Suma Begum (@Sumabe94) header
Suma Begum (@Sumabe94) profile picture

Access Role Holder of @ConcreteXYZ Contributor of @fhenix

Joined March 2025
Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)Photo by Suma Begum (@Sumabe94)
Suma Begum retweeted
Fhenix: The Cryptography, the Researchers, and the Capital Behind Confidential Computing Most blockchain projects talk about privacy as a feature. Fhenix is approaching it as a fundamental computing problem. The idea is simple, but the engineering behind it is not: how can a smart contract process sensitive information without ever exposing the information itself? That question is driving Fhenix’s research in Fully Homomorphic Encryption (FHE), and two contributions particularly stand out. 1/ Threshold FHE Decryption @fhenix researchers Guy Zyskind, Doron Zarchy, Max Leibovich and Chris Peikert published High-Throughput Universally Composable Threshold FHE Decryption, accepted at ACM CCS 2025. Traditional encrypted computation still needs a secure way to reveal authorized results. Threshold decryption distributes that capability across multiple participants, so the system doesn't have to depend on one party holding the entire decryption key. Their research tackles the performance overhead of earlier threshold FHE approaches, replacing noise-flooding techniques with a more efficient MPC-based approach while targeting strong composable security. The paper received the Distinguished Paper Award at ACM CCS 2025, making it a notable research milestone for the team. 2/ DBFV: Making Encrypted Arithmetic More Practical Fhenix is also advancing DBFV (Decomposed BFV), focused on improving exact, high-precision encrypted arithmetic. Why does that matter? Confidential finance cannot always rely on approximate calculations. Balances, financial logic, settlement and smart contract conditions may require exact arithmetic while sensitive inputs remain encrypted. DBFV aims to make these computations more practical by improving the efficiency of encrypted arithmetic. Together, these research directions address two important challenges in FHE: • More efficient computation on encrypted data • Faster, secure and distributed decryption when revealing results is authorized 3/ The $22M Funding Foundation Research at this level requires more than an idea. It requires cryptographers, engineering, infrastructure, time and capital. Fhenix raised: • $7M Seed — September 2023 Co-led by Multicoin Capital and Collider Ventures, with participation from Node Capital, Robot Ventures, Bankless, Hack VC, Tané Labs, Metaplanet and others. • $15M Series A — June 2024 Led by Hack VC, with Amber Group, Collider Ventures, Primitive Ventures, GSR, Stake Capital and Dao5 participating. That brings its publicly announced Seed and Series A funding to $22M. 4/ Strategic Backing Beyond Venture Capital Fhenix's backing goes beyond conventional funding rounds. Tandem, the venture studio and investment arm of Offchain Labs, brings a strategic connection to the Arbitrum ecosystem and confidential computation. BIPROGY and Translink Capital have also been associated with strategic investment, connecting Fhenix's technology ambitions with enterprise technology and broader institutional networks. These relationships matter not simply because of the names involved, but because confidential computation needs an ecosystem capable of supporting real applications, integrations and users. 5/ What I'm Watching Research is impressive. Funding is meaningful. But the real test is what happens when cryptography moves from papers into production. Can Fhenix make encrypted computation faster, easier for developers and useful for everyday on-chain applications? Can it progressively decentralize its trust assumptions? Can confidential DeFi, private payments, encrypted trading and institutional use cases attract real demand? That's the journey I find interesting about Fhenix. Not just making data private, but building the infrastructure that allows applications to compute on secrets without forcing users to reveal them. Privacy should not mean sacrificing utility. It should become part of how blockchain works. That is the vision worth watching. @deivonchain
🤖 Made with AI
19
1
23
253
Suma Begum retweeted
Privacy is not about making financial activity invisible. It's about controlling who can see what, and when. Institutions entering on-chain finance need more than encrypted transactions. They need to execute strategies, manage capital, generate yield, and maintain accountability without exposing sensitive financial information to everyone. This is where @primus_labs takes an interesting approach. With Fully Homomorphic Encryption (FHE), smart contracts can process encrypted values without revealing the underlying numbers. Balances and transaction amounts remain confidential while the system continues performing financial operations. But the real challenge goes beyond encryption. Institutions still need a way to verify specific information, meet compliance requirements, and grant access when necessary. Permanent secrecy isn't practical, and permanent transparency isn't acceptable either. Primus addresses this gap by introducing controlled access to encrypted states, making disclosure something permissioned and auditable rather than automatic. That distinction matters. The future of institutional DeFi may not be fully transparent or completely hidden. It may be privacy by default, with verifiable accountability when authorized. That's the kind of infrastructure real capital needs. Join Here s.kaito.ai/2trAVAx Let's go together. @xxiang_xie @bitvickyzh #PrimusLabs #FHE #DeFi #Privacy #Web3
