NVIDIA Corporation: A Strategic Audit - DigitalCommons@University of Nebraska - Lincoln - UNL Digital Commons
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln Honors Theses, University of Nebraska-Lincoln Honors Program Spring 5-2018 NVIDIA Corporation: A Strategic Audit Daric Teske University of Nebraska-Lincoln Follow this and additional works at: https://digitalcommons.unl.edu/honorstheses Part of the Business Administration, Management, and Operations Commons, and the Technology and Innovation Commons Teske, Daric, "NVIDIA Corporation: A Strategic Audit" (2018). Honors Theses, University of Nebraska-Lincoln. 18. https://digitalcommons.unl.edu/honorstheses/18 This Article is brought to you for free and open access by the Honors Program at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Honors Theses, University of Nebraska-Lincoln by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln.
NVIDIA CORPORATION:
A STRATEGIC AUDIT
An Undergraduate Honors Thesis
Submitted in Partial fulfillment of
University Honors Program Requirements
University of Nebraska-Lincoln
by
Daric Teske
Computer Science
College of Arts and Sciences
April 5, 2018
Faculty Mentors:
Samuel Nelson, PhD, ManagementAbstract
NVIDIA is company that has been involved with the production of computer hardware
known as a graphics processor units (GPUs) since the turn of the century. Over the past decade,
NVIDIA has begun to explore other applications of its proprietary technology, including the
fields of artificial intelligence (AI) and accelerated computing. Looking from a business
perspective, this strategic audit analyzes the history of the company and its current state, as
well as industry outlooks to provide a strategic recommendation. Annual reports, market
reports, and press releases were all utilized to provide the best perspective of the industry
possible. Though NVIDIA has been performing extremely well recently, a reorganization was
proposed. This reorganization moves resources such as people and products around so that the
business can operate more efficiently and make the research and development process more
streamlined than it was previously.
Key Words: NVIDIA, strategic, audit, business, strategydemanding better performance. NVIDIA
Background offers the GeForce GTX line for PC gaming,
SHIELD devices for gaming and streaming,
History
GeForce NOW for cloud-based gaming, and
Founded in 1993 by Jensen Huang, Chris
often partners with console producers to
Malachowsky, and Curtis Priem, NVIDIA
develop a custom graphics solution.
entered into a highly competitive market
with as many as 70 different competitors. In Professional visualization is another
1995, NVIDIA released NV1, the company’s industry that is growing rapidly. Industries
first product, and has grown steadily since like architectural design, medical
then. With a newly invented graphical instrumentation, and aerospace
processing unit, or GPU, in 1999 NVIDIA development require designs and
went public at $12 per share, and was added simulations that are nearly identical to our
to the S&P 500 in 2001. Releasing the physical world. Digital content creators rely
Compute Unified Device Architecture, or heavily on graphical processing through
CUDA in 2007, NVIDIA enabled calculations video-editing and post production, special
other than graphics to be run on the GPU, effects, and broadcast-television graphics.
massively increasing the potential uses for NVIDIA offers the Quadro for this market, a
their chipsets. NVIDIA chips could now be card designed to be used in these high-
used effectively in supercomputers, performance workstations.
autonomous vehicles, and mobile devices
The datacenter market encompasses
too (NVIDIA History, 2018). With the rise of
NVIDIA’s offerings for the AI and
cryptocurrency, autonomous vehicles, and
supercomputer needs. CUDA enables the
virtual reality in recent years, more and
calculations that are required to be
more computational power is needed. GPUs
offloaded to the GPU, and can process many
have become the main source for this
of them at once, significantly speeding up
computational power, and NVIDIA is
the process. As development for these
uniquely positioned to take advantage.
technologies ramps up, computers
Markets accelerated with NVIDIA GPUs are proving to
Today, NVIDIA divides their business into be more and more useful. For NVIDIA, this
four distinct markets — gaming, professional market consists of their energy efficient
visualization, datacenter, and automotive — GPUs, the CUDA programming language,
offering a unique set of products branded specific machine learning libraries, and
specifically for the challenges faced in each software products that allow applications to
market. run over multiple GPUs, as well as
organization-based and cloud-based
The computer gaming industry is the largest supercomputing clusters.
