I thought of buying NVDA just under the $100 range, believing it would rocket on AI hype but then crash after it is over. I didn’t in the end as I was not confident in being able to time this. While AI is a good business, there are uncertainties as to whether NVDA will adapt to this market. I doubt GPU will be the mainstay of AI compute in the future and on any switch to dedicated compute for AI, NVDA would lose the synergy with its gaming market.
On the gaming market front, NVDA runs the risk of losing this business in the long term due to commoditization of gaming (except at the top end) with GPUs being brought onto the CPU package. 7nm NAVI could already be the start of this trend.