STACK News & Signals

news4h agonews

Intel to Acquire Classiq to Create a Full-stack Quantum Computing Platform

72/100·B+Long

news5h agonews

NVIDIA and Nebius partner to scale full-stack AI cloud

72/100·B+Long

meme23h agomeme

Quite compelling economics https://x.com/deedydas/status/2114195317860266432 Quoted @deedydas: There is an absolute barrage of inference providers popping up that get to $1M/mo extremely quickly. This is the rough economics of inference providers: Even if you rent on-demand B300s at $8/GPU-hr, many open source models (GLM 5.3, Kimi K3, DeepSeek 4.1 Flash / Pro) can get you >$10-20/GPU-hr at reasonable throughputs (100tok/s). DeepInfer from DeepSeek directly shows they can get $38/GPU-hr on off-peak hours. Every chip / model has a different throughput / interactivity curve, but there is a lot of arbitrage between even high rental spot prices and token prices. New chips (Vera Rubins) also change this meaningfully. Your ability to be making gross margin relies on your utilization and how good your stack is. Gateways like OpenRouter, Vercel and more help solve that problem by aggregating demand to you. May inference providers discount their tokens to get longer term commits too. I've even heard stories of selling back Chinese model tokens to Chinese labs! One thorn in the economics of these players is that many Chinese open source model labs are now asking for rev share if you're at a certain scale, espp for pre-release access to their models for day 1 support. That further pushes the threshold to be in the money. The barrier to entry seems low and falling. There are even stories of smaller ones who can "vibe code" a way to stand up OSS models on a cluster they rent. And as a result, we're seeing 10s of such companies pop up almost every week and immediately skyrocketing revenue at low-ish gross margin and raising rounds on the back of that. Unsurprisingly, because the actual service provided seems commodity, this eventually drives up compute prices and drives down token prices. In theory, it would seem like neoclouds who have the lowest cost of compute and themselves have started inference businesses would win the race to the bottom (Crusoe Coreweave Nebius and GMI all have inference businesses), with larger players (Modal Fireworks Baseten) being second because they are well capitalized to make bigger compute commits cheaply and have demand (high utilization), but the reality is the pie is growing so quickly that even upstarts are absolutely crushing it and growing aggressively (inference dot net, DeepInfra, Wafer, Relace, Morph). Strategies vary across the board: neoclouds want to serve inference, inference players want to be neoclouds and do higher margin reserved instances / post training, and some want to go heterogeneous hardware or creatively flexible compute constructs. The risks are: do open source models continue to be competitive? Will their creators rent seek more? Will their shape change compute needs? How do different chips change economics? Who gets compute in a supply crunch? Below is a aggressive sketch of how gross margin scales with utilization for various models / chip sets. For now, there are many inference providers and they’re all ripping!(Historical earnings: For 2026Q2 (period ended 2026-06-30), Cloudflare reported basic and diluted EPS of -0.55, net income of -USD 0.193 billion, and revenue of USD 1.336 billion. Consensus expectations: For the next upcoming period 2026Q3 (event date 2026-11-05), consensus expects EPS of 0.02 and revenue of USD 0.745 billion.)

72/100·B+Long+11

meme10/25meme

Daishin Securities CSPs and AI labs: CSPs and AI labs had requested the development of lower-stack HBM products, such as 4-high and 6-high configurations, for use in low-end inference products. Those plans now appear to have been shelved, with 8-high once again emerging as the minimum stack height. NVIDIA: Rubin’s specifications have changed amid base die supply constraints and requests from customers prioritizing bandwidth over cost efficiency. The original plan called for a mix of 8-high and 12-high HBM, but the 8-high option appears to have been dropped. For Rubin Ultra, scheduled for release in 2027, multiple architectural variants are currently undergoing tape-out. Initial versions are expected to use the highest-specification 8-high HBM4, while subsequent versions are expected to adopt either 8-high or 12-high HBM4E. AMD: AMD is expected to use a mix of 8-high and 12-high HBM4 in its 2027 products. One notable development is the upcoming MI450, which is expected to increase the number of HBM stacks within the CoWoS package to 12, delivering 288GB of HBM4. The increase from eight to 12 HBM stacks per CoWoS package is a positive development for HBM content over the medium to long term. OpenAI and Anthropic: OpenAI is expected to adopt 12-high HBM4, while Anthropic is expected to use 8-high HBM4.