entertainment industry, and many factors
propel growth. Virtual and augmented Automobiles are relying on technology more
reality, eSports, and online competitive heavily today than ever before, especially as
gameplay are a few reasons gamers are ever strides towards fully-autonomous vehiclesare being made. Autonomous vehicles will energy-efficient embedded AI computing
rely heavily on sensor data being interpreted platform.
in real-time to make the best possible driving
decision every second. To this end, NVIDIA Major Issues
offers several solutions under the DRIVE The current strategy for NVIDIA hinges on a
brand, ranging from infotainment solutions few key focuses: advancing the GPU
and advanced driver assistance systems to computing platform, extending its
complete autonomous vehicle systems. technology and platform leadership in AI and
NVIDIA also has the ability to provide over- visual computing, advancing the leading
the-air updates, adding new features and autonomous vehicle platform, and
capabilities to these vehicles (NVIDIA leveraging the intellectual property it
Corporation, 2018). already owns. NVIDIA does not manufacture
anything directly, instead designs are sent to
Product Lines suppliers specializing in the technologies
Though NVIDIA offers a wide array of end- needed to create these devices, and the
products, they are all based on a single products are then tested for quality control.
underlying architecture, allowing the
There are common issues that affect all tech
company to customize the offerings to their
companies, like protecting intellectual
specific application without as significant of
property, maintaining strong system security
an investment. NVIDIA divides their products
and data protection, and retaining key
into two business segments: the GPU and
employees. As a multinational corporation,
the Tegra processor (NVIDIA Corporation,
NVIDIA is also subject to the issues that arise
2018).
when operating internationally, such as
The GPU segment includes the GeForce and different tax liabilities and government
GeForce NOW brands for desktop and cloud policies. However, there are a few major
gaming, the Quadro brand for design issues that NVIDIA specifically needs to focus
professionals, the Tesla on. Adapting to the needs of
“[Everything is] based on a single
brand for AI and deep the market and identifying
underlying architecture, allowing
learning, the GRID brand for new products, services, and
[NVIDIA] to customize the
datacenters, and the DGX technologies is crucial for
offerings…”
brand for AI scientists and NVIDIA to stay ahead in the
researchers. market. If it fails to do this, it
will stagnate, and become obsolete. NVIDIA
The Tegra Processor segment includes the
also faces high levels of competition and
DRIVE brand for automotive
must deliver superior products consistently
supercomputers, the SHIELD brand for
to remain an industry leader. A part of the
mobile-cloud entertainment and gaming,
reason NVIDIA has become so successful is
the Tegra processor brand that enables
consistency, so if the quality drops in either
branded platforms like DRIVE and SHIELD,
the design or the third-party manufacturing,
and the Jetson brand that provides an
there will be considerable damage to thebrand, eroding the consumers’ trust in NVIDIA does not own any manufacturing
NVIDIA products. Another major issue is facilities; instead, it engages in long term
being able to meet consumer demand for contracts with several of the industry’s top
the product. If NVIDIA is unable to do that, suppliers for all parts of the manufacturing
consumers will likely turn to competitors for process. As a result, it can leverage the
similar products (NVIDIA Corporation, 2018). suppliers’ expertise to create the highest
quality products possible. With
Situation Analysis manufacturing technology constantly
NVIDIA’s actual business operations are improving, NVIDIA doesn’t have to expend
uncomplicated: it designs new chips and time and capital constantly investing in the
systems for the various target areas, ships latest technology and can instead focus on
the designs to manufacturers, and then tests providing designs that take advantage of
the results, iterating on the design and said technology. On occasion, NVIDIA will
incorporating the latest manufacturing work very closely with a specific
technology to produce the best possible manufacturer to release custom editions of
chipsets each year. This contributes heavily its products as well, but the work is still not
to NVIDIA’s brand and reputation, as it done by NVIDIA (NVIDIA Corporation, 2018).