🤖 Made with AI
22
5
28
364
Suma Begum retweeted
The best traders are the ones who survive long enough to become profitable. If you're a trader and you find yourself forcing entries or trying to catch a falling knife in this brutal market, you have no survival instinct. And the only winners in this game are people who survive long enough.✌️
15
4
26
313
Suma Begum retweeted
最終面接で社長に「趣味の銭湯めぐり、何軒行きました?」と聞かれて、「312軒です」と答えたら、面接が15分延びた 志望動機の話はほぼなし どこのお湯が熱いか、薪で沸かしてるか、そんな話ばかり 雑談だけで終わったし、これは落ちたと思ってた 1週間後、内定の電話のあとに社長がひと言 「入社のとき、312軒の一覧を持ってきてください」
1
1
9
3,777
Suma Begum retweeted
gm btw, almost forgot its saturday xd not gonna discuss the market today. those who survived can see whats happening themselves and for those who didnt... rest in peace anyway, got some amb and creator opportunities for u to check over the weekend filled out a few myself, skipped some others. just a matter of taste i guess • @Flop_Labs - creator / KOL applications docs.google.com/forms/d/e/1F… • @entropyIO - creator program, roles from Voices to Genesis. first deadline is today discord.gg/entropyio • @vistaralabs - ambassador program, monthly rewards and team access. DM the team • @Octos_official - ambassador program, monthly salary + allocation xitter.cf/Octos_official/status/… • @Elyon_Chain - creator program, requires 2k+ followers airdrop.elyon.network/creato… • @sovrentech - ambassador program, $3k monthly reward pool node.sovrentech.io/ambassado… • @uxuycom - ambassador campaign, $1k reward pool xitter.cf/uxuycom/status/2120434… thats pretty much it for today maybe u will find smth interesting here. good luck m8s, and enjoy ur weekend
UXUY Ambassador Program is live now! We’re looking for creators, traders and community builders from around the world to share local perspectives, create original content and connect UXUY with their communities. Participants will also have the opportunity to receive: · $1,000 USD reward pool · Official amplification of standout content · Direct communication with the UXUY team · Priority discussions about future ambassador roles and regional collaborations How to join 1. Quote this post and tell us how you’d like to contribute to UXUY and connect with your local community. 2. Complete the Google Form and include the link to your quote post. Apply here: forms.gle/iPjTCeVikBWghCas7
2
2
15
184
Suma Begum retweeted
JUST IN: Bitcoin falls under $57,000 for the first time in over 2 years. $385 million liquidated from the crypto market in the past 5 minutes.
1
1
9
144
Suma Begum retweeted
DAC's quantum security story has another important question: How do you securely deliver the keys to the people protecting the network? DAC's Entropy-Padded Transaction (EPT) design aims to protect transaction information during broadcast using entropy-based encryption. But there is another side to this process. Validators still need the correct cryptographic material to process and verify protected transactions. And that creates an interesting challenge. How does the network distribute this sensitive material without exposing it to unauthorized parties? What happens if a validator or communication channel is compromised? How can the network coordinate delivery across its different layers without introducing delays or security weaknesses? These questions matter because even strong encryption can be undermined if the material protecting it is exposed during distribution. DAC's published technical material outlines its protected transaction architecture, while further hardening of validator communication and entropy distribution remains an important area to watch. I don't see this as a reason to dismiss the design. I see it as one of the engineering challenges that deserves serious attention as the architecture develops. Generating secure entropy is one achievement. Distributing it securely, consistently, and at network scale is another. That is where the real strength of DAC's design will also need to be demonstrated. @dac_chain
🤝 Paid partnership
9
4
26
166
Suma Begum retweeted
The hidden cost of AI isn't always compute. It's the time GPUs spend waiting. Training data must be staged, batches prepared, and checkpoints synchronized. When storage or network throughput falls behind, expensive accelerators sit idle instead of learning. Adding more GPUs won't fix a data pipeline that cannot feed them efficiently. This is where @shelbyserves has an interesting thesis: make data globally accessible, verifiable, and easier to move across compute environments. AI scalability isn't just about faster chips. It's about keeping every chip productive. @shelbyserves @affinity_matrix
17
5
33
418
Suma Begum retweeted