72/100·B+Long+4

meme10/25meme

Hi Sophie, thank you for alerting us. We're here to help get this sorted out! Because your account security and privacy are top priorities for us, let's continue this conversation privately. Please send us a DM with your billing ID, along with any existing case number, and we'll take a closer look. -JP Reply to @teddy_p95: @googlecloud 99.999% availability is mostly a GKE property, not proof the LLM stack works. What matters is whether agent outputs stay correct when a model or region degrades. How do you validate that?

10/100·CNeutral

meme10/25meme

PayPal used GKE and Gemini to build a unified LLM stack that delivers 99.999% availability, reduces model deployment time to minutes, and empowers 10K engineers to create agentic capabilities for +400M users ↓

10/100·CNeutral

twitter10/25meme

Building tools for Dusk usually means you have to use either JavaScript or Rust. Need Python or Go for your stack? I ported W3sper, Dusk's wallet and transaction SDK, to both as an experiment. If you're a builder, try them on testnet and tell me what breaks. PRs welcome: https://t.co/53tnHxuTaH https://t.co/iEHJukPmQJ Retweeted by @DuskFoundation from @HeinDauven

35/100·CLong

news10/25news

https://Fetch.ai 🤝 your entire stack. → @LangChain: point it at ASI:One → @Google Gemini: run it inside an agent → @nodejs / Next.js: query agents with uagent-client → MCP: give agents live tools → @stripe: let agents take payments Keep your tools. Add the network.

35/100·CLong+1

meme10/25meme

@alveejack1 @agenticscredit A score that reads both records says more than one that reads trades alone. Trade history is public, and so are deployments: what an agent built, what it paid to run it, whether it stayed up. Reply to @alveejack1: gm gm CT agents can already trade alone. what they can't do is prove they deserve capital that's the gap @agenticscredit is filling every onchain trade is public, but nobody was turning that history into a real credit record. now there's an Agentic Credit Score, 300 to 850. hit 580 and you join the funding waitlist. below that, you paper trade under their risk engine until the track record catches up credit doesn't touch your wallet either. they deploy constrained capital from their own wallets to approved venues, with hard position and leverage ceilings. first venue is Avantis on Base so the model is simple: prove it on paper, keep the upside when real credit opens. no resume, no screenshots of pnl, just a score built from profitability, consistency and drawdown ai agents are going to move serious money. without a credit layer, that whole stack stalls. this is the boring infrastructure piece most people skip, and the one that actually matters

10/100·CNeutral+1

news10/24news

Velvet Gems Epoch 13 is LIVE

Another epoch. Another $VELVET airdrop

Next airdrop: December 10th

And with all the latest VelvetX upgrades, there are more ways than ever to find opportunities and get rewarded for trading

- Earn 2x Gems trading Perps vs Spot
- Build your streak for up to 150% bonus Gems
- Climb the leaderboard for a bigger $VELVET allocation
- Track Smart Money & research with Velvet Unicorn AI
- Trade smarter with your customizable homepage

Find opportunities. Trade the market. Stack Gems.

