A friend and I are developing a new fast, space-efficient (read: probabilistic) key-value store. We're using it for some computational biology applications.
It's cool because:
* It scales to really large data sets: We store ~10 billion keys in memory on a single, not-too-ridiculous machine, and the design supports sharding trivially
* It's fast: we have a lot of optimization still, but can do ~1M lookups/sec (O(1) lookups)
* It's a data structure! (Which as non-formal-CS people we find fun)
Would be very interested to hear other applications that people would find the above attributes useful for!
It's cool because: * It scales to really large data sets: We store ~10 billion keys in memory on a single, not-too-ridiculous machine, and the design supports sharding trivially * It's fast: we have a lot of optimization still, but can do ~1M lookups/sec (O(1) lookups) * It's a data structure! (Which as non-formal-CS people we find fun)
Would be very interested to hear other applications that people would find the above attributes useful for!