The real test of privacy tech is what it can protect in a live financial application. @FlutonIO is now live on Arbitrum, using @fhenix CoFHE to bring confidential payments to existing wallets. Users can shield USDC or ETH, transfer value inside the pool without publicly exposing amounts, recipients, or token types, then withdraw to an address of their choice. What makes the architecture interesting is the combination of FHE and ZK. FHE handles encrypted reserve balances and computation, while zero-knowledge proofs validate ownership and transfers without revealing the underlying details. CoFHE also confirms encrypted deposit and payout conditions, helping the pool track its obligations without exposing its books. Deposits and withdrawals remain visible, so privacy isn't absolute. But confidential activity inside the pool is a meaningful step forward. This is what makes Fhenix worth watching: privacy is moving from cryptographic research into real financial infrastructure. @deivonchain @Asterix54907294 @0xkeksgiven
31
11
49
350
Suma Begum retweeted
The most interesting thing about @primus_labs isn't just what it can keep private, but how it's working to make privacy computationally practical and easier to integrate. Looking deeper at its architecture reveals a different side of Primus. 1/ Privacy needs performance, not just cryptography. Homomorphic encryption allows computations on encrypted data, but the underlying operations can be computationally expensive. Primus' GPU acceleration architecture targets this challenge at multiple levels: • Primitive-level arithmetic, including multi-word and finite-field operations • Polynomial computation through CRT and NTT transformations • Kernel optimization, fusion, and memory management • Device-level operations such as encrypted multiplication, rotation, addition, and key-switching Optimizing these layers together is important because the real bottleneck isn't always the encryption itself. It can be the cost of everything happening underneath it. 2/ zkTLS is becoming easier for developers to integrate. The Extension Core SDK is particularly interesting because developers can use Primus' zkTLS algorithms inside their own browser extensions without depending on the Primus Extension. They can configure web requests, select specific response fields, initiate attestations, verify Primus signatures, and use the verified results in their own application logic. That creates a bridge between existing web data and applications that need verifiable information rather than blind trust. 3/ Verification and computation serve different roles. zkTLS helps establish whether web data is authentic and verifiable. FHE makes it possible to compute over encrypted information without exposing the underlying plaintext. These technologies address different problems, but together they point toward a more powerful vision: data can be verified at its source, protected when sensitive, and made useful without unnecessary exposure. 4/ The ecosystem is part of the bigger picture. Primus' partner network and developer tooling suggest an approach focused on making cryptographic infrastructure accessible across different applications, rather than keeping it isolated within one product. The real progress to watch isn't simply another privacy feature. It's whether Primus can turn complex cryptography into infrastructure developers can integrate, users can trust, and applications can use at scale. Privacy should not be an obstacle to building. It should be part of what makes better applications possible. 🧡 Join Here s.kaito.ai/2trAVAx Let's go together. @xxiang_xie @bitvickyzh #PrimusLabs #FHE #DeFi #Privacy #Web3
🤖 Made with AI
18
5
33
387
Suma Begum retweeted
Privacy in blockchain isn't just about hiding information. It's about giving users control over what they reveal. That's what makes @Fhenix interesting. With Fluton now live on Arbitrum, users can shield USDC or ETH into a privacy pool and make private transfers without revealing the amount, recipient, or token on-chain. The distinction matters: deposits and withdrawals remain publicly visible, while transfers inside the pool keep those details confidential. Powered by Fhenix CoFHE for encrypted reserve management, alongside zero-knowledge proofs for private transfers. A meaningful step toward more practical financial privacy. @Asterix54907294 @deivonchain
🤝 Paid partnership · 🤖 Made with AI
1
1
4
25
Suma Begum retweeted
Privacy matters, but verifying computation is equally important. Traditional FHE enables computations on encrypted data, keeping sensitive information private. However, ensuring computational integrity remains a major challenge. This is where @primus_labs brings an interesting approach with zkFHE. By combining encrypted computation with cryptographic proofs, it aims to make computation verifiable without exposing sensitive inputs. The HasteBoots research takes another step forward by making TFHE programmable bootstrapping proofs more efficient. Privacy, integrity, and verifiability are essential for the future of secure computation. It's not just about keeping data private. It's about proving that computations are correct. And @primus_labs is exploring how to bring these capabilities together. Let's move toward a more private and verifiable digital future. @xxiang_xie @bitvickyzh #PrimusLabs #FHE #zkFHE #Privacy #DeFi #Web3
6
2
8
86
Suma Begum retweeted
Gfhenix Fhenix is proving that FHE is ready for real world use. Speed was never the real limit. @fhenix
Speed was never going to be the thing that stopped FHE @GuyZys on when Fhenix knew it had moved past the bottleneck
4
1
12
93
Suma Begum retweeted
Gfluton @FlutonIO one pool. one anonymity set. many tokens. a shared tree, strict per token accounting. private transfers keep the token unnamed.