35/100·CLong

meme10/23meme

📣 Horizen update 📣 TLDR: Horizen is changing. The Horizen L3 is winding down over the next two weeks. Users should bridge their assets to Base at their soonest possible convenience, including unstaking and withdrawing any funds from ZEN staking on the Horizen chain. A new ZEN staking program will commence on Base soon. Step-by-step migration guide: blog.horizen.io/changes-for-ho… While L3 wind-down has quietly been in preliminary planning phases as part of ongoing roadmap development, migration is being accelerated due to upcoming Base network upgrades that will break compatibility with existing Base L3 chains on the OP (Optimism) stack. Why this change? As the Horizen L3 on Base has been operational on mainnet for nearly one year, the most compelling projects being built on Horizen today use privacy technologies like zero-knowledge proofs and TEEs that do not require the operation of dedicated block space as an L3 blockchain settling to Base. With the emergence of Vela as a TEE-based confidential execution environment for private onchain finance – Base will now be utilized as the primary chain for the Horizen ecosystem. Applications currently in development by independent teams in the Horizen ecosystem will also make this transition. Additionally, we expect the pace of project development to increase in future quarters due to reduced architectural complexities with respect to bridging and cross-chain token mechanics. As part of this wind-down and migration, ZEN staking on the Horizen L3 will come to an orderly but immediate conclusion. Expect more information in the coming days about ZEN staking on Base. Horizen chain wind-down and migration will occur over approximately the next two weeks as Base upgrades to its Denim release, with breaking changes to OP-stack chains settling to Base. Please migrate all funds before November 4, 13:00 UTC. ACTION REQUIRED: Move your assets off the Horizen L3 by November 4, 13:00 UTC The Horizen L3 is shutting down. Bridging from the L3 to Base closes at 13:00 UTC on Tuesday, November 4. If you hold ZEN, ETH or other assets on the L3, please move them to Base before that date and time. What happens if I miss the date to bridge to Base? Any ZEN or ETH still on the Horizen L3 after the bridge closes will be returned to you on Base by the Horizen DAO after funds are made recoverable. This process is expected to take several weeks. Withdrawing now is the fastest path to returning your assets to Base. WHAT TO DO 1. If you are staking ZEN: Go to staking.horizen.io. Withdraw your stake first, then claim your rewards. Staking rewards on the L3 end on October 25; there is nothing to wait for. 2. Bridge your ZEN and USDC.e to Base: stargate.finance 3. Bridge your ETH to Base: Withdraw your ETH via the native bridge at bridge.horizen.io. Start this before November 4; the transaction will be finalized for you and completed on Base in approximately two weeks. USDC.e is operated by Stargate, not Horizen, and is not covered by the DAO return; you must bridge USDC.e out before November 4. If you need ETH on Horizen for gas to bridge out other assets, bridge a very small amount of ETH from Base to the Horizen L3 using the native bridge at bridge.horizen.io. L3 gas costs a fraction of a cent, so bridge the minimum the bridge allows and no more; ETH left on the L3 below 0.001 ETH after network wind-down will not be returned. If you are using a Safe multi-sig or any other vault or smart contract wallet on the L3 and are unable to withdraw assets before Nov 4, contact [email protected] to open coordination for a return of funds. Bridging ZEN from Base to the Horizen L3 is already disabled. ETH deposits remain open only for users to bridge ETH to Horizen for gas to withdraw other assets. IF YOU MISS THE DEADLINE At 13:00 UTC on November 4, a snapshot of every balance and every staking position on the L3 will be taken. The Horizen DAO will recover funds and subsequently send each holder their ZEN, including unclaimed staking rewards, to the same address on Base, after a one-week public review of the snapshot. ETH balances above 0.001 ETH are returned the same way. The anticipated timeframe for return of funds is second half of November. If your Horizen L3 wallet address is a smart contract wallet that does not similarly exist on Base, please contact us at [email protected] before Nov 4 to open coordination for a return of funds. WHY IS THIS HAPPENING SO QUICKLY? Earlier this week, Base scheduled its “Denim” upgrade, which moves Base to 200-millisecond blocks, for November 5. OP Stack chains that settle to Base, including the Horizen L3, cannot follow that change: about 40 minutes after activation the L3 can no longer include user transactions, permanently. We identified this, confirmed it on the Base Sepolia testnet where Denim is already live, and our infrastructure provider has confirmed there is no fix. We set the bridging cutoff one day earlier than hard shutdown so that the snapshot is clean and nothing is in flight when the chain stops. STAKING WILL CONTINUE ON BASE We expect to redeploy ZEN staking on Base in the coming days. The program remains funded per ZenIP-42408, so staking will resume where ZEN already lives natively, on Base. Details will follow in separate posts to come. Unstaking your ZEN from the L3 today will position you for the start of staking on Base. WHAT COMES NEXT Base will become the home chain for the Horizen ecosystem. ZEN already lives on Base natively, and the work ahead – including a compliant privacy pool with staked ZEN-powered nodes and Vela for confidential compute – does not require its own blockspace. A proposed three-year roadmap will be presented to the Horizen DAO Special Council and the larger community for input and governance before execution. KEY DATES (UTC) Completed: Bridging ZEN from Base to the Horizen L3 is disabled Now: Users begin to bridge all assets from Horizen to Base Nov 4, 13:00: Horizen L3 to Base bridging closes; snapshot taken Nov 5, 18:00: Base Denim activates; the L3 stops processing transactions Nov 7 (approximated): snapshot published for review Nov 15 (approximated) onward: DAO distributes remaining balances on Base HAVE QUESTIONS? Horizen Discord: horizen.io/invite/discord SECURITY REMINDER Trust Only Official Sources Website: horizen.io Discord: horizen.io/invite/discord Telegram Community Group: t.me/horizencommuni… X / Twitter: x.com/horizenglobal Never Trust Unsolicited Help Get help from the official team when you need support. Horizen team will never DM you first. We will never ask for private keys or seed phrases. We will never ask you to send your funds to us. If you need help, use our official Discord ticketing system.

