# Encryption algorithms larger than 256 Bit for “big data” encryption?

I'm somewhat new to encryption. When looking at encryption programs for big data, I frequently see a maximum of 256 bits.

Why do we generally restrict our (symmetric) keys to 256 bits? Can more powerful encryption algorithms be used practically, or is there a specific reason for that 256 bit maximum?

• Are you asking "why do we generally restrict our (symmetric) keys to 256 bits"? If you aren't asking that, what are you asking? – poncho Jul 31 '15 at 21:30
• I hope you don’t mind… but to be sure everyone grasps what you’re asking, I’ve added that “reformulation” to your question. Please feel invited to modify my edit in case I missed something. Btw: Welcome to Crypto.SE, Erich! – e-sushi Aug 1 '15 at 0:03
• I think the actual issue is more with the blocksize of the cipher and not the key size – Richie Frame Aug 1 '15 at 6:16

This is because there are fundamental physical limits on the energy required to perform a computation. The Landauer limit is one: to switch one bit takes $kT\ln2$, where $k$ is the Boltzmann constant ($1.68\times10^{-23}$), $T$ is the temperature in Kelvin, and $ln2$ is the natural log of 2. On average brute forcing a 256-bit key will take $2^{255}$ operations (half of $2^{KeyLength}$).
Assume a computer exists which operates at this limit, is at the temperature of the cosmic microwave background (about 3K), and only needs to do one operation to test a key. It will then take $2^{255}\times 1.68\times 10^{-23}\times 3 \times \ln 2$ Joules, which comes out to about $2.02\times 10^{54}J$.
For comparison, the if the entire mass of the Earth were converted to energy it would only release $5.4\times 10^{41}J$. So you'd be annihilating 10,000,000,000,000 earth-sized planets just to brute force one key. Or building a Dyson sphere and capturing the total energy output of the Sun for 10,000,000,000,000,000,000 years.