Intel has seemingly dropped their Tick-Tock strategy which is the process of offering a new node (Tick) followed by a new architecture (Tock) during a two year time frame. The Tick-Tock strategy is based on the fundamentals of Moore's Law which was devised by Intel's industry veteran, Gordon Moore, back in 1965 and revolves around the process of a CPU incorporating double the number of transistors in a certain period of time (2 Years).

It has since become a target for miniaturization in the semiconductor industry. However, due to complications in the development of smaller nodes using existing materials and methods, the Tick-Tock strategy is no longer a very viable process to build CPUs.

Intel Says Farewell To Tick-Toc, 14nm and 10nm Process To Utilize 3 Year Cadence With Process, Architecture and Optimization Strategy

With the Haswell generation of processors, we have witnessed the decay of Tick-Tock. After Haswell, Intel was supposed to release Broadwell but the microarchitecture was delayed and available to a small market space in late 2014. The full on Broadwell lineup released in Q1 2015 and delay was mostly due to poor yields with the latest 14nm process. Later that same year, Intel introduced a new family known as Skylake which included the desktop and mobility chips on completely new platforms. But like Broadwell, most of the improvements from Skylake came in the form of better architecture design, faster graphics and support for new I/O capabilities.

In Q3 2016, Intel plans to launch another generation of processors. This upcoming generation will become the 7th generation Core family but it's also going to feature a 14nm node. The Kaby Lake processors will be utilizing the most mature form of Intel's 14nm process node but it will also mark the third processor family to utilize the 14nm node, a first from Intel and the final nail in the coffin for the Tick-Tock roadmap.

Intel's Latest PAO Strategy (Image Credits: The Motely Fool)

As confirmed by The Motely Fool, Intel is now longer following the Tick-Tock roadmap and will now be devising a completely new strategy that is supposed to leverage from a single process node during a three year cadence. In 14nm case, we are looking at Broadwell (2014), Skylake (2015) and Kaby Lake (2016). According to the new strategy, the three year cadence will involve of three phases. First is the arrival of the new process which in the case of 14nm was Broadwell. This will be followed by the Architecture phase which was Skylake on 14nm. The final phase would be optimization which is Kaby Lake on 14nm.

As part of our R&D efforts, we plan to introduce a new Intel Core microarchitecture for desktops, notebooks (including Ultrabook devices and 2 in 1 systems), and Intel Xeon processors on a regular cadence. We expect to lengthen the amount of time we will utilize our 14nm and our next generation 10nm process technologies, further optimizing our products and process technologies while meeting the yearly market cadence for product introductions.

A time frame courtesy of Anandtech, showing the various process node from Intel.

Now we don't know how significant of a difference would there be between the Architecture and Optimization phases but knowing that Intel managed to obtain a small fraction of IPC improvement over Broadwell from the Skylake process, we can expect around the same or slightly lower percentage of improvement from Kaby Lake. As for other optimizations, the Kaby Lake chips can shine in areas involving graphics, media capabilities and I/O support.

Intel's 10nm Lithographic Node To Last 3 Years - Cannonlake (P), Ice Lake (A), Tiger Lake (O)

It is already confirmed that Intel’s 10nm Cannonlake processors have been scheduled for launch in 2H of 2017. This is a year after Kaby Lake would be introduced. Now what Intel plans to do after Cannonlake is reminiscent of what they are currently doing with their 14nm process since it will follow the new PAO strategy. It is reported that Intel will be introducing 10nm Ice Lake processors in 2018 and 10nm Tiger Lake processors in 2019.

While Cannonlake will be designed on a completely new node, most of its DNA would be taken from the Kaby Lake generation of processors as its a Tick. The 10nm process will really start to show its prowess in 2018 with the Ice Lake generation of processors which will be using a brand new microarchitecture, further ensuring the IPC lead and efficiency gains on the smaller node.

Gordon Moore devised the Moore's Law back in 1965, the Tick-Tock roadmap is formulated with it.

We don’t really know much about Ice Lake right now but the previous reports suggest that it will be incorporating the FIVR (Fully Integrated Voltage Regulation) modules which were excluded from Skylake and Kaby Lake processors. Intel will also release an intermediary platform known as Tiger Lake which will serve as their semi-Tock 10nm family. Tiger Lake is a brand new name which we haven’t heard of before but it makes the roadmap slightly clear as to how longer Intel plans to make use of the 10nm node.

There are several challenges ahead for Intel as competition rises from TSMC who are expected to make develop their 7nm node process in 2017 followed by 5nm process node in 2020.

“TSMC expects to start production of 7nm chips in the first half of 2018, said company co-CEO Mark Liu at a recent investors meeting. He did not specify whether the node would be ready for volume production or just risk production.

TSMC also revealed it will be ready to use extreme ultraviolet (EUV) lithography to make 5nm chips. “We’ve made significant progress with EUV to prepare for its insertion, likely in 5nm,” Liu indicated.” via Digitimes

The TSMC 7nm process is much alike to the 10nm process which Intel will release with Cannonlake but knowing that Intel will slow down and stick with a 3 year roadmap for quite while, it might give edge to other companies in the field of semiconductor development. Currently, Intel still holds a strong lead over competitors with their own fabs producing state of the art process nodes.

As we near 2020, Intel will aim to use new technologies for chip development and production such as spintronics and the quantum tunneling that was discussed by Intel’s William Holt who leads the technology and manufacturing group at the cpu company. There are definitely interesting times coming ahead as we reach the end of life for silicon and newer materials will be deployed in the development of next generation processors.