You all convinced me. Okay, I broke down and bought a tiny starter position (0.75%). I can’t guarantee I’ll keep it, but on the other hand it may grow. We’ll have to see.
Saul
26 Likes
A significant purchase made with optimism and enthusiasm of a high order!
8 Likes
Now if you will get on board with TTD the assimilation will be complete!
(Evil Dreamer with a glint in his eye)
11 Likes
I did the same thing a while ago, despite the “Motorola Rule,” which was an expensive lesson that has served me well for many years.
1 Like
Hey! You’re stealing all my moves!
Saul,
I’d love to hear your rationale for taking this position.
NVDA is a company I keep waffling on, and have been for years. They have had a decent run up in the past two years or so. But I just can’t get past the fact that hardware companies as a category, grow much more slowly than software companies. And all it takes is one bad chip design to send them spiraling down for several quarters to undo all the growth they’ve accumulated previously.
I do really like NVDA as a company, their CEO and other executive leadership, and their product. And maybe the combination of cryptocurrencies, AI, and eSports will be the thing that drives them forward for years to come.
But it’s still hardware, and hardware still, overall, tends to grow more slowly, and have more physical limitations/risks associated with it than software.
Perhaps I’m too hung up on the hardware aspect of it.
Anyway, I would enjoy hearing your thoughts.
Paul
1 Like
Paul, you’re ignoring the CUDA aspect to NVDA. “There is no serious AI development on AMD” is something I hear all the time, too. NVDA owns the space, from both HW and SW. From a gaming perspective the two vendors are interchangeable (thanks to OpenGL, etc.). This is not true in the ML/AI space, where the software of choice runs on NVDA and not on AMD. NVDA will almost certainly not be as fast a grower as other names mentioned here. But it will trounce the market over the next few years, and likely do so with lower risk.
3 Likes
Will certainly not try to speak for Saul on the “why” part, but I thought I’d share my own brief thoughts on why I continue to hold/add NVDA (though I did trim some after the 2014-2016 run-up, I have since doubled the position size and it grew even beyond that to ~12% at one point). In mostly significance order, with most important (to me) first:
-
Company leadership. Jensen Huang is a rare breed, IMO, able to convey intelligently what the vision is, and certainly able to invigorate his staff and employees to deliver it. He seems genuine, has been with the company for a long, long time, and clearly has a vested, personal interest in the success of Nvidia. All of which well aligns with what I want in a CEO as an investor.
-
Product moat. Their only real competitor in the GPU market is AMD, who until very recently weren’t even out of the locker room and on the field. AMD may be back into leapfrog mode where AMD and NVDA may trade places now and then, but I think NVDA has actually been holding back because they didn’t NEED to push out their full-awesome quite as soon. In Datacenter, AI and others, there’s no comparison. CUDA platform wins without even starting the clock, and NVDA has the lion’s share of partnerships on the Datacenter piece. AMD sort of wins the now-worth-nothing crypto piece because their cards couldn’t be sold as expensively as NVDA (i.e. you could mine more cheaply), but that point now has no significant forward value (IMO)
-
Market share^H^H^H^H^H^H … no, let’s better label this mind share. This is sort of similar to the product moat, but different enough I think it’s worth mentioning. When developers are thinking about what GPU do they need their upcoming game to really shine on, very few A-list shops seem to lean toward AMD. In the AI realm, CUDA already has the mind share locked down pretty tightly. Not suggesting it’s impenetrable, but they’ve been doing it for a long time and the name is locked in peoples’ heads. I think this also translated somewhat into the /momentum/ part of the stock movement into 2016/2017 (which is now in breather mode, IMO, but likely to jump again). There was SO much press about Nvidia, from stock analysts to crypto analysts to my neighbor’s grandma asking about Bitcoin, it was impossible to not hear it. Not so much for AMD, or if you heard AMD, you usually heard it in comparison to NVDA.
There’s obviously much more, such as the financials (company debt, dividends, and whatever other metrics you want to measure), and I didn’t even really put AI or self-driving cars very high on the list (to me those are somewhat speculative in that they are likely to take 2x longer to actually pan out into something significant), but those are three that I think keep me very interested in the long haul, aside from the dollars and ratios.
6 Likes
Saul,
I’d love to hear your rationale for taking this position. NVDA is a company I keep waffling on, and have been for years. They have had a decent run up in the past two years or so. But I just can’t get past the fact that hardware companies as a category, grow much more slowly than software companies. And all it takes is one bad chip design to send them spiraling down for several quarters to undo all the growth they’ve accumulated previously.
