ARCIUM - THE ENCRYPTED SUPERCOMPUTER
The Lag of Innovation: The Three Waves of Acceleration
In crypto, the question always comes up: "Where’s the next big innovation? What’s the killer use case?" You hear it at every conference—conversations full of vague, noncommittal answers. But the truth is, we’re not stuck. Some things just take time.
Vitalik outlined some while ago three key areas for mass adoption, and we’re making major progress across all fronts:
Scaling: The core scaling primitives are largely solved. The industry is converging on a similar design pattern in which everyone is iterating on an IBRL-inspired mental model. We’re already seeing new applications unlocked with
@megaeth . Fat sequencers, which optimize latency to directly enable CEX-competitive order book DEXs. Fully on-chain games with acceptable tick rates could be the next to follow. These advancements are big, but scaling alone doesn’t create the next step-function in adoption.
UX: Innovations like
@infinex_app are redefining how users interact with blockchains. A gas-free interface lets users navigate assets seamlessly across ecosystems, eliminating the need for multiple blockchain plug-in wallets. From a user’s perspective, does ecosystem lock-in even exist anymore? Crypto increasingly feels like a unified state layer rather than fragmented networks. The modular vs. monolithic debate doesn’t matter to users—it’s just "chains and bridges," a UX mess that needs bundling. We’ve finally reached the point where products can be distributed without requiring users to navigate complex infrastructure themselves.
Privacy: While scaling and UX have laid the groundwork for mass adoption, privacy remains the missing piece—especially for businesses. Privacy isn’t just about hiding transactions; it’s about enabling enterprises to operate on-chain without exposing sensitive strategies. It’s the key to unlocking entirely new asset classes and applications.
With
@arcium , that next wave is finally here. (Our portfolio company—NFA, but I’ll be shilling.)
Arcium and The History of Privacy
Privacy in crypto has historically involved balancing strong anonymity with the need for broader financial integration. Early projects like Monero and Zcash established the foundation with anonymous transactions, yet their use cases were tied to a volatile L1 asset, too limited for widespread adoption beyond the darkweb. As the need for privacy grew, tools like Tornado Cash introduced mixers and shielded pools to obscure transaction trails, yet some operated on the fringes of the larger crypto landscape and often encountered regulatory challenges.
At the same time that Tornado Cash was sanctioned we underwrote Elusiv/Arcium on the premise that on-chain privacy is still essential for a financial system. However, their pitch has always been that the privacy space is far too idealistic, and that true confidentiality must be balanced with regulatory compliance. After the Hangover in 2021, this became one of the key insights that Elusiv provided, guiding our decision. They are a team that wants to accelerate blockchain adoption and build within the privacy space, yet they are pragmatic enough to understand how to deliver on their vision and ship a product that aligns with social norms. It takes a special kind of founder team to double down on this vision in such a time though. Especially when you hear that developers are going to jail for shipping open source code. Without talking those teams down they made it clear to me that a world where everyone uses Tornado Cash or Monero does not represent a bright future we should optimize for and leads to a future where we don't have privacy at all. With the recent U.S. Treasury decision to lift sanctions on Tornado Cash though, privacy-focused founders now appear to have some breathing room again. There couldn't be a better moment for a privacy renaissance.
In addition to compliance options to ship a socially acceptable and scalable product, the other important angles to us are friction- utility of privacy and usability with the current user/liquidity set. As long as people do not feel compelled by fear, they will not adopt privacy solutions on their own if these impose a poorer experience or higher friction. For example, consider how many of you use DuckDuckGo compared to Google Chrome. Ultimately, we must continually work to bring privacy to market while adding the least amount of lag in terms of latency and cost overhead compared to the current industry standard.
While I certainly believe that we are going into a more and more abstracted world where we have more or less different states that we can all access in a bundled way, we can't build an isolated privacy chain and then expect to build a big ecosystem of privacy applications around it. We already have chains that have many users and significant liquidity like Solana, and we must put privacy where the users/funds are, not building in isolation next to them in the hope that users and funds come.
Since we invested in 2022, we have seen several new approaches emerge, including zk-based privacy-focused chains like Aleo, MPC-based architectures such as Arcium or Zama which are pursuing the fully homomorphic encryption (FHE) route.
Here’s my ultra short TLDR to why Multi-Party Computation (MPC) for now is the right approach:
- Zero-Knowledge Proofs (ZKPs) are excellent for verification but not for computing on encrypted data in a shared state setting, since they are cumbersome for efficiently modeling and verifying evolving shared state.
- Fully Homomorphic Encryption (FHE) is too slow and costly for practical use— it may become viable for Arcium in the future.
- Trusted Execution Environments (TEEs) suffer from side-channel vulnerabilities, but there are ongoing security improvements.
- Multi-Party-Computation (MPC) is significantly faster and more cost-effective than FHE while providing stronger security assurances than TEEs.
For anyone wanting to argue about technical infrastructure, there’s only one question that matters in the coming months/years: "Do you even encrypt, bro?”
