Entering text into the input field will update the search result below

Nvidia: Whatever It Takes

Jun. 09, 2020 1:07 PM ETNVIDIA Corporation (NVDA)54 Comments
MangoTree Analysis profile picture
MangoTree Analysis


  • NVDA shares continue to rip higher, with the chip stock's market cap surpassing $200 billion.
  • I am reasonably constructive on NVDA fundamentally. My problem mostly lies in the valuation of the stock. The valuation is getting a little ridiculous on both revenue and EPS.
  • NVDA's launch of the Ampere-based A100 accelerators will likely keep its reaccelerating datacenter business strong. These accelerators have a large performance lead over the previous generation V100 and T4 accelerators.
  • Gaming could come in a little light, at least until CQ3, when NVDA is likely to launch its Ampere-based gaming offerings (RTX 30 Series cards). And when the next gen gaming consoles release, Nintendo Switch could suffer, hurting NVDA.
  • Reiterating HOLD. PT raised from $270 to $300.

Datacenter Trends Accelerating at Nvidia

The vast majority of positivity I notice in the sell-side community on Nvidia (NASDAQ:NVDA) has to do with the company's ever-improving footing in the datacenter market. The datacenter was a reason for me being negative on the stock a few years ago, as some of their hyper scale clients as well as other smaller startups began developing competitive ASICs (application specific integrated circuits). At the time, we didn't have much data on the performance differentiation between Nvidia's GPU accelerators and these new ASICs, except really for the TPU. All three generations of Google's (GOOG) (GOOGL) TPU processor were faster and more efficient than even Nvidia's top-of-the-line GPU offerings.

Since then, however, my mind has been changed by two things: Understanding Nvidia's deep software ecosystem and the launch of the A100 accelerator. First of all, Nvidia's CUDA ecosystem is an advantage that no other competing silicon manufacturer can replicate. Nvidia has millions of developers working in the CUDA software ecosystem, an ecosystem that is very difficult to leave. It will take time and resources for other competitors to attempt to match the size of the CUDA software ecosystem.

With regards to the A100 accelerator, Nvidia is clearly not only ahead of the competition but has also been leapfrogging from generation to generation.

(Source: Nvidia)

(Source: Nvidia)

(Source: Nvidia)

In high-performance computing, inference, and training tasks, Nvidia's latest A100 offering is a generational leap. Remember last year when Nvidia saw an extremely weak datacenter business? The datacenter Capex cycle is a build-and-digest cycle. With the release of the A100 accelerator and A100-based systems, we are likely to see a spurt of growth as datacenters begin adopting these extremely high performance accelerators. So, I believe Nvidia's datacenter business, which has been experiencing rapid growth in recent quarters, will continue to grow quickly in the quarters

This article was written by

MangoTree Analysis profile picture
(10/1/23): MTA is focused on L/S ideas primarily in the TMT space. MTA's investment process is environment agnostic, and views each opportunity through the lens of generating alpha against whatever timetable is appropriate for the idea. MTA employs a mock sell-side coverage process, utilizing valuation methodologies, ratings and price objectives, and takes advantage of alternative data.

Analyst’s Disclosure: I am/we are long GOOG. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

I am not a financial advisor. This is not financial advice. Please do not interpret this as financial advice. Do your own due diligence before initiating a position in any of the securities mentioned.

Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

Recommended For You

Comments (54)

Michael Orwin profile picture

I found an application of AI I like, because it saves lives (at least potentially), "Israeli AI device 'watches' swimming pools to prevent drowning accidents" www.jpost.com/... (Jan 2019). It's for residential pools, and it isn't like pool alarms that go off when something lands in the pool.
puddnhead profile picture
Cool! This is just one of many applications that have already yielded positive human health & safety benefits. We even had a CEO of a nonprofit post post here, maybe around 2017, about how they were using tegra-powered tablets to diagnose eye diseases for referral in remote locations where there wasn't the human medical expertise available.

In fact arguably that is the biggest ultimate payoff from automated driving, automatic braking, etc -- cutting down on deaths and injuries by reducing the number of accidents.
Luke2020 profile picture
NVDA is nothing more than a TULIP MANIA BUBBLE! The stock is trading at Forward P/E=+120.
puddnhead profile picture
It's my observation that most people here who cite the Dutch tulip bulb craze of a few centuries ago actaully have almost zero understanding of what it was all about.