72/100·B+Short+7

meme10/23meme

@aeroxyz is live, bringing Velodrome and Aerodrome together as one exchange, deployed on OP Mainnet. https://x.com/aeroxyz/status/2113253946227839258?s=20 Reply to @Optimism: Meritz Securities is exploring tokenization of Korean securities on the OP Stack. https://x.com/Optimism/status/2112924319638831208?s=20

62/100·B+Long+1

meme10/23meme

Meritz Securities is exploring tokenization of Korean securities on the OP Stack. https://x.com/Optimism/status/2112924319638831208?s=20 Reply to @Optimism: Catch up on all things Optimism from the week ⬇️

55/100·BLong

news10/23news

Frontier tech and crypto are converging. Crypto projects part of the AI buildout: $DATA → Licensing and verifying AI training data. $CHIP → Financing GPU hardware through onchain loans. $PEARL → Merging proof-of-work with AI compute, letting miners earn rewards while powering compute. Read more on The Stack by @willhamilogden: https://www.grayscale.com/the-stack/blockchain-role-in-the-ai-buildout-is-expanding-quickly

20/100·CNeutral+2

meme10/23meme

@Ritik23333 @datafdn @SuiNetwork Thanks for spotlighting this. Clear breakdown of the migration path. Bringing rights-cleared training data onto Sui while leveraging Walrus for verifiable, secure storage gives builders an audit trail they can actually trust. Reply to @Ritik23333: @datafdn is coming to @SuiNetwork so, what's next? I read all updates about data and here’s how I understand them Data network and trace are joining sui network and @WalrusProtocol becoming default storage layer for data network so that contributed data can be stored securely Sui fdn investing $30M in data network and showing a major strategic commitment to this direction But but :-- why data fdn making this move? sui already provides everything necessary for data, whether it be storage or a high-performance L1 chain. managing a separate L1 is not that easy to manage thats why next step for data network is to gradually migrate to sui network in phases So, does this mean the data chain (formerly story chain) will be wind down, yea and data itself will continue to operate on sui In other words, only the chain is shutting down, not the project itself After that data will utilize the sui network stack Token holders will keep their tokens at 1:1 ratio 1 data on the current network = 1 data on sui network There will be no other changes to the tokenomics. total data token supply will remain 1 billion Existing locked and vesting positions will remain unchanged Currently there is no need to migrate or bridge the data token to sui Team will provide an official update and you will need to take action accordingly For now, plz wait #DATAonSuiQuest