If the answer to this is no and you are waiting for innovation to pick up before encrypting, you’re already behind.
What the “added utilities” privacy Arcium Enables? Finding a Wedge in Blockchain
Every post about privacy can't miss talking about the adoption story of VPNs. Much of this growth is fueled by the added utility of VPNs, such as bypassing geo-restrictions for streaming services, which appeals to users more than privacy features alone. Techwise blockchain startups aren't the first to work on MPC or FHE solutions for privacy. Inpher has been pursuing confidential computing since 2015, with a total of $25M in funding. However, due to their non-blockchain positioning, they may have been too early to the market and struggled to find a strong wedge. Despite this, as a technology leader, they accumulated patents and expertise, which Arcium can now leverage through their acquisition of them. As a result, the Arcium team is stacked with PhDs who have been working on this for a decade.
If you combine the knowledge with an actively used shared state (initially Solana) and bring a compliant solution to the market that is also performant, you already have a very strong market position. But what are the “added” utilities that make us use Arcium?
Businesses On-chain
Let’s start with “DAOs” which is the biggest psyop in crypto. Most DAOs are essentially for-profit businesses, yet they operate in full transparency, making them easy targets for adversaries. Take ConstitutionDAO. Its failure highlighted the structural weakness of public blockchain organizations. Without privacy, on-chain businesses will always be vulnerable. Right now, the "privacy layer" for blockchain startups is keeping their balance sheets off-chain.
If we want to see real economic activity on-chain and truly build the Internet Financial System (IFS), we need to give companies privacy, starting with basic treasury management like encrypted transfers, to be able to pay employees without the whole company and world knowing their salaries. If you have to decide something with token holder voting and, for example, have to hold a minimum threshold to see the result, it is possible to hold a vote on strategic decisions such as an M&A deal that only token holders with X% of the voting power can see.
In the end, if we as funds and startups are not using our own rails to run our companies, who will?
Confidential Tokens:
Confidential Fungible Tokens
Confidential tokens enhance privacy while ensuring compliance. Unlike typical tokens, which expose balances and transaction details, confidential tokens use cryptographic techniques to conceal this data while remaining verifiable. This is crucial for tradFi institutions and RWA tokenization, where both privacy and regulatory transparency can be required.
Confidential Non-Fungible Tokens
I am a collector of on-chain assets. Whether from artists like Refik Anadol and XCOPY or game assets like Dark Forest Planets, I’ve often thought about the "right-click save" meme.
In the crypto space, we often have a misguided perspective on scarcity, largely influenced by Bitcoin. Even in the collector market, we tend to think of scarcity in terms of token supply, whereas I’d argue that owning an XCOPY piece that comes directly from the artist, one that only he and I have seen before, would hold immense value. While there is scarcity from a dilution standpoint with the total number of artworks, there is also "attention scarcity" in the unique opportunity to view something exclusive. In this regard, the "right-click save" meme might not be entirely wrong. In an environment where anyone can generate and consume virtually anything, the one thing that remains unobtainable which I would call “Attention Scarcity” feels highly desirable.
Therefore Privacy + NFTs is an immense unlock on the consumer and high-end art market within crypto. Secret Network previewed this in 2021 but failed to align an ecosystem around it. Still, their primitives revealed new user behaviors—like Quentin Tarantino’s Pulp Fiction NFT drop, which made it clear: if I’m not the only one who can see my purchase, I’m just paying for a glorified movie release.
Dark Pools and other Private Marketplaces
Institutional capital won’t fully embrace DeFi as long as on-chain markets remain too transparent. Arcium enables private order matching, shielding trade intent and preventing front-running. The result? True dark pools for DeFi and high-frequency trading without manipulation. A Darkpool built with Arcium therefore allows participants to trade privately without revealing sensitive order details. This preserves transaction confidentiality and prevents issues like market manipulation and front-running, while also supporting other use cases such as large block trades and off-exchange liquidity.
While token exchanges are the first example that comes to mind, the concept of keeping my wallet hidden or maintaining private bids and asks can be applied to all marketplaces. We can also explore new auction and sale mechanisms using this approach
Beyond Blockchains: Secure AI & Data Collaboration
Secure AI Model Training
Imagine two pharmaceutical companies working on AI-driven drug discovery. They want to train a model on their combined datasets but can’t expose sensitive research to each other. With Arcium, they can securely train AI models on encrypted data without ever revealing their proprietary information.
Collaborative Data Analytics
Consider multiple banks trying to detect fraud patterns. Normally, they wouldn’t share raw transaction data due to competitive and regulatory concerns. Arcium enables them to analyze encrypted transaction records together, spotting suspicious activity without exposing individual transactions. The same applies to healthcare providers collaborating on patient data while maintaining compliance with HIPAA and other regulations.
Internally, we have a joke that there are hardly any "chat outcomes" from German founders because, once, when we were discussing a deal with another VC, they made the statement:
"Typically not the most chad outcomes from German founders."
@yrschrade,
@julian_arcium,
@cryptopapi997,
@manzgoeggel
are the outliers
<encrypted>