Can someone please explain to our addled friend that hyperscalars, gamers, etc are buying GPUs for specific & productive application, and not to resell like beanie babies, or plant in the ground and behold their beauty?
Michael Orwin profile picture
The future is AI. The future was never tulips.
Michael Orwin profile picture
"There Never Was a Real Tulip Fever" www.smithsonianmag.com/... has a bit near the end about the historian Simon Schama saying that popular accounts of tulip mania come from moralistic pamphlets by Calvinists.
Not that it's relevant to investing.
puddnhead profile picture
FWIW, just sold a single-digit percent of my position above $372. My first sale since February I think? "Bird in the hand" & all that.
grxbstrd profile picture
how does that feel? :)))))))
grxbstrd profile picture
@kvimax talked me into selling some $410 Jan calls a while ago, I though $400+ no way, those shares are safe. $375 this am and climbing. sheesh! I have to admit to banking some profits as well. It's always fun to sell at ATHs!
puddnhead profile picture
It actually feels really good right now, these are shares I bought for $8. But ask me again in a month (or week, or day, or hour? lol).

I certainly am sometimes wistful about the shares I sold in Jan-Feb between $240 and $315. But I have to keep reminding myself that the only reason I even had those shares is because I stuck my neck way out at the end of 2018, and sticking with that too long was starting to compromise my SWAN abilities. If I had just been a little bolder in March, those sales could have looked very smart in hindsight. I guess, thinking about it now, selling in Jan-Feb completely innoculated me from even thinking about selling in March-April, that alone was worth it.

I'm not gonna pretend I have any idea if it hits $300 or $450 first. The people say they are confident about either target simply mystify me.
DimaP123 profile picture
The article is good, but Geforce Now (online gaming) is not mentioned. At this point it is a large money pit for NVidia (no revenue, large data centers, R&D, etc). But a potential is great, especially when 5G comes into to play. Everyone want to take shares from Steam.
Also, AMD (not mentioned) has started doing the same as NVDA in data centers but at much cheaper prices and comparable performance as top NVidia cards. Yes, CUDA is a way more developed, but at the end of the day money talks. NVDA has VERY power hungry laptop GPUs. Have you seen power supplies sizes? My brand new laptop has RTX 2080 and power supply is a brick and a whole thing heats up really well. It is a disadvantage for consumers. Same AMD powered laptops have it all much more compact in all aspects.
What I'm saying, NVidia is a really great company (historically I have all my PCs and laptops with NVidia and Intel since Riva 128 days if anyone remembers such thing), but there is a "new" guy (AMD) that is *very* aggressive and taking shares of the market from NVDA and INTC. That needs to be accounted in the analysis. Nearest major competition is not miles away. Not at all.
grxbstrd profile picture
@DimaP123 "AMD (not mentioned) has started doing the same as NVDA in data centers but at much cheaper prices and comparable performance as top NVidia cards. "

AMD is not in the same league. The author spends some time talking about NVDA's software moat, CUDA. You spend little time talking about CUDA, "at the end of the day money talks"? You clearly have no clue. Lisa hasn't been investing at any where near the rate needed to compete.

"there is a "new" guy (AMD) that is *very* aggressive and taking shares"

LOL, what share? Historically AMD ships 3 for every 7 dGPUs NVDA ships. Please articulate how that's materially changed recently.

NVDA has NO competition. Not AMD, not Intel. And you have zero data to bolster your position.
Michael Orwin profile picture
I expect Nvidia's sales and high margins in AI and accelerated computing to attract competition and funding for it. I expect competitors to mostly offer some differentiation and target specialist areas, rather than try to challenge Nvidia head-on.

Nvidia has the advantage of CUDA, a powerful server-farm where they use their own tech for R&D, the Mellanox acquisition along with the vision that the whole datacenter is now the computing unit, and versatile hardware that's good for both AI training and inference. While I have some understanding of those, without a professional background in tech, I'm going to struggle to evaluate competitive threats.
"NVDA has VERY power hungry laptop GPUs."

