Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to johnstevens77,Bhoddhisatva,scotia,Anonymous,Cornytiv34, for Donating to support the site

Nvidia

Analysing companies' finances and value from their financial statements using ratios and formulae
TheMotorcycleBoy
Lemon Quarter
Posts: 3245
Joined: March 7th, 2018, 8:14 pm
Has thanked: 2222 times
Been thanked: 587 times

Nvidia

#409018

Postby TheMotorcycleBoy » May 3rd, 2021, 4:12 pm

Hi folks,

I've been spending a few days so far investigating the technology company, Nvidia (NVDA: NASDAQ). This is a firm famous in the PC computing world, especially for graphics cards. They were founded in 1993, and started out in designing PC circuit (graphics) cards which added enhanced texturing of images whilst the computations involved in the creation of the graphics remained on the CPU. Interestingly, two of the founders, Jensen Huang (CEO and President) and Chris Malachowsky (Fellow) remain on board, something I find very encouraging. In the mid to late '90s the term GPU was borne. Some slight contention about exactly when the GPU was "invented" and by who. Nvidia were clearly fundamental in this. The key thing about the GPU is that bulk of the computational work in rendering imaginery, especially for computer games is off loaded onto this unit. The GPU differing from the CPU in the respects:

  • large amounts of matrix and other maths operations (planar and 3-D geometry) occur in parallel
  • specialised units for decomposing shapes into collections of triangular primitives
  • colouring, lighting, clipping etc. of backgrounds whilst a player's viewpoint changes realtime.
NVDA partner with TSMC 1998. NVDA, launches it's first GPU (GeForce) in 1999. From then we have a continuous evolution of ever sophiscated (and expensive!), dedicated graphics cards and GPUs from NVDA, as computer games themselves evolve. From 2000-2005 various partnerships and small acquistions. 2005 - the processor for Sony PS/3. 2005-2015 various mobile processing innovations, 2015 - "nvidia drive", deep learning, AI.
2019 - hpc, data center.

From https://www.nvidia.com/en-gb/about-nvid ... -timeline/. Also rans 2010 some participation in the worlds fastest supercomputer https://en.wikipedia.org/wiki/Tianhe-1/

From the above perusal of their potted history, it's clear that the diversifications into AI, and machine learning, whilst being diversifications, are very natural ones given the emphasis on parallelisation, floating point maths, matrix algebra.

As with similar companies e.g. AMD, this one has a very high valuation - about 87 x earnings as I type. Before looking at various bull and bear scenarios, I'll upload a few numbers. The firm are always outrageous cash rich, it's pointless including leverage or gearing KPIs as they are non-existent.



So whilst the current valuation is crazy high, the firm has fantastic fundamental values, with ROCE/ROE typically between 20-40%, top and bottom line growth CAGR over the past 5 years about 27% and 44% respectively. To me it also seems likely their growth will continue since their IP can naturally move between graphical (multi-dimensional) and AI worlds over time, in addition to reaping the rewards akin to cloud compute and data center growth.

Matt

TheMotorcycleBoy
Lemon Quarter
Posts: 3245
Joined: March 7th, 2018, 8:14 pm
Has thanked: 2222 times
Been thanked: 587 times

Re: Nvidia

#409044

Postby TheMotorcycleBoy » May 3rd, 2021, 5:29 pm

Still the valuation, IMO is too high. Of course this one is incredibly hard to analyse. Not only do we have much forecast growth priced in. We also have the announcement of the ARM acquisition in Sept 2020.

https://nvidianews.nvidia.com/news/nvid ... age-of-ai/

I'm curious as whether how much of this is embedded in the current price. This is definitely a value adding acquisition, there will be synergies, new market places, and merging of mobile, embedded, graphic etc. compute functionalities.

Image

However looking at the chart, it seems that the market have been more concerned by 1) earnings updates (Q2 and Q3 were good, quarter against comparable last year's quarter being up 26%-33% top and bottom, various measures, hence uplift/consolidation there), (Q4 not so good *only* 5%-9% up!) and 2) general SP500 madness. The green line marks the original announcement of the ARM deal, which is yet to go through, and may not happen.

Anyway valuations, $500-$545 seem more reasonable. Using $525 and the latest EPS of $6.90 we have PE=76. Wow still very high. Looking at tech firms with only about 5% CAGR (e.g. INTC, QCOM, CSCO) these have PEs between 12x and 22x. So for a PE of about 76, I guess we need about growth of 50%-60%.

Well from my spreadsheet NVDA has approached this kind of growth:



But whether this can continue is down to debate.

Matt
Last edited by TheMotorcycleBoy on May 3rd, 2021, 5:32 pm, edited 1 time in total.

TheMotorcycleBoy
Lemon Quarter
Posts: 3245
Joined: March 7th, 2018, 8:14 pm
Has thanked: 2222 times
Been thanked: 587 times

Re: Nvidia

#409045

Postby TheMotorcycleBoy » May 3rd, 2021, 5:32 pm

That's all for now. I'm still researching NVDA - and hoping for an SP500 crash of course to give me an opportunity.

Feel free to add. I'll try to write up some more in the future. Alas I'm pretty busy on other things presently.

