Tensor parallelism (TP) changes how some of that shelf space is used.
About SN10 Pareton
Pareton@Pareton_aiTensor parallelism (TP) changes how some of that shelf space is used. With TP4, the model is distributed across 4 GPUs. With TP8, it's distributed across 8. So each GPU holds a smaller portion of the model, and distributed memory like the KV cache can scale across more GPUs. But not every calculation is distributed this way.
SN10 Pareton at a glance
SN10, Pareton (α token Ⱂ), is one of the independent markets on the Bittensor network, each with its own α token and its own price. Its α currently trades at 0.006469 τ (≈ $1.99), against 16,576 τ of pool liquidity. On-chain, the best-delegated validator on it yields about 40.86% APR (simple) and 50.47% APY (compounded), emissions add 0.000003 τ to the pool each block, epochs run every 360 blocks. It was registered at block #2,869,647.
More from SN10 Pareton
Taosis indexes announcements, articles, posts and releases about Bittensor subnets from their own channels and keeps a record of each. The text above is the source’s; the figures are read from the chain by Taosis.