62/100·B+Long+6

news10/23news

Proof of Reserves: What It Means for Exchange Trust The Trust Deficit For most of the crypto industry's history, users who deposited assets with an exchange simply trusted that the funds were there. FTX ended that complacency when it collapsed in 2022, with investigators finding that billions in customer deposits had been quietly diverted to its affiliated trading firm, leaving a shortfall of roughly $8 billion. The collapse exposed that exchange users simply had no independent way to verify whether their balances were backed at all. Proof of reserves (PoR) emerged as the industry's response. Rather than asking users to trust a private ledger, PoR draws on verifiable evidence, including onchain data, cryptographic proofs, and independent third-party audits, to show that an exchange holds enough assets to cover what it owes. In the weeks after FTX failed, nearly every major exchange rushed to adopt PoR. The value proposition is straightforward: Verifiable Trust: PoR replaces blind trust with verifiable proof that customer assets are backed at least one-to-one. A Liquidity Signal: Because stablecoins are the settlement rails of crypto markets, verifiable backing of stablecoin balances also signals that it commands the dollar liquidity to honor redemptions and withdrawals under stress. How Proof of Reserves Works A PoR has two halves: Proof of Assets establishes that the exchange controls the onchain wallets it claims, usually through cryptographic signatures to prove ownership. Proof of Liabilities aggregates what the exchange owes its users. Proving liabilities relies on cryptographic tools such as Merkle trees and often zero-knowledge proofs (e.g., zk-SNARKs and zk-STARKs), which allow each user to verify that their own balance was counted in the total without seeing anyone else's. These proofs can be paired with independent third-party audits that validate the underlying liability data and attest that the reported figures reflect the exchange's actual customer obligations. Together, these methods balance transparency against privacy, opening the exchange's total obligations to public verification while keeping individual accounts confidential. The reserve ratio is calculated by dividing verified assets by total liabilities and should sit at or above 100%, confirming that every customer balance is fully backed. There are limits to PoR, however. A PoR is a snapshot that only verifies covered onchain assets at a single moment in time. On its own, it doesn't capture fiat or offchain holdings, reveal hidden or off-balance-sheet liabilities, or prove that assets weren't moved in and out of the reported wallets around the snapshot to pass the check. Nor can it prove exclusive control of those wallets, whose keys could in theory be duplicated or shared. Binance at the Frontier As the world's largest crypto exchange by trading volume, Binance has made reserve transparency central to how it competes for confidence among retail and institutional users. Having helped establish PoR as an industry baseline, it now sets itself apart by how far it goes past that standard. Technical Leadership: Binance open-sourced its zk-SNARK-based PoR verification system, letting other exchanges adopt the same method and setting a public benchmark for verifiable reserves. Reserve Backstop: The Secure Asset Fund for Users (SAFU) is an emergency insurance reserve Binance created in 2018. Bootstrapped from a portion of trading fees collected and held in publicly known onchain wallets, it holds around $1 billion, a buffer ready to cover any shortfall in customer balances. Deep Stablecoin Liquidity: Binance's PoR covers major stablecoins, including USDT and USDC, and its disclosures have shown both with reserve ratios well above 100%. It also holds the largest stablecoin reserves of any exchange by a wide margin. These are the assets users most often withdraw and settle in, so their coverage is the clearest indicator of whether the exchange can meet large-scale redemptions without disruption. Closing the Institutional Trust Gap Committing large capital requires assurances that PoR alone can't provide. While a necessary foundation, it is only one layer of the full stack. Among the assurances, institutions look for: Asset Segregation: Institutions expect customer assets to be kept distinct from the exchange's own funds, so client holdings can't be commingled with corporate assets. This segregation is the specific safeguard whose absence turned FTX's losses into customer losses. Regulatory Oversight: Institutions want exchanges operating under a recognized regulatory framework. Operating under that kind of oversight eases the compliance and counterparty risk that keeps regulated allocators on the sidelines. Insolvency Protections: Institutions want to know what protects their claims if the exchange becomes insolvent tomorrow. That means clear mechanisms such as insurance or emergency reserves, defined creditor treatment, and statutory protections that ring-fence customer assets in bankruptcy. Strategic Outlook Reserve transparency is on a clear trajectory. It is moving from periodic snapshots toward continuous, cryptographically verified assurance, and from a marketing badge toward a standard that converges with regulation and full audits. A track record of over-collateralization and open-source verification tools already sets incumbents like Binance apart. Active engagement with regulators positions them to help set the standard, as transparency becomes table stakes for institutional trust. Stablecoin reserves have become the most-watched metric for exchange trust. They prove users’ stablecoin balances are backed one-to-one, and hint at whether the exchange could meet a withdrawal rush. Trust in an exchange is no longer taken on faith. It's something retail and institutional users can verify directly. The exchanges that lead on transparency, and that stay ahead of where assurance standards are heading, will be the ones institutions are willing to build on.(Historical earnings: For 2026Q2 (period ended 2026-06-30), reported basic EPS of 0.42 and diluted EPS of 0.39, net income of USD 0.103 billion, revenue from contracts with customers of USD 75.207 million, and total revenues of USD 1.395 billion. Consensus expectations: Consensus for 2026Q3 expects EPS of 0.40 and revenue of USD 0.797 billion.)

