC The SpaceX side of the story is equally telling: Grok downloads have fallen from 20 million in January to 8.3 million in April, and SpaceX is ceding Colossus capacity rather than using it for internal model development. The most powerful AI supercomputer in the world is now running Claude workloads, not xAI's. Bitget Intelligence needs a factory now. Anthropic just built three of them. #AnthropicComputeRace

Gavin Baker on SpaceX's booming datacenter business: $15B from Anthropic is just the beginning @Jason: “ Gavin, your take on the S-1, and I think specifically @elonmusk Web Services?” @GavinSBaker: “$15 billion (from Anthropic), that means (SpaceX’s) AI business is going to quadruple. It has already effectively quadrupled. Their first data center was 122 days. For the second one, it took them 91 days. The third one was, I think, 66 days. They build data centers dramatically faster than anyone else at a lower cost. And now that you have a clear offtake partner, and I would expect ‘partner’ to become ‘partners,’ there is no reason they can't start stamping these data centers out really fast. And having watched Jensen for a long time, it is important to Jensen that his GPUs be used. And so GPUs will be allocated to who can plug them in, turn them on, and start converting electrons into tokens. And so I think this business can grow dramatically faster than maybe what anyone could have contemplated three months ago. But $15 billion from Anthropic is extraordinary.”

#AnthropicComputeRace is getting serious. AI companies are no longer just racing on models they’re racing for compute, power, GPUs, and infrastructure. Anthropic locking down massive compute capacity shows how aggressive the AI race is becoming. At the same time, MSFT remains one of the strongest AI infrastructure plays, with cloud, data centers, and AI partnerships keeping it central to the narrative. Watching AI-related crypto coins closely this week. No hype just narrative + volume + confirmation. #AnthropicComputeRace $ANTHROPIC $MSFT

🚀💹 Global AI Compute Race Intensifies Anthropic is reportedly expanding its access to large-scale computing power, with more than 220,000 NVIDIA GPUs and around 300 MW of total infrastructure capacity through partnerships involving high-performance clusters, including SpaceX-associated Colossus systems. The company is also diversifying its hardware ecosystem, with potential interest in Microsoft’s Maia AI chips alongside its existing compute partners. 🔹 220K+ NVIDIA GPUs deployed 🔹 ~300 MW of total compute capacity 🔹 Multi-cloud + hybrid infrastructure approach (AWS, Google, Microsoft, SpaceX ecosystem) 🔹 Rapid scaling to support next-generation AI workloads This trend underscores a major shift: competitive advantage in AI is increasingly driven by access to compute at scale, not just algorithmic improvements. 👀 The real question is whether compute availability will become the ultimate strategic advantage in the AI era. $NVDA 💹

If you do some basic math on the Anthropic/SpaceX deal, you realize that SpaceX is making ~50% margin after depreciation, which is incredible. Shows you where the leverage was. A stark contrast to 2 years ago when neoclouds were selling GPUs at 0% margin.