consistently churns out high-performance, External Analysis
reliable chips year after year. NVIDIA is also
NVIDIA mainly competes in two different
very intentional with branding, often
segments, GPUs and embedded solutions,
keeping their logo visible whenever possible.
which aligns with its product segments. In
The resulting value created for the customer
the GPU segment, competitors include AMD,
is that the brand is recognizable, reliable and
ARM Holdings, and Intel. These competitors
performant. Even if competitors release
produce discrete and integrated GPUs and
chips that out-perform NVIDIA’s, consumers
accelerated computing solutions. In the
are likely to still choose NVIDIA due to its
Tegra segment, competitors include AMD,
high-quality reputation. By engaging in
Broadcom, Intel, Qualcomm, Samsung, and
partnerships with various companies,
Texas Instruments. These competitors
NVIDIA also can design custom chips for
produce devices that are embedded into
unique applications, delivering a superior
automobiles, autonomous vehicles, smart
experience to the end user, and further
televisions, and gaming devices (NVIDIA
increasing its reputation.
Corporation, 2018).
NVIDIA spends a bulk of its time in research
NVIDIA competes on two main fronts:
and development activities, as well as
quality and service. Because of lofty
marketing its new innovations. The company
standards and thorough testing at every step
also focuses on providing superior service to
of the process, NVIDIA products are trusted
the end users. This is most visible through
to have high degrees of performance and
the frequent releases of “game-ready”
reliability. Drivers are frequently updated,
drivers, where popular video games are
optimized to run on NVIDIA’s GeForce GPUs.industry rather than the cryptocurrency
Q4 2017 Total GPU Market Share
market. Due to high demand and the long
production cycle for video cards, there is
currently a GPU shortage, causing mid- to
AMD
14% high-tier cards to sell significantly above the
original price point. This accounted for
NVIDIA inflated revenue for both NVIDIA and AMD
18% over the past quarters. However, if the
Intel
downward trend of cryptocurrency stocks
68% over Q1 2018 continues, the demand should
return to more reasonable levels
(Cryptomining Inflated, 2017).
Because of its global nature, NVIDIA must be
Figure 1: Intel's large market share is attributed to
integrated graphics, where graphics technology is included
wary of various international policies,
on the CPUs. Integrated graphics have low performance especially those concerning protection of
and can only accomplish basic computing tasks. intellectual property. Different countries
and a suite of software products exist to offer different protections, with the United
better manage NVIDIA’s offerings. States having some of the strictest policies,
and most easily enforceable. In countries
The GPU industry contains three major overseas, where the chips usually produced,
competitors: NVIDIA, Intel, and AMD. Intel the policies are not as stringent, so NVIDIA
held a 67.4% market share in Q4 2017, due must constantly evaluate the necessary
to dominance in the CPU market and the measures to take to protect their designs
default inclusion of integrated graphics. (NVIDIA Corporation, 2018).
NVIDIA held 18.4% of the market, and AMD
held the remaining 14.2%. When considering Privacy is a major social issue within the
only discrete desktop GPUs, NVIDIA holds a technology industry. Major breaches of
66.3% market share to AMD’s 33.7%. Market social trust, where companies are not
share is down over the last quarter for protective enough of their user’s data, are
NVIDIA, due to seasonality of the industry likely to become more common, as people
and the pressures of cryptocurrency mining continue to put more of their lives online.
