Jump to content

NVDA - Nvidia


berkshire101

Recommended Posts

Has this stock popped up on anyone's radar?  What's surprising is the massive expansion in margins in the recent years, thus stock trading for much higher multiples.

 

What contributed to the expansion in margins?  Is it sustainable or typical trend we see in NVDA's business cycle?

Link to comment
Share on other sites

What contributed to the expansion in margins?  Is it sustainable or typical trend we see in NVDA's business cycle?

 

I'd say crypto mining and AI. Although AMD may have benefited more from GPU shortage for mining.

 

Is it sustainable? Who knows.

 

I'd think crypto mining is not sustainable. It will either move to specialized solutions (and in fact for bitcoin + ??? other cryptos it has already) or die off (don't tell that to proof of work crypto weenies).

 

AI as such is sustainable. The question though is whether NVDA will continue to have near-monopoly position in it and whether it won't move to specialized solutions too (although NVDA is trying to cover that somewhat). I'd say short term (couple years) there's not many alternatives. Longer time frame there's gonna be companies that try to get into this space.

 

Consumer/gaming GPUs are still bread and butter. This will take a hit in recession or if business cycle turns. Any business cycle peaking would now be masked by crypto mining though... From what I've heard - here on CoBF too - the pricing on gaming GPUs has been crazy. I haven't looked at the trends recently. Maybe resident experts will chime in.  8)

 

On the positive side, if - a big if - they can get their solutions into majority of upcoming (self-driving or not) cars, that would be a big market. If they cannot, the current market cap is discounting a lot.

 

Disclosure: I hold tiny position in NVDA.

Link to comment
Share on other sites

This will take a hit in recession or if business cycle turns.

 

It'll be interesting to see how long pent-up demand might sustain them, but also, what'll happen as crypto moves off: there are are hordes of PC gamers who've been putting off (or just unable) to upgrade their video cards in the last few years because the street price is 2x+ retail, when you could get them. Many people were selling their old cards for way more than they paid.

 

I don't have enough data to refute those 10% numbers, but the effect on the market's been sustained shortages with price inflation. If crypto buying lets off (as it must -- specialized chips are so much better), you'll see both new cards coming on to the market at the traditional price points but also the super-cards of the last few years dumped onto the market in huge quantities.

Link to comment
Share on other sites

Whether it's for crypto / AI or not, if you want to speed up your number crunching (i.e., matrix computation), GPU is the way to go these days. At some point, some better alternative technology will probably come along (perhaps FPGA) but who knows when / what.

 

As for NVDA vs. AMD regarding the use of GPU for high-performance computing task, NVDA has the edge because of CUDA. It's not the race to cheaper / better hardware as it used to be with the typical semiconductor business. CUDA really made programming with (NVDA) GPU a lot easier, which I believe is the significant moat for NVDA.

 

When I realized this I bought NVDA shares back in 2015... but then sold them in early 2017 when the price started getting astronomical... I guess I  sold way too early  :'(

Link to comment
Share on other sites

The FPGA path is not yet there in terms of productivity, and has a long way to go. One can however port SW running on Nvidia cards onto one (I've experience only with Altera's OpenCL), and get the same speed, at a ridicules power consumption compared to GPUs.

That being said, my current employer relies heavily on Nvidia platforms, as most of the companies I know, and it seems that it is not going to change soon.

To my mind, CUDA would eventually be replaced by OpenCL (Supported by AMD, Intel, Xilinx and even "supported" by Nvidia), just as Glide is now dead and OpenGL is used for graphics.

 

Disclosure: I'm a former Intel employee, but don't hold a single share.

Link to comment
Share on other sites

  • 3 weeks later...
  • 1 year later...
  • 2 months later...

https://www.telegraph.co.uk/technology/2020/08/14/arm-sale-talks-accelerate-amid-calls-protect-cambridge-hq/

 

Talks between UK chip designer Arm and US firm Nvidia are accelerating with a $40bn (£30bn) deal expected to be thrashed out by the end of the month.

 

The exclusivity period is expected to last 30 to 45 days, after which the companies would walk away or the talks would open to other parties.

Link to comment
Share on other sites

I usually don't sell, but when I do, I really f&ck things up.

I held this baby from $100-120 range, sold out in the middle of mayhem in March, to raise even more money for re-deployment, and it literally went up every week since I sold it.

 

An giant opportunity cost of more than 100% and counting.

Link to comment
Share on other sites

  • 4 weeks later...

This is a watershed moment for NVDA. No opinion on the current valuation but if this transaction gets past the regulators and the ecosystem opposition, it would pave the way for a trillion $ NVDA market cap in a few years. He has the best GPUs (AMD is well behind esp on software), the best network interface cards (intel really lags) and now with access to ARM, he secures the last piece - CPUs, plus broadens his market to mobile where NVDA had practically no presence. Jensen is criminally underrated among the big CEOs for someone who has run his company as CEO since the early 90s.

Link to comment
Share on other sites

On one hand, ARM has been a great company pre Softbank acquisition and they pretty much own non-x86 CPU space.

 

On the other hand, they've been design and license shop, which made ARM architecture so attractive and adopted. If NVDA decides to manufacture ARM CPUs, ARM licensees won't be happy. If NVDA continues to design and license, they are somewhat limited in revenues, features, products, etc. (though margins/profits from licensing are very high).

 

It's gonna be quite interesting integration and growth challenge IMO.

 

Disclosure: I hold NVDA stock. No plans to buy or sell here.

Link to comment
Share on other sites

On one hand, ARM has been a great company pre Softbank acquisition and they pretty much own non-x86 CPU space.

 

On the other hand, they've been design and license shop, which made ARM architecture so attractive and adopted. If NVDA decides to manufacture ARM CPUs, ARM licensees won't be happy. If NVDA continues to design and license, they are somewhat limited in revenues, features, products, etc. (though margins/profits from licensing are very high).

 

It's gonna be quite interesting integration and growth challenge IMO.

 

Disclosure: I hold NVDA stock. No plans to buy or sell here.

 

Check out the call this morning, Jensen's pretty clear about their plans there, both to keep the open license model, and how they will integrate the ARM stack to compete in the data-center.

 

If you want the cliffs notes, i'll publish my notes via email on the call on Wed morning.

Link to comment
Share on other sites

  • 3 weeks later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...