Intel (NASDAQ:INTC) is widely known to have slipped from its traditional tick-tock product release cadence. For those of you unfamiliar with tick tock, the idea is that each year, Intel releases either a fundamentally new processor design on a proven manufacturing technology or brings a familiar design to a brand new manufacturing technology.
This has proven a very effective means of managing risk for the world's largest chipmaker by revenue, but as of late Intel has fallen off of that traditional schedule. It seems that with its 2015 "tock" product, Intel is trying to get back on this cadence.
The historical pattern explained
Here's a recap of Intel's recent processor launches:
- Jan. 7, 2010: Intel launches the 32-nanometer shrink of Nehalem known as Westmere.
- Jan. 3, 2011: Intel launches Sandy Bridge, the first new architecture on the 32-nanometer node.
- April 29, 2012: Intel launches a 22-nanometer shrink of Sandy Bridge known as Ivy Bridge (note that high volume manufacturing began in the third quarter of 2011).
- June 4, 2013: Intel launches the first ground up design on 22-nanometers known as Haswell.
- September/October, 2014: Intel is set to launch the 14-nanometer shrink of Haswell known as Broadwell.
The move from Westmere to Sandy Bridge was flawless, but there was a slight hiccup in the transition from Sandy Bridge to Ivy Bridge, and then another in shifting from Ivy Bridge to Haswell. It is unclear if the Ivy Bridge and Haswell slips were due to technical issues or simply inventory management or demand problems.
By Broadwell, the slips had piled up
With Broadwell, Intel suffered production issues. The 14-nanometer processor wasn't yielding well enough to be commercially viable at Intel's desired gross margin levels, leading the company to delay volume production.
Intel is very sensitive about its gross margins, and given that the chipmaker lacks any meaningful competition from Advanced Micro Devices (NASDAQ:AMD) in the PC space, it could afford to keep the fight going with the current Haswell products until Broadwell was ready to go at high yields (in a more competitive environment, Intel may have chosen to take a gross margin hit). That said, too many of these slips could open a nice window of opportunity for AMD.
Broadwell delay doesn't push Skylake
In October 2013, an analyst asked Intel CEO Brian Krzanich if the Broadwell delay would delay Skylake. Krzanich, surprisingly enough, said the Broadwell delay would not push Skylake out. Seemingly confirming this is a recently leaked Intel road map slide:
As you can see, Intel plans to launch Skylake for mainstream and premium desktops during the second quarter of 2015, less than a year after the first Broadwell systems will hit the market.
While it's unclear what the road map will look like for Skylake for notebooks and convertibles, it may not be aggressive since Intel seems to be just outright skipping Broadwell for the majority of its mainstream and premium desktop products but is bringing Broadwell to mobile.
Why did Intel do this?
In its core PC space, Intel must ensure that it maintains and grows its market segment share, as the long-term secular trends in personal computers remain uncertain. While AMD may be down, Intel still needs to take its rival seriously. Despite AMD's multiyear weakness in PC CPUs, a company with essentially nothing to lose will move aggressively, particularly if it senses that its larger rival has become complacent.
Additionally, as mobile system-on-chip products add additional functionality like HEVC/H.265 video encoding/decoding, Intel must keep pace in adding this functionality on the PC side of things to stay competitive on features and efficiency. The Gen. 8 graphics block inside Broadwell doesn't seem to include some of these key features, so bringing Skylake's much-improved (from a media standpoint, anyway) graphics block should be a key priority.
Returning to its tick-tock cadence by pulling in Skylake would be absolutely fantastic for Intel. Though Intel's competitive problems are in tablets and phones, and not in the PC space, it needs to sufficiently advance the PC platforms in order to maintain a sizable performance and feature delta against the quickly improving mobile products.