$2Z This is the *2Z/USDT 3D chart*. It has a "New" tag, meaning it was recently listed. Let's analyze it: 1. *Current Situation* - *Price*: 0.11304, up +0.15% today. The 24h high is just below 0.11984. - *Trend*: There has been a solid recovery from the March 2026 low of 0.06551. It is currently making higher highs and higher lows. - *MAs*: Price is above MA5 at 0.10238, MA10 at 0.09597, and MA20 at 0.08837. The MAs are also upward sloping, which is a strong bullish sign. 2. *Key Levels* Level Type Note **0.11984** Immediate Resistance 24h high, price has stopped here for now **0.12000** Psychological Resistance Round number, short-term target **0.14000** Major Resistance Previous swing high area **0.10238** Immediate Support MA5, short-term hold level **0.09597** Major Support MA10, pullback limit **0.06551** Strong Support Recent bottom, trend invalidation level 3. *Analysis* - *Structure*: Dumped from 0.20080 in Nov 2025, bottomed at 0.06551 in March 2026. Since then, a clean uptrend has been ongoing with buying on every dip. - *Momentum*: Price is above all MAs and MAs are upward. Up +37.55% in 30D and +55.48% in 90D, indicating momentum is with buyers. - *Volume*: Volume is increasing with the recovery. 24h volume is 7.97M 2Z, liquidity is decent for a new token. - *Risk*: The 0.12 zone is psychological resistance where profit booking may occur. If it doesn't break, a pullback to 0.102-0.096 is possible. 4. *Next Target* If it breaks 0.11984 and 0.12000 with a 3D close: 1. *Short term target*: 0.14000 - old supply zone 2. *Medium term target*: 0.16000 - 0.18000, if momentum continues 3. *Extension*: 0.20080 retest possible, but this is strong resistance If rejected, support is at 0.10238, below which a pullback to 0.09597 may occur. A 3D close below 0.06551 will break the trend. --- *Note*: This is a new/low-cap token with high volatility and slippage. #AnthropicComputeRace #BTCReserveCodified

A TON OF THINGS HAPPENED IN THE STOCK MARKET TODAY. Here's a full recap: 1. Anthropic is reportedly in early talks to rent Azure servers powered by Microsoft’s $MSFT Maia AI chips, per The Information. The move would further deepen Microsoft’s relationship with Anthropic, after Microsoft pledged up to $5B to the company, Anthropic committed $30B in Azure spending, and Microsoft began reselling Claude through Azure. The chip angle matters because Maia is being positioned as a lower-cost option than Nvidia’s $NVDA GPUs for certain inference workloads, rather than a replacement for frontier AI training. Anthropic already uses chips from Amazon $AMZN, Google $GOOGL, and Nvidia $NVDA, and Microsoft now wants Maia to become part of that mix. 2. Nvidia $NVDA PT's were raised across the street with the highest of $500 at Baird. The firm says Nvidia continues gaining market share in AI inferencing and across hyperscalers, while adoption of the upcoming Vera Rubin platform at frontier model companies is expected to exceed that of Blackwell. Baird also believes Vera’s standalone performance could significantly outperform traditional x86 CPUs, opening an additional $200B TAM opportunity for Nvidia, which already has visibility into nearly $20B in CPU revenue this year alone. Looking ahead, the firm expects annual AI infrastructure spending to grow from over $1T in 2027 to roughly $3–4T by 2030 as agentic AI proliferates across industries, while reiterating Nvidia as one of its top investment ideas. 3. U.S. initial jobless claims came in at 209,000, slightly below expectations of 210,000, signaling the labor market remains relatively stable. 4. OpenAI reportedly generated about $5.7B in Q1 revenue, roughly $1B ahead of Anthropic, per The Information. Anthropic could narrow that gap quickly, with its recent annualized revenue reportedly nearing $45B compared to OpenAI’s $25B annualized run rate in February. Anthropic is also projecting nearly $11B in Q2 revenue and about $600M in operating profit. OpenAI’s Q2 outlook was not reported, but The Information said growth has been supported by Codex, enterprise sales, ad tests, GPT-5.5, and image generation. 5. Walmart $WMT reported earnings this morning, with EPS of $0.66 coming in line with expectations and revenue of $177.8B beating estimates of $175B. Despite the revenue beat, the stock fell 7% after the company lowered guidance and warned it would raise prices on certain consumer goods. 6. Starbucks $SBUX has reportedly retired a worker-facing AI tool designed to automate inventory counts and help address product shortages, according to Reuters. The program was shut down this week, just nine months after being rolled out across North American stores. 