Intel Launching First 10nm Cannonlake CPUs in 2017, Succeeded by 2 More 10nm CPU Families
Intel is planning to launch their first 10nm CPU, called the Cannonlake, in 2017. Though before that, they will bring their Kaby Lake CPUs to the market, sometime this year.
For years, Intel has been releasing new processors following a Tick-Tock model, with two generations based on a single processing node. The Tick moves to a new processing node, improves efficiency and brings minor changes to architecture, while the Tock makes some major architectural changes to the CPU.
The 10nm process node will mark the first time when the Santa Clara, California-based chipmaker will move away from their traditional Tick-Tock model towards a Tick-Tock-Tock model, with each process node being used in three generations of CPUs. That is primarily to address difficulties experienced in an attempt to continue making their manufacturing technology smaller and smaller.
Intel may already be breaking their Tick-Tock cycle with the release of Kaby Lake, unless it’s merely a re-release of Skylake with higher clock speeds. Both Broadwell and Skylake processor architectures are based on the 14nm processing node, with Kaby Lake making up the third CPU generation on this node.
The 10nm process node roadmap suggests that we’ll have Cannonlake in 2017, Ice Lake (Tock) in 2018 and Tiger Lake (semi-Tock) in 2019. It will then be replaced by the 5nm CPU architecture in 2020. The 5nm CPUs are said to use the Extreme Ultraviolet Lithography (EUL) technology that should allow sub-10nm processing tech to be possible.
Source: Motley Fool