Matt

PhaseThree

Re: Nvidia

#409055

Postby PhaseThree » May 3rd, 2021, 6:13 pm

Somewhat counter-intuitively I tink that the ARM aquasition may well harm Nvidia, if it is allowed to proceed. Although the ARM CPU architecture is very widely accepted and employed there are other games in town*. ARM exists by being a neutral player in that it licenses its IP to all and sundry (for a price). So currently NVidia, Apple and Samsung both license from ARM as a third knowing that their competitors don't get insider knowledge.
If Nvidia buys ARM then the relationship changes and all bets are off as to what happens next

* Risc-V is the largest alternative to ARM , is open source and widely supportd.. (https://riscv.org/members)

Technical analysis of the pros and cons here
https://www.nextplatform.com/2021/03/30 ... to-buy-it/

TheMotorcycleBoy
Lemon Quarter
Posts: 3245
Joined: March 7th, 2018, 8:14 pm
Has thanked: 2222 times
Been thanked: 587 times

Re: Nvidia

#409079

Postby TheMotorcycleBoy » May 3rd, 2021, 7:49 pm

PhaseThree wrote:Somewhat counter-intuitively I tink that the ARM aquasition may well harm Nvidia, if it is allowed to proceed. Although the ARM CPU architecture is very widely accepted and employed there are other games in town*. ARM exists by being a neutral player in that it licenses its IP to all and sundry (for a price). So currently NVidia, Apple and Samsung both license from ARM as a third knowing that their competitors don't get insider knowledge.
If Nvidia buys ARM then the relationship changes and all bets are off as to what happens next

* Risc-V is the largest alternative to ARM , is open source and widely supportd.. (https://riscv.org/members)

Technical analysis of the pros and cons here
https://www.nextplatform.com/2021/03/30 ... to-buy-it/

Thanks for the NextPlatform article, PhaseThree. I'm starting to read through it. Being in tech, I'm aware of Risc-V, and it's gaining *some* momentum, most crucially since adopters of this architecture can forgo the ARM licence fee.

However, the negatives there are the porting costs and loss of considerable developer tool advantages and familiarity, with regards the ARM stuff. Without getting too technical, the ARM CoreSight Debug https://community.arm.com/developer/ip- ... ics-part-1 architecture, which is part of their chip packages, is the most obvious one that springs to mind. My own personal experience of ARM cores is limited to the CortexM0-M4, but they do impress as being an incredibly compact, no-surprises, functional design.

But, TBH, I assume (without finishing my read of your article yet!) one the biggest pluses of the ARM deal, if it happens would be access to a larger market and picking up 1000s of highly skilled tech professionals, which are tough to hire at the best of times, and many of whom having decades of experience.

Thanks again for your thoughts,
Matt

TheMotorcycleBoy
Lemon Quarter
Posts: 3245
Joined: March 7th, 2018, 8:14 pm
Has thanked: 2222 times
Been thanked: 587 times

Re: Nvidia

#409084

Postby TheMotorcycleBoy » May 3rd, 2021, 7:56 pm

https://blogs.nvidia.com/blog/2020/05/2 ... sing-unit/
Now we have DPUs in addition to CPU and GPU.

TheMotorcycleBoy
Lemon Quarter
Posts: 3245
Joined: March 7th, 2018, 8:14 pm
Has thanked: 2222 times
Been thanked: 587 times

Re: Nvidia

#409089

Postby TheMotorcycleBoy » May 3rd, 2021, 8:08 pm

Unfortunately I've not got time to write up all that I'd like to right now. Here's a couple more links to some of my researchs,

Bull case:
https://seekingalpha.com/article/441282 ... gain-price
https://www.forbes.com/sites/paulteich/ ... -you-think

Bear case:
https://seekingalpha.com/article/441288 ... ytime-soon
https://seekingalpha.com/article/441666 ... -priced-in

UncleEbenezer
The full Lemon
Posts: 10691
Joined: November 4th, 2016, 8:17 pm
Has thanked: 1459 times
Been thanked: 2965 times

Re: Nvidia

#410251

Postby UncleEbenezer » May 8th, 2021, 12:27 pm

PhaseThree wrote:Somewhat counter-intuitively I tink that the ARM aquasition may well harm Nvidia, if it is allowed to proceed. Although the ARM CPU architecture is very widely accepted and employed there are other games in town*. ARM exists by being a neutral player in that it licenses its IP to all and sundry (for a price). So currently NVidia, Apple and Samsung both license from ARM as a third knowing that their competitors don't get insider knowledge.
If Nvidia buys ARM then the relationship changes and all bets are off as to what happens next

* Risc-V is the largest alternative to ARM , is open source and widely supportd.. (https://riscv.org/members)

Technical analysis of the pros and cons here
https://www.nextplatform.com/2021/03/30 ... to-buy-it/


As a former ARM shareholder who was reluctant to sell when Softbank bought, I agree entirely that a major part of its strength was its independence. No ARM licencee need fear discriminatory or anticompetitive behaviour from a rival, because ARM itself was visibly neutral. A position analogous to Microsoft's vendor-neutrality during the rise of the clone manufacturers in the 1980s and 90s, and ARM's product was immeasurably superior to Microsoft's.

If Softbank raised a questionmark, NVidia drives a bulldozer through that independence. It will surely concentrate minds in the rest of the industry and hasten the rise of Risc-V. But that's very much a long-term prospect: bear in mind that 30 years of Linux has yet to relegate MS Windows to the museum.

It's also far from the first spur to the industry to focus on an alternative. Most importantly, when Trump attacked Huawei, he demonstrated that ARM was vulnerable to political nonsense and couldn't be relied on. To be honest I think I'd be more afraid of Trump and his potential successors than NVidia - which is at least subject to regulators, and doesn't (so far as I know) have a powerful Fifth Column in our own parliament.


Return to “Company Analysis”

Who is online

Users browsing this forum: No registered users and 7 guests