Early indications are that the
SoC is very clearly on a collision course with the CPU. The Surface
tablet from Microsoft highlights the choice that OEMs now have when
choosing processor architecture. This tablet will be made in two
versions – one using an ARM-based SoC processor (Nvidia Tegra 3) and
another using an Intel x86-based CPU processor (Ivy Bridge). Table 1
compares the relevant specs of these two versions while Figure 1
shows a die-photo and block diagram of the two chips. It is evident that
the SoC based design is better suited for the ultra-mobile consumer
form factor where light weight and long battery life are valued more
than having the highest raw performance.
Click on image to enlarge.
1 Comparison of SoC (NVIDIA) and CPU (Intel) based MS Surface tablets.
It is clear that the tablet with SoC is more amenable for mobile use
while the one with CPU would be less amenable as a mobile device.
(Source: Microsoft Surface website)
The CPU based design
relies on extra chips to achieve the necessary hardware integration and
consumes more power, resulting in a 30% heavier and 40% thicker product
form factor. It is also interesting to note that the NVIDIA SoC design
is made using lagging 40nm lithography while the Intel CPU design is
made using the two generation more advanced 22nm lithography. The
comparison will only get more interesting next year when the SoC players
move to 28nm technology with hi-k/metal gate transistors. The Surface
tablet will serve as an important benchmark since it will provide a
direct comparison between the incumbent/leading-edge CPU and the
disruptive/trailing-edge SoC – not only in terms of functionality but
also in terms of cost.
The unique cost structure enabled by the
SoC has the potential to truly disrupt the business model in the
semiconductor industry. The ASP for the NVIDIA Tegra SoC chip is in the
range of $20 while the ASP for a leading edge Intel IvyBridge CPU chip
is in the $150 range. The CPU chip will also need to be supported by
other chips to provide the functionality that is provided by a single
SoC. When OEMs compete on price, it will be very difficult for the CPU
product to compete while retaining historically high profit margins. As
the SoC gets better and encroaches into the ultrabook and laptop space,
the cost differential will have an even larger impact. The rising
influence of a low-cost, low-end technology (SoC) and its potential to
eat into the profit margins of a high-cost, high-end technology (CPU) is
an example of the classic segment-zero phenomenon articulated by Andy
The NVIDIA Tegra 3 SoC and the Intel Ivy Bridge CPU
are both best-in-class products – but they are designed for different
form factors and cost and value metrics. While it is conceivable that
the low-margin SoC may be able to serve the high-end laptop market well,
it seems unlikely that the high-margin CPU will be able to serve the
low-end mobile market as well.
Click on image to enlarge.Figure
1 A typical CPU design (Intel Ivy Bridge) dominated by core/graphics
compared to a highly integrated SoC (NVIDIA Tegra). The integrated SoC
design has obvious advantages in the tablet and ultrabook formfactors
(Source: Intel/NVIDIA websites).
a strong and cash rich incumbent in the sustaining CPU market, Intel’s
response to this growing segment-zero threat should not be
underestimated. Intel is expected to address this threat head-on and has
taken steps to make up for lost time in winning mobile market share.
Earlier this year, Intel released Medfield, its first SoC processor
aimed at the smartphone market. Intel also released its first reference
phone design, enabling OEMs to quickly integrate smartphone products
based on Intel technology.
The 32nm Medfield processor is
expected to be replaced over the next year by an advanced 22nm processor
(Merrifield) which is expected to sport a dual core Atom CPU for lower
power. Even Intel’s flagship CPU products are seeing more on-chip
integration. The GPU block grew significantly (an additional 400 million
transistors) when Intel transitioned from the 32nm Sandy Bridge chip to
the more advanced 22nm Ivy Bridge chip shown in Fig 1.
has a significant lead in CPU process technology and is at the
forefront of Moore’s Law. However, radical changes to architecture (e.g.
Tri-Gate) may actually slow down the integration of on-chip
functionality. In spite of acquiring baseband technology from Infineon
in 2011, it is unclear when Intel will be able to integrate it with
Tri-Gate transistors on Atom cores. Intel’s SoC product offering has
traditionally lagged its mainstream CPU offering by 1-2 years. That gap
is expected to narrow in the coming years as Intel addresses the growing
need for on-chip integration and the growing threat from seemingly
low-end product offerings which are rapidly becoming more competitive
and cost-effective in the high-end.