The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.
Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most. A picture that shows something else, or shows it badly, is looked at too, and moved down or replaced.
| Deployment | 2015 internally; disclosed May 18, 2016 | Deployment year and public announcement differ[1][2][4][5][7][8][9][11] |
|---|
| Packaging | Board fits a hard-drive slot | 2016 maker description[1][2][4][5][7][8][9][11] |
|---|
| Product | Published context |
|---|---|
| Tesla P100 | Pascal data-center accelerator announced April 2016; a contemporary general-purpose GPU alternative.[3] |
Product context, not a benchmark ranking. GPU, TPU and cloud-instance peak rates and memory units may differ.
| Generation | Public milestone | Role |
|---|---|---|
| v1 | 2016-05-18 | Inference only[1][2][4][5][7][8][9][11] |
| v2 | 2017-05-17 | Training and inference[1][2][4][5][7][8][9][11] |
| v3 | 2018-05-08 | Training and inference[13] |
| v4 | 2021-05-18 | Training and inference[1][2][4][5][7][8][9][11] |
| v5e | 2023-08-29 | Training and inference[1][2][4][5][7][8][9][11] |
| v5p | 2023-12-06 | Training and inference[1][2][4][5][7][8][9][11] |
| v6 | 2024-05-14 | Training and inference[1][2][4][5][7][8][9][11] |
| v7 | 2025-04-09 | Training and inference[1][2][4][5][7][8][9][11] |
| v8t | 2026-04-22 | Training optimized[1][2][4][5][7][8][9][11] |
| v8i | 2026-04-22 | Inference optimized[1][2][4][5][7][8][9][11] |
Main thread follows public generation reveals, with 8t and 8i announced together. Availability dates belong to each pin timeline.
Pascal data-center accelerator announced April 2016; a contemporary general-purpose GPU alternative.
Source