Total nonsense.
Have you seen Navi based laptop ? Me not, nowhere to be seen.
On slides, it looks good. On reality, it falls short.
Even in 7nm, Navi can't compete against what Nvidia offer in the laptop space, even they're still on 12nm.
Your comment is just utter bullcrap and just show that your opinion has no value, whatsoever.
I think there may get be a small mix up somewhere in the gaming analysis: people loading up on laptops to work from home arn't usually Gamers putting out $ for preformance. In other words, even gamers who buy laptops for work, are also going to be buying proper rigs with full / higher spec'd GPUs. Turing *has been slow to be adopted, with the 1660 TI / 1660 Super by far and away seemingly the most popular.... Even with slow adoption though, NVDA never lost more than a couple % TAM of Gamers. My beleif is that People who get a 1660 level card wanted a stop gap. Most gamers still rock a Pascal gen card (myself included). Therefore, Ampere will trigger an enormous upgrade cycle. I hope they really max out the specs and annihilate the competition. Ampere leaks don't seem to disappoint on that front.
Michael Orwin profile picture
I think AI and accelerated computing now smells of money, thanks to Nvidia's datacenter sales topping a billion dollars in the latest quarter. While that could push the stock higher, some market and macro factors could pull the other way (e.g. the the Buffett Indicator). My guess is "smell of money" will win, but as a long term holder I don't need to be right.
I like Nvidia and the capabilities of this company and the stock is on my shortlist. But, sorry, the storytelling is getting ever more creative by the Wall Street herd in finding sound arguments to push the stock further into outer space. And this is not the only one being so strikingly disconnected from all reasonable fundamentals and earnings projections. The tech bubble from 2000 sends warm regards. With all due respect, but we have to see lower levels then your 15% downside.
09 Jun. 2020
I get your point, though I value NVIDIA as one of the most potent players in the fields of cloud computing/datacenters, gaming, AI and autonomous driving. Hence, my humble opinion is that no matter what price you pay for a share of the company today it's going to be worth it. If you bought at the top of the tech bubble, its a 22x gain on today's market price, even though the price movement was basically zero for 15 years. Nowadays I think its different in a sense that NVIDIA today is a disruptor and leader in key markets and not a small competitor. Also, technological advances/disruptions now take place within years, not decades. I think up to 2030 it really doesnt matter at what price you buy today, the probability you make above-market, risk-adjusted returns with NVIDIA is solid. This however is based on the idea that all forcasts about NVIDIA (and similar players) are wrong because no once can really forecast what the world will look like in 2030. I think that the TAM of future key markets NVIDIA can be/is the leader in will be much, much higher than anyone can imagine today. But, based on conservative assumptions, the valuation today is too high, I agree.
However, I dont think that a solid margin of safety on tech stocks like NVIDIA is that important because they have most of their value sitting in technology that is very hard to value correctly, especially when looking at the future value.
Thanks for this valuable comment, which I agree in almost all the arguments. Still, Nvidia is in the hardware business (which IMO is a shrinking element of the valuation, though) and for me its hard to imagine the pure speed and size of expansion such companies have to stem with respect to the forecast numbers, which circulate. Ok, this bottlenecking will just be a comforting element for earnings hikes, but not in the scale which justifies the current price levels. The intrinsic value of techs like NVIDIA is more and more a question of a good rating of its intangibles, and this is where it gets blurry. My feeling is that we are in the "big bet" territory today. Something to the past: back in 2000 NVIDIA was a smaller player and after the big tech burst the company was priced around its intrinsic value (even long periods after 2008). I wish I had it on the radar then ... but right, with hindsight to even the peaks of the 2000 its a no-brainer. The winning was 22x .. 30x. BTW, there are similar stories of (med) tech stocks like German Sartorius (SRT3.DE) which went up >150x from 2009.
An excellent and comprehensive analysis and conclusion.
09 Jun. 2020
wait for forward rev/eps to catch up and the stock will be even higher. are you considering negative real interest rates and the impact on your macro thesis?
fabskxha profile picture
Sometimes you just have to pay a little more for the best ideas. Quality stocks don't always fall to become a bargain. After the next ER, the higher prices you pay today will make your buys the bargains of tomorrow.
"Sometimes you just have to pay a little more for the best ideas."

It was $196 three months ago LOL.
grxbstrd profile picture
So overall a fair analysis.

The only problems I had with this article is the re-painting of history around DC biz, like

- "At the time, we didn't have much data on the performance differentiation between Nvidia's GPU accelerators and these new ASICs, except really for the TPU. All three generations of Google's (NASDAQ:GOOG) (NASDAQ:GOOGL) TPU processor were faster and more efficient than even Nvidia's top-of-the-line GPU offerings."

No one ever claimed that but Google. And it was an impossible claim to validate as there was no 3rd-party test bench to plug a TPU into. Only with the advent of MLPerf could Google's claims be validated, and now we know the TPU accels in only a handful of tests while Volta/Turing basically run everything, and for the most part run faster than TPU3. www.MLPerf.org/results. It appears Google has abandoned the TPU effort and is all in with Ampere.

