Riot Platforms Shifts Focus to Data Centers for HPC and AI
The company Riot Platforms, formerly known as Riot Blockchain, is following the steps of other firms in the sector and diversifying its operations. It is now directing its efforts toward hosting workloads for high-performance computing and artificial intelligence. To achieve this, it purchased more than 80 hectares in its Rockdale, Texas complex. According to reports, the company will lease space in this new facility to AMD. 🔄
Financing the Shift with Digital Assets
Riot recently announced that it acquired the entire land in Rockdale, where it already operates. The firm paid 96 million dollars for the area. It obtained these funds by selling nearly 1080 bitcoins that it held. This movement of digital assets reveals a defined strategy to reorient its business model.
Key Details of the Transaction:- Purchase of the entire land in Rockdale for 96M$.
- Financing came from selling approximately 1080 BTC.
- Demonstrates a clear willingness to change the core of the business.
Going from digging up digital gold to renting peaks to forge the algorithmic future.
AMD Deploys Power in Phases
The agreement with AMD involves installing the computational load progressively. The first 5 megawatts will be ready in January. By May, the tech company plans to use 25 megawatts. The contract gives AMD the option to obtain an additional 75 megawatts and the right of first refusal to purchase another 100 megawatts of capacity.
Phases of the Capacity Agreement:- First phase: 5 MW operational in January.
- Second phase: Expansion to 25 MW by May.
- Future options: 75 MW additional and preferential right over 100 MW more.
A Complete Strategic Transformation
If AMD exercises all its options, the total power it would lease would reach 200 megawatts, which is the maximum capacity of the two buildings on the campus. This step consolidates Riot's transformation, shifting from being a miner to a provider of infrastructure for expanding sectors. The heat generated by ASICs for mining could now warm processors for training AI models. 🤖