7. A reported draft agreement between the U.S. and Iran has reportedly been reached through Pakistan-mediated talks, with an official announcement expected within hours. The deal is said to include an immediate comprehensive ceasefire across all fronts, along with commitments from both sides not to target infrastructure. The agreement would also guarantee freedom of navigation through the Persian Gulf and Strait of Hormuz under a joint monitoring mechanism. In exchange for Iran’s compliance, sanctions would reportedly be gradually lifted, while negotiations on unresolved issues are expected to begin within seven days. 8. $AMD AMD announced more than $10B in investments across Taiwan’s ecosystem to expand advanced packaging capacity and accelerate next-generation AI infrastructure deployments. The company highlighted products ranging from its 6th Gen EPYC CPUs, codenamed “Venice,” to its Helios rack-scale AI platform featuring Instinct MI450X GPUs, with multi-gigawatt deployments expected to begin in the second half of 2026. AMD also said it reached another key production milestone with TSMC, as Venice EPYC CPUs are now ramping on TSMC’s 2nm process technology in Taiwan, with future plans to expand production to TSMC’s Arizona fabrication facilities. 9. The U.S. is preparing to award $2B to nine quantum companies through the CHIPS and Science Act while also taking minority equity stakes. IBM is expected to receive $1B, GlobalFoundries $375M, and other recipients including $QBTS D-Wave, $RGTI Rigetti, $INFQ Infleqtion, Atom Computing, PsiQuantum, and Quantinuum are expected to receive $100M each. Diraq is expected to receive $38M. The deals would combine grants with government equity stakes, though they still need to be finalized. 10. The Pentagon is testing AI models from OpenAI, Google, and others as it looks for alternatives to Anthropic’s Claude, per Bloomberg. The tests began in March with 25 heavy users of the Defense Department’s AI tools across five military theater commands. The move follows Defense Secretary Pete Hegseth designating Anthropic as a supply-chain risk over its model guardrails, a designation Anthropic is now challenging in court. Claude is currently used heavily in the Pentagon’s Maven Smart System, including classified operations. 11. Oura, the maker of the popular smart ring, has reportedly filed confidentially for an IPO and plans to go public later this year, per Bloomberg. The company’s growth is the main story: revenue is expected to reach $1.5B in 2026, up from $500M in 2024, while total rings sold have climbed to 5.5M from 2.5M in mid-2024. Oura was last valued at $11B. 12. A Bank of America survey shows investors are increasingly worried the Federal Reserve may need to raise interest rates again. According to the survey, 14% of investors believe conditions for a hike are already in place, while 38% think the Fed would tighten policy if core inflation rises to 3.5%–4%, regardless of labor market strength. Overall, more than half of respondents see rate hikes as possible under certain inflation scenarios, reflecting renewed concern about persistent inflation pressure. I will also be launching a product tomorrow morning at 10AM EST that I think many retail investors will enjoy. 100% free. See you then! WALL STREET IS THE GREATEST SHOW ON EARTH.

SCOOP: Anthropic is in talks to use Microsoft's Maia 200 chips for inference, and has been upping its spending on Azure as its customers surge. Talks are early, but would be a coup for Microsoft's in-house chip effort.

Seeing the news that Anthropic is going to use Maia made me think… At this point, does performance even matter as long as the chip works decently well? How can they use Maia? Maybe you can just go to TSMC, ask them to make a reasonably functional chip, and sell it to Anthropic… lol What exactly is the moat for ASICs?

ANTHROPIC EYES MICROSOFT CHIPS Anthropic is in early talks to rent Azure servers powered by $MSFT’s Maia AI chips, per The Information. This would deepen a relationship that is already moving fast: Microsoft has pledged up to $5B to Anthropic, Anthropic committed $30B of Azure spend, and Microsoft is reselling Claude through Azure. The chip angle matters because Maia is being pitched as cheaper than Nvidia for some inference workloads, not as a frontier training replacement. Anthropic already uses Amazon, Google, and Nvidia chips. Now Microsoft wants Maia in that mix.

Anthropic is reportedly in talks to rent servers powered by $MSFT Maia AI chips as it looks for more compute to run Claude. This shows how severe the AI compute shortage still is with frontier labs needing every viable source of capacity they can get.