demand (GPU Market, 2018). Luckily chip makers like NVIDIA, AMD, and
Intel are mostly protected from these issues
Mining cryptocurrency is a computationally because their focus lies in the low-level
expensive process, so miners use GPUs to hardware they produce. However,
accelerate those calculations. AMD’s supplemental software has fallen under
hardware has a performance advantage of scrutiny due to concerns about the type of
NVIDIA’s, and as a result AMD was able to data being gathered. NVIDIA has promised
take market share away (Mujtaba, 2018). that any personally identifiable data will
Both companies have announced that they never be shared, only allowing partners to
intend to focus on the consumer gaming see aggregated and anonymized statisticslike gameplay and hardware configuration
(Frequently Asked, 2018). Q4 2017 Discrete Desktop GPU
Market Share
Artificial intelligence is a rapidly growing
technology and popular research topic. This
broad area can cover anything from AMD
autonomous vehicles to enhanced modelling 34%
and decision making. The most practical
aspect of AI is currently neural networks, a
machine learning algorithm that is trained to NVIDIA
recognize patterns between an input and an 66%
output and can then begin to predict outputs
of any input. The parallel-processing-nature
of GPUs allow for thousands of these
Figure 2: Discrete graphics are used for graphics when
calculations to be run simultaneously, higher performance is necessary, such as video games or
greatly increasing the viability of AI-related image processing, and this is where NVIDIA excels.
algorithms. NVIDIA made this a possibility
the cost of switching for consumers is little
with the original development of CUDA in
to none. Discrete GPUs have been
2007, which unlocked the potential for GPUs
developed with standard interfaces, so
to be used for purposes other than graphics.
switching is as easy as purchasing a different
More recently, NVIDIA has focused on
one. Quality has been a big focus for NVIDIA,
offering customized products for a variety of
and as a result it has the reputation of
AI applications, and is the only one out of its
producing more reliable chips. When
competition to do so.
considering GPUs of similar capabilities,
For the GPU industry, there is relatively little consumers often choose NVIDIA even
threat of new entrants into the market. The though the price is typically higher because
capital investments required to design and of the reliability. Overall, it appears good to
produce GPUs has created a huge barrier to be in this industry if the competition with
entry. The threat of substitution is higher, rivals remains manageable.
but still low because people are going to
need GPUs for performance reasons. Given Internal
the high number of suppliers in the industry, In FY 2018, NVIDIA reported $3.8 billion in
they have relatively little bargaining power. total liabilities and $7.5 billion in total
Suppliers must continually upgrade to the stockholder equity, giving a debt-to-equity
latest technology to be able to compete for ratio of 50.7%. This is down over the
contracts. Buyers also have relatively little previous year’s ratio of 70.3% and can
bargaining power because there are only a mainly be attributed to a $1.7 billion
few options available, and this is a necessary increase in retained earnings over the past
year without any other significant changes.
component for computing. The competitive
This is still indicative of a financially stable
rivalry is still high between the few
business, though FY 2018 benefited heavily
companies in the industry, mostly becauseNVIDIA Annual Revenue by Segment
12
Billions
10
8
6
4
2
0
2014 2015 2016 2017 2018
GPU Tegra Other
Figure 3: Revenue has increased over the past 5 years for NVIDIA, with GPUs making up a large portion of that revenue.
from a revenue stream that has proven to be to attract that crowd and see increased sales
extremely volatile: cryptocurrency. Revenue as a result.
