Pune
SuperBrain pools spare or dedicated RAM from multiple machines into one unified, shared memory space that acts like a single large block of RAM any application can access. It divides the pooled memory into fixed-size blocks, allocates them across nodes on demand, stores every block with two copies on different machines for durability, and enables direct read/write access from clients to the nodes holding those blocks. A high-availability Raft coordinator cluster keeps track of block locations, node health, and allocations, automatically electing a new leader in under two seconds if the current one fails, while the client SDK handles everything transparently, providing a block map on allocation, managing direct node communication for data operations, and automatically switching to a replica if a node dies. This creates a fault-tolerant, in-memory data fabric where applications can store and retrieve large amounts of data quickly and continue working even when individual machines or coordinators crash or disconnect.