Some would argue $QCOM is undervalued. It's not that simple. $QCOM is just a chip designer and sends all their designs to $TSMC. They don't own the end customers, no vertical integration. Amazon had $AVGO design their TPUs, and guess what, AVGO/Broadcom sends their chip to TSMC. The only value added is designing chip. Now Amazon is doing a deal with QCOM, a company that fell behind, to design the next chip, pressuring them to undercut AVGO. Amazon serves end user on its AWS platform, one of the largest hyperscaler. They have the pricing power. Now they have 2 companies competing for the price. While $NVDA also design and sends their chip to $TSMC. They controls the CUDA stack developer ecosystem connecting to end developers. Chances are NVDA design will far outpace the price/performance of the rate of designs by QCOM/AVGO. The math is not good for $QCOM. Pure chip designer without the distribution to end developers, nor manufacturing, pressured by some of the largest company on earth. I could be wrong in the short term, but it's not a risk I'm willing to take.