ICYMI — Aethir Twitter Space Recap

On March 20th, Aethir launched their Checker Node Sale that would go on to sell over  65,000 nodes at a record-breaking 28,000 ETH.

As proud partners of Aethir, we sat down with Dan Wang, Co-Founder and Chief  Executive Officer of Aethir, Ilya Abugov of Sanctor Capital (also investors in Aethir), Garlam Won, and Virtual Bacon of Momentum6 in an Impossible Spaces session to chat all things Aethir and the node sale.

For those who missed the session, fret not. Listen to the recording here, or check out our written recap below!

TLDR

• Aethir's Origin and Focus: Co-founded by Dan Wang, with extensive experience in gaming infrastructure, Aethir aims to revolutionise cloud infrastructure by leveraging decentralisation to address latency and real-time scenarios.
• Web3 Intersection: Aethir intersects with Web3, AI, and gaming industries, foreseeing the potential of blockchain technology to transform gaming and AI sectors, making content more accessible and breaking hardware limitations.
• Partnership Highlights: Aethir's partnerships with leading enterprises like Well-Link and major telcos demonstrate its enterprise-first approach and successful product-market fit in solving latency-sensitive issues.
• Sanctor Capital: An early investor of Aethir, Sanctor Capital sees Aethir as a key player in driving down costs and delivering games at scale, bridging the gap for users in regions with limited access to hardware and connectivity.
• Demand-Side Focus: Aethir's emphasis on addressing demand-side challenges distinguishes it within the Web3 space, ensuring a balanced orchestration between supply and demand in decentralised networks.
• Enterprise Focus: Unlike consumer-focused competitors, Aethir exclusively targets enterprises, ensuring service reliability and SLAs crucial for industries like gaming and AI.
• Utilisation and Real Computation: Aethir prioritises increasing marginal utilisation of compute to benefit the industry as a whole, fostering collaboration rather than competition among decentralised ecosystems.
• Checker Node Sale: Aethir's Checker Nodes play a vital role in ensuring ecosystem stability and decentralisation. The node sale aims to democratise participation, allowing anyone to contribute and earn Aethir’s $ATH tokens with minimal hardware requirements.
• Ease of Node Operation: Running a node on Aethir's network is straightforward, with lightweight software and options for deploying on VPS or utilising Node as a Service partners, making it accessible to a wide range of users.
• Future Outlook: Aethir's vision extends beyond technology, focusing on solving real business problems and driving meaningful impact in gaming, AI, and beyond, within the broader Web3 ecosystem.

08:01 - Introduction of Dan Wang and Aethir

Dan:

Hi everyone I’m Dan, one of the co-founders of Aethir. I started out in banking, but the bulk of career has been in gaming (I’m a hardcore gaming nerd at heart). I started at Riot Games, built League of Legends leading international publishing, management, global expansion and infrastructure. After acquisition by Tencent, I transitioned over to the Shanghai office as COO for Riot China.

I had the itch to go all-in Web3, being a big believer of the confluence of blockchain technology and gaming industry. I met the other Aethir co-founders when I joined a fund, and they had just finished building Tencent cloud gaming, scaling it up to ~38 million monthly active players. They were looking to come out to utilise a distributed decentralised hardware node network to help solve the biggest pain points with centralised cloud infrastructure, specifically in relation to latency, sensitive and real-time scenarios.

And Aethir was born

10:30 - Introduction of Ilya Abugov from Sanctor Capital

Ilya:

Hi everyone I’m Ilya, one of the partners at Sanctor Capital. We invest in early-stage projects in Web3 space with a focus on metaverse and gaming. We also look at infrastructure to tooling and distribution layers. (NFA disclaimer)

12:00 - What do you think is going to be the main focus of Aethir moving forward? How does Aethir intersect between three growing narratives of AI, DePIN and gaming?

Dan:

When we set out to build Aethir, the term “DePIN” wasn’t coined yet. We just knew that for latency-sensitive real-time scenarios, infrastructure didn’t really work.

We always wanted to build a distributed hardware network in order to solve these pain points, specifically around distance and performance being inversely related, and the overall cost.

