NVDA Q2 results

I wrote earlier today that I had recently increased my position in NVDA from about 5% to 10%. In the after hours NVDA is down 11% to about $154 on heavy volume for after hours (more than 7 million shares traded in after hours).

I looked at the results and listened in on the conference call. I must say that there’s nothing in the results that I didn’t like. I think guidance if well below what I was expecting but based on what I heard on the call, I think they will easily beat the guidance.

up 56% y/y
up 53% full year growth

up 91% y/y
up 99% full year growth

TTM P/E is now 39 (down from about 50 before earnings)
1YRPEG = 0.39

I have said before that the huge growth is in Datacenter and Automotive so let’s look at how those did:

Datacenter Revenue
revenue of $416M
revenue up 175% y/y
revenue up 2% sequentially

I think a big reason for the after hours reaction is the datacenter sequential revenue growth. They launched a new product for this market in the quarter. The product is the V100 GPU and nothing in computing comes close to it. It is a 10x improvement over their last product for the datacenter market. Customers knew it was coming and sales in Q2 were affected because customers did not want to buy the last generation product. This is a highest end product on the market and the DGX-1 which has 8 V100s in it is priced at $149,000 or $18,625 per GPU. The product first came available in Q2 so they had sales of it in Q2 but I expect them to sell a lot more in Q3. CEO thinks all datacenter should be using at least some of NVDA’s datacenter GPU. This is the AI enabling product and I think it will be largely implemented by the world in the cloud; this would mean that companies like Amazon and Microsoft will be buying these GPUs for their datacenter and end user customers would largely use in the cloud versus on-premises. I wrote earlier today that I think that the market potential of AI is hard to predict. CEO Huang echoed this when asked about the longterm of the datacenter market potential. Basically, the potential is going to be huge but the rate of adoption is less predictable.

Automotive Revenue
revenue of $142M
revenue up 19% y/y
revenue up 1% sequentially

I think this was another disappointment for the market. Maybe people think that this segment should be adopting faster. Now, we must remember that the Automotive segment revenue includes both revenue for infotainment in vehicles as well as revenue for self driving automotive. The big potential is for self driving but this will be small until cars with self driving becomes available. Huang was asked to break out the revenue for the 2 automotive applications but we would not do that. Huang did say that fully autonomous cars (taxis in some cities) would hit the market in 2020 or 2021. Tesla uses NVDA in all their cars and the Model 3 will be ramping up production in the second half of this year. The Audi A8 2019 model will have NVDA for their level 3 self driving features. The Toyota partnership will be quite large in a couple of years. There’s a partnership with Daimler which may start bearing fruit next year. And the partnership with Baidu for self driving cars in China may be quite large. Huang said we can expect more automotive partnerships in the coming year. Automotive for NVDA is still in moderate growth but I expect it to really take off in 4-8 quarters.

The Intel licensing deal ended in March so the OEM & IP revenue segment should have really dropped off in Q2. But it actually increased sequentially to $251M from $156M. This was due to cryptocurrency which they said was about $150M of the $251M. They also said that some of their gaming GPUs were purchased for cryptocurrency mining. NVDA has made a special product for this market. Huang thinks cryptocurrency mining is here to stay and they will watch the market and make sure that they can take advantage of any demand. This is in contrast to what the AMD CEO said; she thinks it’s a more transitory demand.

One analyst asked about the low guidance and asked whether NVDA would have enough supply to meet demand should the guidance be too low. Huang assured the analyst that NVDA would have no problem meeting demand. I think this means that we could see significant upside to the guidance.

A couple of other things that stood out to me:

  1. Huang said that NVDA has never been better positioned as a company. I think he meant by this that NVDA has the products that the market wants for each segment and the NVDA is farther ahead of the competition in all market segments: gaming, datacenter, automotive, professional visualization.

  2. The V100 was successfully ramped up in Q2. This means that production costs (and margins) were higher (lower) than we will see improvement to margins now that the product has been successfully scaled in production.

Overall, the results make me confident that NVDA is executing beautifully and that the future looks as bright or brighter than I previously thought. Tech stocks are correcting and NVDA is down 11% in after hours trading (although it’s still up 25-30% since the end of the last quarter). As I mentioned, I have a 10+% position and I will probably add to it in the coming days or weeks.


Huang did say that fully autonomous cars (taxis in some cities) would hit the market in 2020 or 2021.

I thought I heard him say the first self driving taxis would be on the road in 2018. With significant adoption in 2021.

This will be disruptive.


I thought I heard him say the first self driving taxis would be on the road in 2018. With significant adoption in 2021.

You’re probably right. I was going from memory and the written transcript isn’t out yet.

One thing is clear. Huang thinks self driving cars are coming fast. My big question about the timing is not the technology that will make it possible but the government regulation regarding letting those vehicles on the roads.

I am not so worried about the regulatory environment. It will be like Tesla fighting with the dealership laws. Some states will be reluctant and go slow, and some will be open about it and go fast.

There are multiple states, including California, that are competing to be the autonomous driving state. When such cars come closer to larger distribution I am sure that political resistance may increase, but as things stand now there will probably be a few well populated states that will welcome the autonomous vehicles. Plenty for an early roll out.

I don’t know how this will play out in Europe or Asia. I suspect Asia will welcome autonomous cars and that Europe will be more reluctant, at least at first.


The automotive market at this point is 95% talk and 5% implementation . It’s very early maybe even s Tech Enthusiast stage…
Left on it’s own Big Auto would adopt autonomous driving over perhaps a 15 or 20 year period. Tesla will probably prod them along , but even if Tesla sells a half million cars a year, that still does not have much impact on a 100 million per year light vehicle market. But I believe it will come, is inevitable, and NVDA will profit. And that the early Tesla implementation will cause ICE makers to proceed at a faster pace even if it is mostly out of wounded pride.

Analysts don’t look much more than a quarter two ahead. If they were to pick a stock that went down for 4 years then made a 5 bagger the next year, they would have been fired long before year 5 .
Money managers also are judged quarter to quarter, they have giant holdings and need the cover of a decent earnings report to unload all those shares.

No stock goes straight up forever, and the whole tech sector has been looking weaker over the last couple months. Why I bought those QQQ puts a few days back.
I worry more about the whole sector than I do about NVDA

Looking 4 or 5 years ahead Nvidia seems in good shape. Nothing about the report discourages me.
I may buy more.