1
1
11
60
Suma Begum retweeted
Just applied for the @meetmotes allowlist 👀🔥 Motes is building NFT-native AI agents on robinhood chain, where each Genesis NFT comes with its own AI agent and smart account Ending in few hours, apply now 👇 meetmotes.com/apply
32
11
47
392
Suma Begum retweeted
Storage is only the beginning. Access is what matters,Data becomes truly valuable when it can be accessed quickly, verified reliably, and used wherever computation happens.That's the vision behind @shelbyserves . Shelby is exploring a decentralized hot-storage infrastructure designed to make frequently accessed data faster, more reliable, and verifiable. Its approach highlights an important shift: - Faster data retrieval for demanding applications. - Verifiable storage through cryptographic auditing. - Erasure coding for data reliability. - Infrastructure connecting storage with different computing environments. - Incentives aligned with actual data usage. What makes Shelby interesting is its focus on more than simply storing information. It's about making data accessible, dependable, and useful when applications need it most. The future isn't just about where data lives. It's about how efficiently we can access and trust it. Building a better foundation for the next generation of decentralized applications. @affinity_matrix
10
10
20
166
Suma Begum retweeted
Proof of Reserves Is More Than Publishing Numbers Transparency in Web3 shouldn't depend on promises. Publishing reserve numbers is one thing. Proving those numbers are authentic, accurate, and backed by verifiable data is another. That's where @primus_labs brings an interesting approach with its Proof of Reserves infrastructure. By combining zkTLS, Trusted Execution Environments (TEE), and zkVM technology, Primus aims to make reserve verification more transparent while protecting sensitive financial information. Why does this matter? → Verify financial data from authentic sources. → Protect sensitive institutional information. → Generate cryptographic proofs of computation. → Make reserve information independently verifiable. The bigger vision is simple: institutions shouldn't need to expose their entire financial infrastructure to demonstrate accountability. **Transparency and privacy can work together.** That's an important step toward building greater trust in Web3 finance. #PrimusLabs #ProofOfReserves #zkTLS
9
5
19
217
Suma Begum retweeted
Fhenix is expanding its mainnet ecosystem The integration of @0xAlphaEngine brings another step toward confidential onchain finance. Privacy isn't just a feature; it's the foundation for the next generation of DeFi. Exciting to see the Fhenix ecosystem growing @fhenix
AlphaEngine (@0xAlphaEngine) joins fhenix mainnet.
4
4
17
129
Suma Begum retweeted
Today, we're Introducing Fluton. Private transfers. Encrypted balances. Your existing wallet. Powered by FHE & ZK. Mainnet is now live on @arbitrum.
103
87
140
288
31,680
Suma Begum retweeted
AutoHoG: Can Primus Make Fully Homomorphic Encryption Fast Enough for DeFi? Fully Homomorphic Encryption (FHE) has enormous potential, but one challenge continues to limit its adoption: computation speed. Imagine sending tokens, calculating balances, or executing financial operations without revealing sensitive information. The technology makes this possible, but encrypted computation can be computationally expensive. This is where Primus introduces an interesting innovation: AutoHoG. Traditional FHE systems often require expensive cryptographic operations, particularly bootstrapping. These operations can create performance bottlenecks when applications need frequent encrypted computations. AutoHoG takes a different approach. Instead of relying entirely on predefined homomorphic operations, it focuses on automatically designing and optimizing custom homomorphic gates. Its approach includes: * Automatic gate generation: Creates specialized gates for encrypted operations. * Circuit optimization: Combines compatible operations to reduce unnecessary computation. * Bootstrapping reduction: Minimizes expensive operations where possible. * GPU acceleration: Uses parallel computing to improve execution performance. * Deterministic computation: Targets reproducible results across executions. Why does this matter for DeFi? Consider a confidential financial application where users want to transfer assets, calculate balances, or execute transactions without exposing their financial information. If encrypted computation remains too slow, developers face practical limitations. But improving FHE performance could make confidential financial applications more practical. Primus reported impressive AutoHoG benchmarks using eight NVIDIA RTX 5090 GPUs: * 90 ms for a 64-bit confidential transfer. * More than 210 transfers per second. These are project-reported benchmarks, not independently verified production-network performance. The bigger picture is important. FHE is not just about hiding information. Its real potential is enabling applications to process sensitive information without revealing it. AutoHoG represents an interesting step toward making that vision more practical. Primus is exploring how privacy and computational performance can work together to unlock the next generation of confidential DeFi. s.kaito.ai/tQrhnub #Primus #FHE #DeFi #Privacy #Web3
🤝 Paid partnership
9
7
22
235