Intel is preparing to launch its eighth processors. Intel held its annual Investors Day last week, where the company revealed more details about its next generation of processors.
With a Coffee Lake moniker, the chips will offer a performance boost of up to 15% – something that has become the standard for Intel.
The new chips will use the same 14nm process as the previous generation Broadwell, Skylake, and 7th generation Kaby Lake processors.
In 2007, Intel adopted a ‘Tick Tock’ strategy for its processor refresh. A ‘Tick’ represents a die shrink and a ‘Tock’ represents a new microarchitecture.
So, Intel would refine an existing processor microarchitecture by shrinking it down and making further adjustments – tick; next, it would release a brand-new microarchitecture using the same manufacturing process as the last tick – tock.
Unfortunately, Intel officially abandoned this strategy in early 2016 as it simply couldn’t meet its own targets. The Broadwell processors came out in 2014, yet the 8th generation Coffee Lake processors will use the same 14nm process as Broadwell did.
Still, there are refinements to be made, and that’s what Intel is doing.
According to ArsTechnica, Intel will launch its 8th generation processors in the second half of 2017; though, Intel is maintaining its silence with regards to more details about the Coffee Lake lineup.
A few rumors have suggested that Coffee Lake’s i7 lineup will include a few six-core processors, which will be a first for Intel’s consumer offering. Intel only offers more than four cores in its Xeon and Extreme Edition lineups, at the moment.
The Santa Clara giant also hasn’t explained how it expects to squeeze out 15% more performance out of the same 14nm processor, but that’s probably a trade secret anyway.
AMD’s Ryzen CPUs are expected to arrive in March. While these CPUs are based on a 14nm process – much Coffee Lake – they are also meant to be cheaper.
A few rumored – and unconfirmed – pricing details have suggested a massive price gap, giving AMD an advantage.
The first 10nm processors to arrive in consumer products won’t come from the once-leading Intel but Samsung and Qualcomm. The two companies are working on the Snapdragon 835 SoC in collaboration, Qualcomm will utilize Samsung’s 10nm process for the chip.
The Snapdragon 835 is also what Microsoft used to demo x86 app emulation on ARM with Windows 10 last year. It’s clearly a high-performance processor, thanks to the die shrink and further improvements. These SoCs are expected to launch in the first half of 2017 – perhaps at MWC 2017.
However, fret not – Intel is working on building a 10nm process for its Cannon Lake processors. Cannon Lake doesn’t have a release date, but Intel has hinted that it would come to data centers and enterprise before it comes to the average consumer.
Moore’s Law is in trouble since the advancement in processor technology has slowed down; perhaps it’s time for that big push into exploring graphene-based processors.