in the GPU segment has increased from $4.2
Assets
billion in FY 2016 and $5.8 billion in FY 2017
NVIDIA often takes advantage of leasing
to $8.1 billion in FY 2018, due to strong sales
whenever possible. The only properties it
of high-end graphics and overall strong
owns are its headquarter complex in Santa
demand from the market. It is estimated
Clara, California, and a research and
that revenue from GeForce GPU products
development facility in Hyderabad, India. It
increased by over 90% due to
leases various research and development,
cryptocurrency mining. Revenue in the Tegra
sales, and administrative offices around the
market has increased from $559 million in FY
United States, Europe, and Asia, and
2016 and $824 million in FY 2017 to $1.5
billion in FY 2018, indicating that devices and datacenters around the world. Patents are a
embedded platforms tailored for specific significant intangible asset for NVIDIA, with
applications have been successful and expiration dates ranging from April 2018 to
demand is increasing for these customized January 2037. These patents range from
solutions. Operating results have fluctuated basic GPU technology to the customized,
in the past for NVIDIA, and seasonality plays application-specific devices. Reputation is
a significant role in this. Sales tend to pick up also a significant intangible asset for NVIDIA,
in the second half of the fiscal year, and this protecting it by enforcing rigorous quality
trend can be seen across the computer testing. NVIDIA must litigate to protect its
entertainment industry. Many video game intellectual property but is subject to similar
producers and hardware manufacturers suits from competitors and the possibility of
target release dates near the holiday season unfavorable outcomes. NVIDIA employsover 10,000 people, with about 70% covers a good portion of the card’s expected
engaged in research and development, and lifetime. Intentional branding has also
about 30% engaged in sales, marketing, served NVIDIA well, and it seems to be
operations, and administrative positions something that it does better than the other
(NVIDIA Corporation, 2018). This high major players. Even though every single one
number of employees means the company of NVIDIA’s products are based on the same
can target many different areas that would underlying architecture, each device has
benefit from GPU technology, but means the been optimized for a specific function, and
company might be slower to adapt to a an appropriate brand has been assigned.
changing marketplace, and that any GeForce for gamers, Quadro for content
restructuring efforts could prove to be creation, Tesla for scientists, and DRIVE for
costly. autonomous vehicles are all based on a
single platform, customized to provide the
Competitive Advantage
best experience for the intended use. As a
To be successful in this industry, all
result, reports like a GeForce product
competitors must demonstrate
functions horribly in a self-driving car or a
competencies to adapt quickly to the latest
Quadro product can’t run the newest AAA
manufacturing methods and to continually
video game rarely surface. Instead, the
improve performance. Moore’s law
products perform excellently at the
estimates that performance of computers
expected task, further increasing NVIDIA’s
doubles about every two years, due to
reputation. Herein lies NVIDIA’s competitive
improvements in fabrication methods and
advantage: with superior branding,
energy efficiency (Britannica, 2017).
reputation, and support, consumers are
Competitors must make regular
willing to pay more for a product by NVIDIA
performance increases to keep the
than a similar product from a competitor.
hardware they produce following this trend.
As a company, NVIDIA still has room to grow.
NVIDIA also excels in a few distinctive
It has been very successful in the discrete
competencies. Support of its products is a
GPU market and has started to branch out
main one. Driver updates are frequently
into other areas. This seems to be a good fit
released, making sure all devices can take
for where the organization is at, and founder
advantage of the latest software
Jensen Huang still seems to be the correct
optimizations. Long-term warranties, lasting
person to lead this company. Huang is best
between 1 and 3 years based on the device,
described as a visionary-type CEO, focused
are also a great example of support. The 3-
on new and exciting innovations that will
year warranty applies to all the discrete
allow NVIDIA to continue to grow. As a
GPUs and covers hardware and
company, it is still expanding into new
manufacturing defects (Frequently Asked,
markets and exploring new products, so
2018). Common advice is to upgrade GPUs
overall this seems like a good fit.
when a new card can double or triple the
performance of an old card. With that
happening every 2-3 years, this warrantyStrategy Recommendation significant investment required to develop
specialized hardware like this, it’s possible
Goals the market could be non-existent by the
NVIDIA should have two major goals moving time a product would be released.
forward. The first one is to maintain its Additionally, NVIDIA has released
discrete GPU leadership (Chauhan, 2017). statements publicly declaring it would
For this to happen, NVIDIA will have to support the gaming industry over
consistently provide performance increases cryptocurrency, so this action would likely be
in line with competitors. Two metrics to seen as a hypocritical one and could tarnish
track progress on this goal are market share NVIDIA’s reputation.