Beyond TPU, competitive ASICs have proven to be a lot of smoke and little else. They still don't show up on mlperf despite years of development.

There were voices here on SA saying these things, and talking about the goodness of the CUDA ecosystem years ago. A few smart investors listened. And some others are still waiting around for ideal entry points that may or may not ever come.

Now with Ampere raising the bar, many of the home grown and chip startups in ASIC land will be dropping by the wayside, imo.
grxbstrd profile picture
re: "Google . . . is all in with Ampere."

"To continue to help you meet your goals, we’re excited to announce forthcoming support for the new NVIDIA Ampere architecture and the NVIDIA A100 Tensor Core GPU. Google Cloud and the new A100 GPUs will come with enhanced hardware and software capabilities to enable researchers and innovators to further advance today’s most important AI and HPC applications, from conversational AI and recommender systems, to weather simulation research on climate change. We’ll be making the A100 GPUs available via Google Compute Engine, Google Kubernetes Engine, and Cloud AI Platform, allowing customers to scale up and out with control, portability, and ease of use.

In addition, Google Cloud’s Deep Learning VM images and Deep Learning Containers will bring pre-built support for NVIDIA’s new generation of libraries to take advantage of A100 GPUs. The Google Cloud, NVIDIA, and TensorFlow teams are partnering to provide built-in support for this new software in all TensorFlow Enterprise versions, so TensorFlow users on Google Cloud can use the new hardware without changing any code or upgrading their TensorFlow versions." cloud.google.com/...
shaderhacker profile picture
The problem is even if the valuation was overbought. If it becomes oversold, people are going to buy it back up again. Its' a hot stock and there is no way anyone will be able predict what it's price *should* be. People are ready to buy this stock at any chance of it falling down consistently. Since most stocks went down to reasonable valuations in March, that opportunity is lost.
Antiochus profile picture
Here’s the thing - if you think the valuation is “ridiculous”, but the current price is only 20% above your price target...that’s not really ridiculous, is it? If the stock is gonna be 80% lower in 5 years, it would be overvalued either way. It it goes 10x in 5 years, it would be irrelevant if it was 20% higher now. 20% or less should never be your margin of safety when investing.
I think that is the correct framework to have if you're going to be a long , nobody can predict the Macro events like war or virus or market recession which cause the super dips; in the meantime its either hop on board or get left
puddnhead profile picture
"predicting" is the wrong mindset in the first place. one should be calculating rough odds, in order to make some kind of risk/reward judgment. so for simplified hypothetical example you think it's 60% likely something will go up, but the downside might 2x the upside, it's irrational to buy even though you "predict" it will probably go up.

Something Kwan-Chen Ma used to do here. though not starting from the same places I would -- among other things he would treat each stock like it's own little distinct universe of possibilities, while I'm personally more focused on outside (macro/market) factors.
puddnhead profile picture
p.s. I should clarify I was only responding to the reply, not Antiochus' post, which I think was spot on -- I also really scratched my head at points when I read this, for example the conclusion saying "my valuation sees ~15% downside in Nvidia ... I would wait until the name is well south of $300 before buying in." Say wha? I'm just too tired to push back at incoherencies here much anymore (props to grx and others for continuing to put up the good fight there).
Actually I bought at the low and this is my best investment ever. As for your thoughts I think you are wrong. See you $100 higher from here.
Yes we know you are long Goog.
You going to factor in Mellanox EPS numbers or nah?
grxbstrd profile picture
aren't they in the DC #s?
Quarter Nvidia just reported pretty sure they weren't in there on either non or GAAP side.
grxbstrd profile picture
agree, but I think the author accounted for it in his projections. 1.14-> 1.7, no?
Disagree with this article? Submit your own. To report a factual error in this article, . Your feedback matters to us!

About NVDA

SymbolLast Price% Chg
Market Cap
Yield (TTM)
Rev Growth (YoY)
Short Interest
Prev. Close
Compare to Peers

More on NVDA

Related Stocks

SymbolLast Price% Chg
To ensure this doesn’t happen in the future, please enable Javascript and cookies in your browser.
Is this happening to you frequently? Please report it on our feedback forum.
If you have an ad-blocker enabled you may be blocked from proceeding. Please disable your ad-blocker and refresh.