For us, gaming is in our DNA. It’s our expertise. The initial onset of the enterprise clients we onboarded have been from the gaming ecosystem. I think this translates well to getting players to access content they previously couldn’t because it’s now in a cloud version – there’s no hardware barriers or reliance on localised compute. Essentially, they have one-click instant play to content. It makes sense for studios because it lowers the barriers to entry, increasing the total addressable market. For players, they can finally access the content they actually want to interact with, rather than accessing content based on hardware limitations. So gaming took the forefront for our initial partnerships.

What really took us by storm was the amount of interest and opportunity in AI space. There’s so many interesting applications of AI — we’ve seen some pop up in virtualized AI shopping assistance in TikTok or Lazada etc. —  that we didn’t account for, but are able to spring up and manifest because of service fees pushing down for compute. I think what’s really exciting is that gaming is massive (there’s 3.3 billion gamers in the world) and the adoption for a cloud-ready solution is huge there. But we’re seeing the rise in really creative AI applications starting to take hold, which we believe will exponentially increase once service fees start to come down.

Calvin:

I think you’ve touched on something really important. People don’t realise that the demand for AI and GPU compute as a whole is the harder part of the problem.

What we currently observe are people bragging on stats like the supply of GPU cards they have or the infrastructure they’re bringing in. But ultimately, I believe what's most important is getting people to actually use these infrastructures and build real business around them. Right now, many teams gearing up for this wave. And while many are realising that this is a relevant problem, there are not enough folks working on driving real demand and deals to be able to make use of any decentralised networks that we’ve built. At this point it feels like we are still in a very early stage.

16:07  - I would love to hear how Aethir fits in to drive this demand, and if you could you highlight some partnerships?

Dan:

The ecosystem is large enough that multiple marketplaces can exist. What Aethir offers is a fairly unique solution and different tech stack from most of the other competitors out there.

Especially because we’re largely enterprise focused, and most of our existing customers and upcoming partners that will be announced are some of the world’s largest telcos, 3 of the 5 world’s largest games, some of the largest applications of cloud rendering solutions that you can possibly find, such that a single contract’s ARR is greater than all revenue paid out in DePIN ever. This shows that Aethir’s enterprise-first approach with a focus on real-time and latency sensitive solutions has hit a product-market fit.

One good example is our partnership with Well-Link, one of the biggest cloud gaming solutions in China and exclusive operator of Genshin Impact Cloud. Globally, we’re running Genshin Impact on Aethir infrastructure exclusively.

One of the biggest telcos in the world — we can’t reveal who they are just yet — is using Aethir infrastructure to do virtualized smartphone instances. This means their users anywhere in the world access the same content and set of apps without being geo-restricted.

Last but not least, we have some other massive gaming partnerships with 300-400 million players in the world, monthly active users. People are looking for solutions to be able to access content without being restricted by their hardware, and we see this demand only growing as overall cost for a cloud-first solution starts to drive down.

We’re beginning to see other interesting cases pop up: maybe people are considering cloud-first solution, but use it as an onboarding tool or user acquisition method to get people to engage with their content because now the cost of user acquisition is so low.

Calvin:

Sometimes, when we live in our own crypto bubble, we forget that people have real cost margins and other things to think about. So it’s refreshing to think about the real business problems we can help solve with the tech that we have.

21:13 - Ilya, would love to hear about your thesis on what caught Sanctor Capital’s attention on Aethir, with Sanctor Capital being on Aethir’s cap table for nearly 2 years now.

Ilya:

What caught our attention was Aethir’s infrastructure allows for that greater, broader vision for the metaverse and gaming space. A lot of the time we talk about a digital universe, and gaming worlds being populated worlds inside of that universe. In the Web3 vision where gamers are citizens who are able to talk and connect with each other etc.

As games become more advanced visually and computationally, delivering these games at cost to a lot of the regions becomes quite expensive. We find that some games end up excluding themselves from regions based on computational or hardware requirements. Studios think about cost per user for more advanced games and gaming systems.