and benchmarks. NVIDIA has held a majority
market share for over a decade, so it is Another alternative would be to expand into
important to maintain. Benchmarking is a CPU development with the purpose of
process that tests a GPU in various situations building more efficient CPUs for accelerated
and reports a score that can be used in direct computing purposes. NVIDIA has
comparisons with other GPUs, so metrics experimented in this area with the
should be used to make sure performance development of the Tegra processor for
improvements continue to match pace with inclusion on the small and energy-efficient
competitors. Acceptable benchmarking embedded platforms, but the development
scores and market shares will ultimately could be expanded to more intensive
depend on the specific recommendation. applications like data centers where physical
The other major goal for NVIDIA will be to size is not as much of a limiting factor. Given
grow the AI and accelerated computing that NVIDIA’s main competitors AMD and
markets. Over 80% of NVIDIA’s revenue Intel are already heavily invested in CPU
comes from the GPU business segment, and development, it is unlikely that NVIDIA
it can be assumed that a majority of this would be able to compete with the
revenue stems from discrete GPU sales economies of scale that exist. Since CPU and
(NVIDIA Corporation, 2018). By diversifying GPU design are fundamentally different, it is
NVIDIA’s revenue streams, it won’t be as also unlikely that NVIDIA would be able to
heavily affected in the future by rapidly use any intellectual property to give it an
changing and highly violate markets such as advantage.
the cryptocurrency mining market. In 2007, CUDA was first released, opening
the first potential for accelerated computing
Alternatives
on a GPU. Ten years later, serious advances
Creating hardware for the specific purpose
have been made, but there are more
of cryptocurrency mining would be one way
applications that could benefit from GPU
to grow and maintain market share in the
technology outside of AI and computer
GPU market. This would allow NVIDIA to
vision. NVIDIA could work on exploring other
recapture market share from AMD, who has
applications, outside the current areas. This
released such devices. However, with the
type of experimental research is difficult to
recent downturn of the market and the
justify to shareholders, though. Since thecompany already has 70% of employees grow, it will be important for NVIDIA to
engaged in research and development accept technological advances that might be
activities, it is unlikely that an increase in only be beneficial to one area. NVIDIA should
resources would yield significantly different restructure the company and resources to
results. be better aligned with these markets.
Operations should be reorganized into two
Recommendation separate segments, and the product lines
NVIDIA has been performing well recently, redistributed accordingly. Furthermore,
however some improvements could still be research and development on the Tegra
made, especially in positioning for the processor should taper-off, as GPU and CPU
future. Back in a 2002 interview with Wired design are significantly different, and it is not
Magazine, Jensen Huang stated, “Anywhere the best use of NVIDIA’s resources, when
there's a pixel, that's where we want to be.” partnerships with CPU manufacturers can
With recent announcements of customized easily be reached.
gaming monitors, it seems that this vision is
still holding true, however there is so much This proposed restructure would allow
potential from other markets to explore. NVIDIA to co-locate people and resources
NVIDIA needs to redefine the focus as an dedicated to solving very similar problems
enterprise to be narrower. The goal should and developing products and applications
be to advance the fields of gaming, artificial with very clear overlapping concerns in close
intelligence, and accelerated computing proximity to each other. Likely, this
using its unique technology, organization is already used to
without becoming distracted by “Anywhere with a some degree, but making it more
development of peripherals. In pixel, that’s where intentional could enable increased
relation to other organizations, we want to be.” productivity, since more co-
NVIDIA must continue to seek out - Jensen Huang,
workers would be working in the
partnerships, within the gaming Wired 2002 same domain and have a better
and autonomous vehicle industries understanding of the challenges
especially. Remaining relevant in these faced. There also may come a time when the
industries allows NVIDIA to reinforce the underlying architecture will need to split,
brand name it has created for itself. and if that time comes, allocating resources
like this will enable a more efficient
NVIDIA appears to have two main areas it is transition. By making these two segments
focusing on: graphics and accelerated more independent, performance can be
computing. To date, it has been able to tracked easier and adjustments can be made
utilize a unified underlying architecture and easier to achieve a more even distribution of
apply it to two very different use cases, revenue than the estimated 85-15 split
however this might not always be possible. currently seen between the GPU and Tegra
Changes in architecture for one application segments of today.