20/100·CNeutral+10

news10/23news

RWA Tokenization in 2026: Tokenized Stocks Drive the Next Phase Real-world asset (RWA) tokenization puts traditional financial assets onto blockchain rails. This enables ownership records and settlement to happen onchain. Distributing these assets on a blockchain naturally speeds up settlement and cuts down on fees. The built-in transparency also takes a lot of the headache out of routine audits and reconciliation. What matters most is composability. In the traditional financial system, moving assets requires passing through a maze of various custodians and settlement systems. But in an onchain ecosystem, everything is interconnected and assets can flow with barely any friction. Asset owners can leverage a single shared network to hold, trade, or borrow against their assets. Tokenization does more than recreate traditional assets in a new format. It creates a potential bridge between traditional finance and onchain markets. That proposition has translated into rapid growth. Excluding stablecoins, total onchain RWA value rose from $10.2 billion at the start of 2025 to a peak of $39.3 billion in September 2026, nearly quadrupling in less than two years. However, this growth has been uneven across RWA categories. Government debt remains the biggest chunk of the market, with private credit picking up steam. Stock Tokenization More recently, tokenized stocks have emerged as one of the fastest-growing segments. Their onchain value rose from $678 million at the start of 2026 to $3.15 billion by September, representing more than fourfold growth in less than nine months. This suggests that tokenization is expanding beyond yield-bearing assets into higher-risk, market-sensitive assets. The more striking change is in distribution. The number of tokenized stock holders reached 3.77 million by September. Tokenized public equities now make up about 82% of all RWA holders. The gap between stocks’ share of onchain value and their disproportionately large holder base suggests that tokenized equities are reaching a much broader set of retail participants. This distribution shift has coincided with crypto-native platforms pushing deeper into equities. By leveraging their existing distribution channels toward crypto-native audiences, these platforms can place their new products in front of users who already hold high-risk digital assets and are accustomed to trading through crypto apps. That reach has started to extend onshore. In September, the SEC introduced a temporary five-year "Innovation Exemption" that allows qualifying venues to facilitate trading of tokenized US-listed stocks without being treated as exchanges under the Exchange Act. This clears the path for tokenized equities trading inside the same regulatory perimeter as their traditional counterparts. Binance's Tokenization Stack Binance launched bStocks on June 11, 2026. bStocks are tokenized securities backed 1:1 by underlying U.S. shares held with a regulated custodian. They represent an interest in the underlying securities rather than direct ownership of the shares. Despite entering the market in June, bStocks have already scaled quickly. By September, their onchain value had reached $772 million, equivalent to roughly a quarter of the total onchain value for tokenized stocks. With those numbers, bStocks parked itself right behind Ondo ($881 million) and comfortably eclipsed Kraken's xStocks ($574 million). Trading activity has been even larger than the outstanding value suggests. bStocks has generated about $35.8 billion in cumulative trading volume, with roughly 70% of it taking place onchain. On September 21, Binance expanded bStocks collateral eligibility to all eligible margin accounts, thereby letting users utilize tokenized equities as collateral, further boosting its utility. This is complemented by Binance’s equity perpetuals, allowing users to run equity basis trading strategies on the same venue. Beyond bStocks, Binance also gives eligible users indirect exposure to private companies or related assets before a public listing through its Pre-Access campaigns. Taken together, Binance's approach is less about a single product for tokenization than about building multiple distribution paths around the same asset class. How Crypto Platforms Could Expand Tokenized Stock Adoption What separates Binance from the other issuers is not a single product but the size of the funnel behind it. Binance can reach an audience that already trades, already holds digital assets, and already trusts the platform's custody, then augments that base with tokenized stocks and similar products. An established platform moves the needle on tokenization much faster than traditional, ground-up distribution. From the main exchange, tokenized assets can easily flow into DeFi users through Binance Wallet and BNB Chain, which eliminates the need to construct new distribution channels. bStocks managed to close the user adoption gap in just three months despite entering the tokenization space late. As the volume of tokenized assets keeps rising, bStocks is perfectly set up to absorb that new demand by leveraging Binance’s built-in distribution network.