With Aethir, we saw their ability to drive costs down and deliver games at scale to regions and unlock new users. We often talk about the 3 billion user number — but that’s a tricky number because it includes mobile gamers and gamers across all sorts of different devices and connectivity. The number of gamers that can play more advanced games are much smaller because others don’t have the same access to hardware and connectivity. Aethir can help bridge some of that gap.

Aethir is also one of the teams that really thought about the demand side. With a lot of these DePIN markets, we’re talking about an orchestration system between supply and demand. If you have a vibrant market, the flywheel starts working and it makes sense. But if it's lopsided, where you have a lot of supply and no demand, things can fall apart quite quickly. We saw the Aethir team really think through their demand side of that equation.

This is also one of those examples where distributed technology brings value almost immediately outside the Web3 space. A lot of what we’re building oftentimes has planned implications within our space, and we hope it gets bigger. But with what Aethir is building, we see potential for meaningful impact in the gaming and AI space overall. This fits in our vision with Web3 being part of the overall gaming stack rather than a genre or business model.

29:08 - In terms of DePIN, did Sanctor Capital invest in anything more DePIN-focused recently?

Ilya:

I think it was pointed out that DePIN as a term only came about more recently. In reality, we can position some things like Oort or Coin Network that are our portfolio companies as DePIN as well. But again, I tend to be quite cautious with that term (DePIN) — because right now it’s popular, and gets used alot.

We (Sanctor Capital) look a lot at decentralised physical infrastructure quite a bit in the context of these orchestration markets, something I mentioned in the beginning with Aethir. We do have a lot of inefficiencies in how we’re using computational and storage resources. At the same time, that does cause a lot of unfortunate price situations for the end user, which also creates limitations around what gaming companies are able to deliver and explore. So we’re looking at these orchestration systems to see where those inefficiencies can be alleviated. The community nature of Web3 has allowed for better opportunities using these resources: distributed networks are able to address inefficiencies with CDMs, storage and computation.

We’re excited about this opportunity but we also always try to keep in mind: whose problem you are trying to solve from the get-go? Who is your target consumer? With Aethir, we saw they had good focus and a lot of experience around gaming and which they were able to expand from that as the opportunities became available to them. Similarly with DePIN, we ask if you are building for the consumer, or for a narrative? That becomes very important to us because it has a lot of implications for real growth of the product.

32:07 - Absolutely. When you say consumer, do you just mean an end, retail consumer, or some sort of audience who is willing to pay?

Ilya:

A little bit of both. In these orchestration systems, your immediate consumer is often another business. Especially with computation, consumers are normally enterprises who need these resources. You’re orchestrating businesses amongst themselves but at this point we’re starting to add more consumer-level GPU resources. From a market perspective, supply and demand is often B2B, at least at the beginning.

But what’s important is that what you’re offering (ie. to gaming companies) is a computational solution that addresses a real problem which relates to the end user (ie. the gamer). Are you solving a real problem for them that allows them to do something meaningful and different such that they will change their existing behaviour? When you’re working with B2B cycles, these companies can be slow-moving and conservative even if they’re trying to make a change. So you need to give them a very compelling reason to change existing processes that they have. Implementing infrastructural solutions, particularly, is not an easy decision to be made. Again, what got us excited with Aethir is that they were able to demonstrate a reason compelling enough for some of these companies to think about reorganising their business processes, take advantage of this opportunity and step in on the demand side.

Calvin:

That’s an interesting segway to our next question around the competitor landscape as a whole. Some of what you (Ilya) mentioned is already what we’re seeing with other competitors in the space.

34:35 - Dan, can you share more about the competitors and how Aethir differs?

Dan:

We’re really supportive of all the other DePIN projects out there. In fact, any increase marginal utilisation of compute is a good thing for the industry. Just because anyone is providing compute on our network, doesn’t mean they can’t do the same on other platforms as well. Any time there is an increase in utilisation, the overall ROI for their hardware increases. This means they can afford to lower actual service cost because they’re getting utilisation and return on their investment.

Lowering service costs is what we’re really about because it lowers the barrier to entry for really interesting use-cases. Thai can be done on any distribution channel as well, our goal is for compute to be less underutilised.

