Cryptopolitan has had the pleasure of sitting down with Tory Green, the not-so-new CEO of io.net, to talk about the specifics of his vision for the company, and his mission in this industry.
io.net is a decentralized network that’s all about making GPU power more accessible to everyone.
It’s designed to cut costs and speed up projects for engineers and businesses by offering quick access to a huge pool of GPUs whenever they need them.
The network, called the Internet of GPUs (IOG), pulls together GPUs from all over the world, letting users tap into tons of computing power for things like AI, machine learning, and cloud gaming.
Here is all we learned from Tory:
QUESTION: Nice to meet you, Tory. Let’s get right into it. So, you stepped into the CEO role at a uniquely important time for io.net. What steps did you or are you taking to make sure everything runs smoothly and stays on track during this leadership change?
ANSWER: I officially stepped into the CEO role in June, but I ran most operational aspects of the business during the previous 15 months. This allowed me to transition seamlessly into the new role. My first priority as CEO was to establish a clear strategy. I took the time to articulate our long-term vision and the steps required to achieve it to the global IO team. Without this alignment, it would be very difficult to keep everyone on track to hit our goals.
Next, I wanted to build a strong leadership team, so I reviewed the various roles we had and focused on finding the right profiles with the skills and talent we need to grow as an organization. These are incredibly talented people who have navigated challenges and come from reputable organizations.
I’m also a big believer in ensuring there are high degrees of transparency and communication. Therefore, to drive people in the right direction, we established strong cadences for check-ins to maintain focus. Not only internally, but the leadership team makes themselves available and visible to our community on a weekly basis to regularly communicate transparency and vision. We are also heavily focused on reporting to ensure we have the necessary data to keep the team informed, allowing us to stay agile during the leadership transition.
Lastly, I ensured we kept our eyes focused on execution. Transitional periods can sometimes be challenging; however, by prioritizing our operational goals, we were able to sustain momentum as we continued to innovate and scale IO.net.
QUESTION: Io.net has been growing its decentralized GPU network really fast. What specific technical issues have come up as you try to keep everything running smoothly, especially with low latency and high reliability, as you scale up to manage hundreds of thousands of GPUs all over the world?
ANSWER: As an innovator within this space, it’s only natural that we face some technical hurdles to overcome, for example when we scale our decentralized GPU network. One of our goals at io.net is to always guarantee consistent GPU performance across geographically dispersed nodes. We were able to solve this with data processing and simultaneously minimizing latency, while keeping a strong focus on real-time AI tasks.
Another area where we placed a lot of emphasis on was reliability. We implemented advanced monitoring, validation and redundancy systems to prevent node failures and maintain io’s high performance standard. These strategies are essential when managing a global network.
Scalability is also a very important priority for us. Our tech team worked hard to develop algorithms to efficiently allocate resources and balance loads and by doing so, we can ensure a more seamless customer experience as we grow.
All of these challenges were tackled head on, we take a lot of pride in maintaining a high standard for our customers as we grow our decentralized GPU network.
QUESTION: How do you see DePin working in other industries like healthcare and energy, beyond just AI and cloud gaming?
ANSWER: Great question! We believe that DePIN’s potential far exceeds conventional AI & cloud computing. There’s a lot of innovation happening across the globe, for example I’ve seen transformative applications being developed in industries such as healthcare and energy which is exciting.
Decentralization is incredibly important for healthcare as it not only improves data sharing, but also provides access to computing power for critical tasks such as medical imaging, genomics, and AI-driven diagnostics.
The opportunity to transform healthcare is huge if it is able to leverage underutilized computing power on a global scale. This will result in healthcare organizations being able to efficiently process large datasets, which in turn can improve the speed and accuracy of patient care.
I would also argue there are opportunities in the energy sector. For example, companies can use DePIN to support smarter energy grids and even renewable energy management.
Decentralized networks have the power to support industries in balancing supply and demand through the utilization of distributed computing power. This enables them to monitor energy use, optimize distribution and even manage storage solutions in real-time. Not only can this potentially reduce cost, but also improve sustainability and grid resilience.
I would also add that the DePin decentralization model offers significant flexibility and scalability which can enable strong cost control & cost efficiency, therefore making it an ideal solution for industries such as healthcare, energy and potentially many more that require a robust infrastructure to process complex and data-intensive tasks.
QUESTION: What new innovations are you bringing in to make sure your network can handle the increasingly complex tasks needed by AI and machine learning applications?
ANSWER: There’s a lot of exciting innovations happening at io.net! We’re focussed on implementing numerous innovations to process AI tasks and applications in a way that allows customers to scale within our performance-driven decentralized GPU network.
For example, we developed our proprietary Proof of Work system, which includes a detailed and robust hardware verification, VRAM checks and stricter CPU requirements to maintain performance and stability across the network.
Moreover, we’re also optimizing scalability. By pooling global compute resources into decentralized clusters, we allow ourselves to flexibly scale to meet demand and enable dynamic expansion, whilst simultaneously reducing latency especially for inferencing. As AI taks become increasingly more complex, our network can scale and therefore deliver the computing power required to meet these demands.
In order to ensure we’re providing high quality computing resources, we’re also introducing a tiering system which will require verification from enterprise providers. Therefore, tiering, along with staking and slashing mechanisms not only improves the integrity but also the reliability of io’s network. Therefore, such innovations will allow us to provide and manage hundreds of thousands of GPU’s.