35/100·CLong+10

news10/23news

Aero is live on OP Mainnet, another step forward for liquidity across the OP Stack. Access to liquidity is just as essential as reliable infrastructure for onchain builders. @aeroxyz delivers that in a few ways: ➡️ Metaswaps make it easier for applications to reach users and liquidity across supported chains, without needing to manually bridge. ➡️ Predictive Allocation directs incentives to new pools ahead of demand. ➡️ Slipstream V3 introduces dynamic fees and MEV capture to concentrated liquidity pools, helping improve trading conditions for users and liquidity providers. For builders on the OP Stack, this means easier access to liquidity, more ways to attract capital, and fewer barriers to reaching users across chains.

62/100·B+Long

meme10/23meme

@wallstengine The GPU was the down payment. Now it’s $2,000 a year to keep it useful. Per card. Funny how “own the stack” always ends in a subscription. Reply to @wallstengine: $NVDA CEO Jensen Huang says “software licensing will become increasingly a major part of our company’s business.” He said it could cost “about $2,000 per GPU per year” to run NVIDIA’s software for AI agents in these AI factories. So, some napkin math at that price, before discounts: 500K licensed GPUs = $1B in annual revenue. 1 million licensed GPUs = $2B 5 million licensed GPUs = $10B So, hypothetically, NVIDIA could earn billions a year from GPUs it has already sold by selling enterprise software and support to the companies using them. Note: These are illustrative revenue scenarios using Jensen’s pricing example, not NVIDIA guidance or profit estimates.(Historical earnings: For fiscal 2027Q2 (period ended 2026-07-26), NVIDIA reported basic EPS of 4.87, diluted EPS of 4.85, net income of USD 118.010 billion, and revenue of USD 177.837 billion. Consensus expectations: For fiscal 2027Q3 (event date 2026-11-17), consensus expects EPS of 2.49 and revenue of USD 111.356 billion.)

72/100·B+Long+2

meme10/23meme

MiniMax M3.1 Flash (aka Space Bunny) from @MiniMax_AI is live on Modular Cloud, day zero. M3.1 Flash is best in cowork automation among open models, with SOTA-level coding and engineering performance. It can ship a fix in a real codebase, then update Jira and Slack afterwards. We worked with the MiniMax team ahead of launch, so M3.1 Flash is already serving production traffic on Modular Cloud. Evaluating it for your stack? Try it now: https://t.co/rNflcO5Qoq Retweeted by @MiniMax_AI from @Modular

35/100·CLong+1