I do really like NVDA as a company, their CEO and other executive leadership, and their product. And maybe the combination of cryptocurrencies, AI, and eSports will be the thing that drives them forward for years to come. But it’s still hardware, and hardware still, overall, tends to grow more slowly, and have more physical limitations/risks associated with it than software. Perhaps I’m too hung up on the hardware aspect of it. Anyway, I would enjoy hearing your thoughts. Paul
Hi Paul, You just said everything I’ve been thinking (on both sides of the argument). I wouldn’t get too carried away by my 0.75% position. It may not last.
Saul
3 Likes
hlygrail,
Thanks for that response!
I agree 1000% that Jensen Huang is a rare breed in the corner office! If there were one reason I’d invest in NVDA it’s Jensen and his leadership ability. I’ve been watching him for years, so I have no doubt about his ability to lead the company and make great things happen.
I guess, generally speaking, I don’t disagree with a single thing you said. None of it. In fact, all of it is what has me generally interested in NVDA.
My hesitation is summed up entirely by the fact that they make a physical widget. And most, if not all of what you outlined, held true at one time for IBM, Sun, Cisco, etc. And companies that build physical widgets have more overhead and more risk generally speaking, than software companies.
I’m fairly new here and spend far more time reading than writing on this board. And the thing that attracted me to this board first and foremost was that there seemed to be a bunch of people who, like me, had noticed over time that the software/cloud companies grow incredibly quickly and carry lower risk of customer alienation than hardware companies do. And NVDA flies in the face of that whole premise.
I don’t want to say this board is limited to investing only in software/SaaS/cloud companies, but we do seem to spend a lot of time discussing high growth, incredibly agile, software companies vs. hardware/physical widget makers. The one exception beyond NVDA seems to be ANET, but even they’re moving away from the physical widgets and the real magic sauce is all in their software that can run on anyone’s widgets.
So, while I 100% agree with your statements above, the simple fact that they’re making physical widgets to me implies that their overhead and risk profile is higher, and as a result, their margins and growth will be lower. And, for my limited funds, I don’t think putting my money there is as good as other opportunities where the growth and margins are higher due to lower overhead and risk.
–
Paul - who will likely continue to waffle on NVDA for a while longer…
4 Likes
Hi Paul, You just said everything I’ve been thinking (on both sides of the argument). I wouldn’t get too carried away by my 0.75% position. It may not last.
Hi Saul,
Thanks. So I should take your .75% position merely as you’re watching to see how it does and will make up your mind at a later date…
Makes sense to me. I will, for now, continue watching from afar until something convinces me that NVDA is the best place for my limited dollars 
–
Paul
My hesitation is summed up entirely by the fact that they make a physical widget.
If we were talking about Intel or Micron I would agree entirely but Nvidia is much more of an IP business than a chip maker, more like ARM Holdings.
NVIDIA Corp.'s Relationship With Taiwan Semiconductor Manufacturing Is Deepening
There are clear signs the collaboration between the graphics chip designer and the contract chip manufacturer is getting stronger.
Ashraf Eassa (TMFChipFool)
May 17, 2017 at 8:30AM
For many years, NVIDIA (NASDAQ:NVDA) has relied on Taiwan Semiconductor Manufacturing Company (NYSE:TSM) to manufacture the graphics chips it designs. Today, most of NVIDIA’s Pascal architecture-based graphics processors – from the GeForce GTX 1060, designed for mainstream PC gamers, all the way through its cutting-edge Tesla P100 datacenter accelerators – are manufactured by TSMC.
https://www.fool.com/investing/2017/05/17/nvidia-corp-relati…
Denny Schlesinger
4 Likes
I’m not Saul and most certainly I do not speak for him, but it’s well to keep in mind that NVDA is much more a platform company than a hardware company. Approximately 50% of the engineering staff are software engineers. CUDA (the s/w companion to heir h/w) is one of their moats (plural).
The company may not grow as quickly as some smaller, niche s/w players. That’s most likely true. Though Saul has time and again said he does not know what the market will do in the future and makes no attempt to “time” his buy/sell activity based on technical indicators or any other supposed predictors of the future (other than public information about specific companies or released by those companies), he has nonetheless demonstrated an uncanny ability for exiting positions in favor of other opportunities.
Since I’ve been one of his students (which has been close to the time he started this board) it seems to me that most of his sell decisions have been at a time when the company was still growing and the stock price still rising. I’ve not kept records, and in that Saul has asserted that once he sells he does not continue to follow the company I assume he too does not have such records, it would take a fair amount of research to determine if my assertion is correct, but I’m quite certain it is.