may not be beneficial for the other
application. As these markets continue toImplementation back towards graphics. Losing the edge
Restructuring and reallocation of resources NVIDIA has maintained over the past 15+
can sometimes be costly. However, since years in both market share and performance
NVIDIA’s main resources are people, it is a would severely impact the profitability of the
little easier. A plan should be put in place to company. As research turns more industrial
move people around so that they can be in in the AI and accelerated computing fields,
close proximity to those working on similar there will be potential applications for
areas. The timeline could vary depending on NVIDIA to capitalize on. However, if one year
the extensiveness of the reorganization of goes by without seeing sustained growth
the work spaces. Moving desks or buildings quarter over quarter, the company should
in the same complex could be accomplished re-evaluate the profitability of those
in hours or days. A typical remodeling markets. These will serve as the trigger
project could take up to 3-4 months to points to reallocate resources back towards
complete, if the working spaces themselves the more traditional structure the company
need to be adjusted. Relocation into other had, focusing more heavily on the gaming
states or countries would probably not be and design markets where profits have been
necessary given modern levels of the strongest historically.
telecommunication connectivity, but could
prove to be beneficial, and would likely take Conclusion
6 months to a year before transitions could NVIDIA is performing well as a company,
be completed. enjoying increased revenue and high
demand for their products. A
Since this recommendation looks at
recommendation to shift the business
adjusting the relative allocation of resources
segments and accompanying resources was
within NVIDIA, some contingency plans
made to increase efficiency and streamline
should be developed to address possible
operations. This recommendation will also
negative impacts. One possible impact with
allow the company to be more transparent
the increased focus on AI, accelerated
in its resource allocation, further increasing
computing, and other similar areas is that
stockholder confidence. For the most part,
the gaming and design markets could suffer.
no changes to NVIDIA’s operations were
As mentioned previously, market share and
recommended due to excellent
benchmarking should be used to determine
performance. Overall, NVIDIA is well-
when changes to this plan are necessary. If
positioned to remain a leader of the
market share drops below 55% in discrete
graphics, artificial intelligence, and
GPU sales or if competitors’ benchmark
accelerated computing industries.
scores exceed NVIDIA’s consistently at a
similar price, resources should be focusedSources
Britannica. (2017, November 29). Moore's law. Retrieved April 02, 2018, from
https://www.britannica.com/technology/Moores-law
Chauhan, H. (2017, December 07). Nvidia Is Running Away With the GPU Market. Retrieved
April 02, 2018, from https://www.fool.com/investing/2017/12/06/nvidia-is-running-
away-with-the-gpu-market.aspx
Cryptomining Inflated Prices Affect Q4 2017 Discrete Graphics Card Shipments, AMD Climbs to
33.7% Market Share. (n.d.). Retrieved April 02, 2018, from
http://www.anandtech.com/show/12499/q4-2017-discrete-graphics-card-shipment-
report
Frequently Asked Questions About GeForce Experience. (n.d.). Retrieved April 02, 2018, from
https://www.nvidia.com/en-us/geforce/geforce-experience/faq/
GPU market declined seasonally in Q4; cryptocurrency provides smaller offset as AIB prices rise.
(n.d.). Retrieved April 02, 2018, from https://www.jonpeddie.com/press-releases/gpu-
market-declined-seasonally-in-q4-cryptocurrency-provides-smaller-offset
Mujtaba, H. (2018, February 26). AMD GPU Market Share Increase Big Time in Q4 as NVIDIA /
Intel Decline. Retrieved April 02, 2018, from https://wccftech.com/amd-nvidia-intel-q4-
2017-gpu-market-share/
NVIDIA Corporation. (2018). 2018 10-K form. Retrieved April 02, 2018, from
http://investor.nvidia.com.
NVIDIA History. (n.d.). Retrieved April 03, 2018, from
http://www.nvidia.com/page/corporate_timeline.html
Wired. (2002, July 01). Nvidia. Retrieved April 02, 2018, from
https://www.wired.com/2002/07/nvidia/You can also read
