Besides the technical case for a pullback, with $NVDA's price currently trading near 40% above a 200 day moving average, hedge funds are piling on:

https://seekingalpha.com/article/4045214-nvidia-short-opport...

Personally, I love $NVDA. Great products, superlative engineering and management. I believe Masayoshi-san's thinking a long time horizon for this investment. Cloud GPUs are eating software. $NVDA's 100 Tflop Volta GPU could be a monster seller. This is a bet on a new revolution.

It's not just deep learning. PC gaming is also booming and Nvidia is the leader by far in that space. The Switch is running on a Tegra X1, so they're back in the console market at a very low cost to them. Toyota and Tesla will be utilizing their chipsets in autonomous cars, which is a pretty huge win, and Tesla already uses their chipsets to power in car entertainment/nav.

> The Switch is running on a Tegra X1, so they're back in the console market at a very low cost to them.

I think you seriously overestimate console market. Back in 2011 Sony reported 56 million PlayStation 3 units sold and Nvidia made $500M in royalties from it.

https://www.extremetech.com/gaming/150892-nvidia-gave-amd-ps...

Considering there is a lot more alternatives in SOC market now I seriously doubt Nvidia margins become any better. I don't mean that $500M is little amount of money, but it's simply can't justify market cap growth.

UPD: My bad, there actually some data on it and I can be wrong:

http://www.gamesindustry.biz/articles/2017-05-10-nintendo-sw...

Though I still don't think growth can be really related to console itself.

You're right. The gaming opportunity for Nvidia is groing to at most grow at a ~20% CAGR through 2022, and this is largely being driven by the international markets. The US gaming GPU market is only growing at low-single-digits for the foreseeable future. The autonomous driving and datacenter opportunity are a completely different story. The datacenter business is already growing at >100% y/y and this market has only just started to kick off about 6-9 months ago. How big it will be in 5 or 10 years is still up in the air, but it looks like it will be massive. The auto opportunity hasn't picked up yet but it's becoming increasingly clear GPUs will play an integral role in any level 4 or level 5 ADAS vehicle. Both of these markets were a completely neglible portion of the NVDA's story just two years ago.

Today, with the stock up more than 250% in the last 18mo., the implied market size (and market share) estimates in the stock are already pretty ambitious. In my opinion, they're not too ambitious, but the question now becomes, when will it become clear that they are conservative? What will it take? SoftBank has a long time horizon, they're comfortable waiting for when this happens, whether it's in 4 quarters if datacenter growth grows from even these remarkable levels or in 4 years when the autonomous driving market with be beginning to take shape...

The Switch is highly profitable for Nvidia. They could be looking at upwards of $100 million a year in profit on the Switch, and upwards of $400 million a year in revenue, almost doubling the size of their Tegra division. $100 million in profit and $400 million in revenue for a company that does just under $7 billion in revenue is 2016 is very significant.

https://www.fool.com/investing/2016/09/26/will-the-nintendo-...

I have no doubt it's significant amount of money, but still nowhere near to justify growth. And even if it's one of factors it's still very small one since both big consoles use AMD hardware and in next 3-4 years there not going to be another one.