I’m long NVDA for a few reasons. Most of them have to do with the company itself, their product offerings, their management, their ability to identify and develop new markets for their products and their unassailable position in the market. While h/w may be at the core of their business one should remember that all that s/w needs a platform, it does not run on other s/w. Ultimately, h/w is the foundation of the internet, the cloud, the digital enterprise and all the ensuing s/w enablers. NVDA holds a unique position in this environment. All their competition (basically one also ran company, AMD) is literally years behind them, and one of the primary reasons they can’t really keep up and are unlikely to catch up is CUDA.
So, my guess is that Saul will not hold his position very long. He will soon see that his 0.75% position is not performing as well as other positions he holds and he will sell it. But for folks like myself, who are not nearly as nimble and not nearly as experience and not nearly as prescient as Saul, NVDA is a solid investment. With NVDA I don’t feel like I have to read between the lines of the ERs in order to understand what’s really going on.
I might also add that other experienced and successful investors who post here (Tinker, Denny, Dreamer, et al) are also long NVDA (I think, they do not report their holdings as Saul does, but they’ve written favorably about the company). As an apprentice approaching journeyman status, I have learned a great deal from all these masters. Saul is more like a guru.
14 Likes
My hesitation is summed up entirely by the fact that they make a physical widget.
Not an insignificant correction. NVDA designs and sells a widget. They have not incurred the overhead and sunk costs of actually making it. They’ve offloaded all of that to a fab shop.
Jensen is a genius.
6 Likes
You might also consider the fact that Saul speaks highly of and puts his money where his mouth is with respect to Square. Square is at least partially a h/w company. Most of their s/w (with the exception Cash) is dependent on their hardware offerings. Is that so very different?
2 Likes
My hesitation is summed up entirely by the fact that they make a physical widget.
I think there is a mistake in making a simple distinction between hardware and software like this.
When a company commits to an ERP system, often with substantial customization, this can be a very sticky decision. Yes, there are some companies which sell ERP packages and “throw them over the transom” so that the company pays for all the modifications. With that business model the selling company gets the original sales price and maybe some maintenance for a while until the company decides that it has made so many mods that it can never upgrade to a new version. That isn’t very different than selling hardware.
On the other side, selling someone a server is a purchase which is unlikely to be replaced for a while and there is no guarantee that the replacement will resemble the original. But, with Nvidia and their graphics cards and, I suppose GPUs generally, the buyers are people who demand to be on the leading edge of capability and it is only a couple of years before the leading edge is in a significantly different place, so there is a lot of repeat business replacing what one sold a couple of years before. Not SaaS, but not like servers or appliances either.
2 Likes
Well, for Square, the hardware piece is just enabling technology … it is not what is actually being sold, which is the service.
4 Likes
With Square, hardware updates don’t figure large for customers, while they often do with Nvidia. Square makes most of its money by selling their services. Nvidia makes their money by selling hardware. Does CUDA (and whatever else Nvidia software used to program their hardware) get included in the hardware price, or is it paid for separately? It’s also possible that it’s free, but there are saleable services connected with it. (Red Hat does this with Linux - the vanilla Red Hat version of Linux is free, but you can buy various add-ons [and custom software?] from Red Hat.)
1 Like
Here is an as of today ‘deep dive’ into Nvidia?
My apologies if it’s ‘old news’.
https://wccftech.com/review/nvidia-geforce-rtx-2080-ti-and-r…
I am not an IT person, therefore lots of jargon passes over my head.
I skimmed some of the pages, and really only understood that the Nvidia Turing GeForce etc, etc, is AMAZING!!!
At least these guys think so.
Eventually, I got to the Conclusion page, and this part stood out to me:
Ray tracing is coming, but we’re still waiting for Windows to update to make DXR available, and then the wait on games to implement it. Until DXR and the game are ready there’s quite a bit of unused real estate on these dies.
Thinking about the sideways stock price movement over the last 6 months, does the above explain part of it?
Nvidia is way ahead of not only AMD, Intel, etc, but the whole infrastructure system as well. i.e. there’s not much out there at this time that can actually use the Turing’s capabilities.
This has already been said many times on this board and NPI, about Nvidia.
ralph
6 Likes
You all convinced me. Okay, I broke down and bought a tiny starter position (0.75%). I can’t guarantee I’ll keep it, but on the other hand it may grow. We’ll have to see.
I’m sorry, but I just couldn’t hold it. When everything was down there were so many other places I’d rather put my money, so I sold my little position in Nvidia yesterday morning (Wednesday).
Best,
Saul
13 Likes