With regard to the competitor landscape, Aethir is predominantly and exclusively enterprise-focused. With Render, IO etc. (we love what they’re doing; aggregating consumer GPUs) you can contribute your home computer and spare GPU to their network and earn tokens.

However, the issue is that there’s no real enterprise use-case or scalability in offering these types of consumer services; Epic Games is not going to trust Fortnite to run on someone’s home computer. Meta is not going to run their LLMs on your personal home computer. There’s no service level agreement (SLA) or guarantee of service. For example, say you’re playing PUBG Mobile or Fortnite; what happens when someone at home turns on Netflix and the bandwidth chokes. Do players drop from the game? If you’re on your home PC and you want to play a Cyberpunk 2077 — does that mean the resources are now limited and you can no longer provide LLM training or gaming instances for other folks? There’s no way to guarantee this level of service. So there is no enterprise use-case that would accept this type of risk because it interrupts the actual promise of service that these companies offer to their users. Because of that, while there is a growing use-case for stable diffusion images etc., I think what’s happening here is very job and task-based workflows and non-latency sensitive, non-real time, there’s no risk of interaction and disruption of service. Again, nothing wrong with what they do, I think we’re just serving 2 different markets. There’s a huge need for what companies like Render and IO do, but it doesn’t serve the enterprise market properly.

But because Aethir is enterprise-focused and the compute we’re onboarding is hundreds of millions dollars at a time from IDCs, some of the largest telcos, infrastructure providers and mining partners in the world, they’re able to maintain uptime and SLAs that are being demanded by the actual enterprise use-cases like gaming companies, large AI LLM training companies etc.

Calvin:

I think especially there are so many teams out there competing in this space,  what I found most refreshing working with Aethir is being able to have this mindset of WAGMI approach towards having better real utilisation and real computation being contributed to any of these decentralised ecosystems, rather than trying to be the sole winner. It’s a powerful position to be in when you realise the goal is now how to solve bigger technical demands, and work with projects to to help solve the business needs of companies both in and outside of crypto.

41:20 - For those who are new to the concept of node sales, could you share anything to help users better understand what Aethir’s Checker Node Sale is about?

Dan:

Aethir’s Checker Nodes are essentially inspectors in our ecosystem. They check that GPU compute is actually delivering on the SLA I mentioned earlier. This includes liveliness, capacity, uptime, quality of service, and that is why checkers are so important. They prevent collusion, uphold SLAs, ensure level of trustlessness so the ecosystem is fairly decentralised and stable.

For checker nodes, we’ve allocated 15% of our total token supply for checker node operators. We knew this node sale was one way we could ensure as many participants contribute as possible. Not everyone can afford a Nvidia H100 ($300,000), or deploy a data-centre. But everyone can deploy the checker software, operate in our ecosystem and contribute to the Aethir network. It is a lightweight program, very low compute, memory and storage requirements. All you need is a stable and  bandwidth connection.

To run a node, you’ll need to download the software, and simply click “run” to start contribute and earning Aethir tokens.

To run a few nodes at a time, you can deploy on a VPS. Alternatively, we have several partners running Node as a Service, and users can just delegate to them. I don’t want anyone to be intimidated by running a node — there’s really nothing special to it.

You just purchase a licence which gives you the right to operate a node and download the software. You can read more at https://docs.aethir.com.

About Aethir

Aethir is revolutionizing DePIN with its advanced, distributed enterprise-grade GPU-based compute infrastructure tailored for AI and gaming. Backed by leading Web3 investors like Framework Ventures, Merit Circle, Hashkey, Animoca Brands, Sanctor Capital, Infinity Ventures Crypto (IVC), and others, with over $32M in funds raised for the ecosystem, Aethir is paving the way for the future of decentralized computing.

Website | Twitter | Discord | Telegram | Blog

About Impossible Finance

Impossible is an on-chain native, research advisory firm with a DeFi launchpad and accelerator. We help projects kickstart, fundraise, scale and launch their tokens while enabling users to learn, discover and invest in high quality crypto opportunities.

Website | Twitter | Discord | Telegram | Blog