Intel has announced bad news regarding its 8th generation processors saying they cant upgrade to a smaller manufacturing process this year as well. The company had more than its fair share of problems during its transition from 22nm to 14nm nodes, but for the time being it is refraining from any further developments in that domain.
In other words, Intel will be using the 14nm process for the fourth generation in a row, making it the slowest progress in chip making history. The process was first used with 5th generation chips and its going to be used again in the 8th generation series.
Despite the 7th-generation Kaby-Lake representing the ‘Optimization’ stage of the new three-step process that succeeded the legendary tick-tock model, Intel has decided to keep the microarchitecture for the fourth time in the row.
Intel is also indicating that whenever those long overdue 10nm chips do arrive, the shifts towards the adoption of technology will be gradual, either seeing a Xeon or a mainstream chip make the jump shift in the first round.
However, there are also reports that the 8th-gen series will focus on the low-voltage Y or U series, making it difficult to say anything for sure.
Intel is also expecting to make gains of around 15-percent set to be achieved with the latest update, similar to the level achieved by the last generation.
Difficulty Keeping Up With Moore’s Law
Intel has publicly talked about the struggles associated in keeping up with the Moore’s law in this day and age, pointing out that the real cadence was closer to two-and-a-half years today than two.
In 2016, the company’s famous tick-tock model, where a shrinkage in die followed by a change in microarchitecture repeated every two years, was finally retired. A new third stage called ‘optimization’ was added to the fray.
Whatever scenario pans out, when the transition to next die shrinkage does happen, it will bring with it a lot of lost sleep at Intel.