Intel is the king of the tapering kingdom. Yes, most
processors on PC, laptops and servers are made by Intel, however the modern
market of products is not limited only to these segments. Annually millions,
tens of millions of new devices are equipped with chips of another company -
ARM. Even Apple and Google use chips based on ARM. The same Microsoft announced
Windows under ARM. More importantly, changes in the field of server needs, due
to the aggressive development of cloud technologies and services. And they all
need power, which Intel is not yet able to provide.
Of course, Intel is one of the most influential companies in
the field of processors. However, the future is not so categorical.
Throughout 2016 it was possible to observe the confident and
strong growth of shares of Nvidia and AMD. The cost per share of Nvidia grew by
250%, AMD and even higher - more than 370%. Both companies are now estimated to
be several times higher than a year ago. Thank you say to the AI market,
which uses sophisticated and resource-hungry machine learning systems and
in-depth training to solve problems. All these technologies require enormous
processing power. It so happened that AMD and Nvidia already have ideal
solutions: graphics cards.
Intel is the king of the tapering kingdom. Yes, most
processors on PC, laptops and servers are made by Intel, however the modern
market of products is not limited only to these segments. Annually millions,
tens of millions of new devices are equipped with chips of another company -
ARM. Even Apple and Google use chips based on ARM. The same Microsoft announced
Windows under ARM. More importantly, changes in the field of server needs, due
to the aggressive development of cloud technologies and services. And they all
need power, which Intel is not yet able to provide.
Of course, Intel is one of the most influential companies in
the field of processors. However, the future is not so categorical.
Throughout 2016 it was possible to observe the confident and
strong growth of shares of Nvidia and AMD. The cost per share of Nvidia grew by
250%, AMD and even higher - more than 370%. Both companies are now estimated to
be several times higher than a year ago. Thank you say to the AI market,
which uses sophisticated and resource-hungry machine learning systems and
in-depth training to solve problems. All these technologies require enormous
processing power. It so happened that AMD and Nvidia already have ideal
solutions: graphics cards.
GPU or graphics processor blocks can be considered as a
separate type of CPU. Unlike conventional processors, graphics chips can not
boast of a wide range of tasks, but they are ideal for carrying out a huge
number of parallel tasks. Graphics processors are not equipped with six or
eight cores ... they have hundreds and thousands of them. Moreover, each core
is incredibly effective in solving small repetitive tasks again and again.
Graphics accelerators have already become a standard in machine learning, and
the new Google cloud platform coming out this year will be equipped with AMD
FirePro server graphics cards.
AMD already has a whole line of GPUs specifically geared for
accelerating machine learning. It is called Radeon Instinct. And they can be
used to solve many problems: autonomous transport, autonomous drones, personal
robots for financial companies, nanobots, advanced medicine. Investors buy up
AMD shares because they know - the future behind the huge power of thousands of
parallel computing on a single GPU.
Since Nvidia is also engaged in graphics and also works in
the field of machine learning and AI, the demand for AMD has also pulled
interest in Nvidia. Moreover, Nvidia and IBM have already announced the
creation of the world's fastest business solution for in-depth training.
In addition, both companies are growing and at the expense
of consumers' interest in the virtual reality market. This is due to the
release of Vive and Oculus Rift in 2016. Both VR-solutions for a comfortable
game require top-end iron. And everyone interested in VR is a potential AMD or
Nvidia client, so they can produce even more video cards.
Who would have thought that such a giant as Intel would
quickly lose a new gold mine from its own hands, which will feed all our
computing needs in 5-10 years.
Comments
Post a Comment