Our focus is to always maintain a robust enterprise-grade platform that equals or exceeds the performance standards of the traditional cloud providers in this space. We do this by offering superior performance, and reliability for AI and machine learning applications at a fraction of the cost.
QUESTION: You’ve recently brought in some people from top tech companies to your executive team. What are your personal expectations from io.net’s current staff?
ANSWER: We hire talent from top tech companies to scale io.net and meet our ambitious goals. My personal expectations for our team revolve around three key areas: accountability, innovation, and collaboration.
Accountability is non-negotiable. Every team now has a KPI that they are expected to meet. Every person must take responsibility for their KPIs and deliver high-quality work. In order to do this, one must think like an owner. Every individual at io has a role to play in making our company successful. Our culture values results more than anything.
Innovation is important. Our industry evolves rapidly. Excellence requires that we stay agile and proactive, while pushing the boundaries with creative problem solving. If we continue to create this culture and live these values, io.net will be positioned to disrupt the cloud and AI markets.
Collaboration is important to our success. Our talent comes from many different backgrounds, whether it’s from Web3, AI, or top cloud providers like AWS and GCP. One way we bridge these gaps is collaboration. I expect our staff to not only bring their expertise but also foster a collaborative culture where ideas are shared openly and in a timely manner.
My expectation is that every member of the io.net team will embody these values, moving us forward as we continue to grow and innovate in the decentralized computing space.
QUESTION: You’ve talked a lot about focusing on operational excellence and discipline. Can you give some real examples of how you’re putting these ideas into practice within io.net’s decentralized setup, where the usual top-down structure isn’t as strong?
ANSWER: io.net’s success is built on operational excellence and discipline, especially regarding our decentralized network. We don’t have a traditional hierarchical structure but we do create accountability by requiring KPIs for everyone.
Transparency is an important company value. Our success is built on transparency in every level of the business. At io, transparency means that every stakeholder must provide status updates and blockers. If there are blockers or challenges, the team comes together to solve them quickly.
To manage our network, we use automation and monitoring tools. This provides us with a view of how well the network performs. This approach allows us to maintain network health with minimal manual intervention. This efficient approach allows us to maintain discipline while we grow the company.
io.net fosters a culture of accountability, transparency, and smart technology to maintain and scale our decentralized environment.
QUESTION: Your mission to make GPU compute power accessible to everyone is central to io.net. How are you making sure this access doesn’t end up favoring certain areas over others, especially in less-developed regions?
ANSWER: io.net believes that GPU compute power should be available to the entire globe, including less-developed areas; it’s central to io.net’s mission. Our decentralized network leverages underutilized GPUs worldwide. This enables us to distribute compute power equitably across 138+ countries, rather than concentrating all of our efforts in developed regions.
We implemented a tiered system with KYC/KYB verification to deliver high-quality computing to a global market. Our staking requirement is fair and allows for broad participation and rewards, regardless of location.
By working with local partners, we further tailor access to meet the individual needs of different regions. No area is left out as we expand.
Our decentralized network provides equitable access to GPU power. High-quality compute is available to everyone, everywhere, at any time to meet AI and machine learning needs.
QUESTION: Io.net has onboarded a significant amount of GPUs to help AI startups. What are the main metrics you use to measure how well these deployments are doing, and how does that influence your plans to expand?
ANSWER: We measure the success of our GPU deployments with metrics: compute hours delivered, utilization rates, and uptime. These metrics provide us with a detailed view of how GPUs serve the compute needs of AI startups. The benchmarks for these metrics enable io to deliver high-performing GPUs.
We also track user feedback and customer retention rates. For example, our customer satisfaction metrics often provide actionable data that highlight improvement opportunities in different parts of the product. This data enables us to monitor customer satisfaction and solve problems fast as we scale, using the knowledge that we provide high-quality networks.
When we monitor these metrics, we’re able to make smart decisions regarding the adjustment of the network, optimize resource use, and determine the right times and places to onboard additional GPUs as demand grows.
QUESTION: The collaboration between io.net and Chainbase involves integrating an omnichain data network into your AI projects. What specific challenges have you faced in making different blockchain networks work together seamlessly?
ANSWER: There’s many challenges that disruptive startups face, in our case integrating an omnichain data network, like Chainbase, into our AI projects presents some interesting hurdles to cross alongside creating seamless communication between different blockchain networks and interoperability. However, our tech team is constantly and consistently innovating and breaking the mold to continuously make progress.
QUESTION: How do you make sure compute power remains reliable and consistent across such a varied and sometimes unstable supplier base, especially when dealing with mission-critical applications?
ANSWER: The reliability and consistency of GPU supply in our network is obviously a top priority for us. We implemented advanced monitoring and validation systems that continuously track the performance and health of every node in our network. It’s our responsibility to deliver high-quality computing power and the validation system we developed helps us to quickly identify and address underperforming or unstable suppliers.
We also built redundancies into our network. These redundancies allow workloads to dynamically shift to other nodes if disruptions in quality service arise. As I mentioned earlier, we deployed a tiered system that requires verification for high-quality suppliers, prioritizing those with a proven track record of reliability.
INTERVIEWER: Alright, that’s our time. Thanks for doing this, Tory.
TORY: Thanks.
