Qualcomm Q2 2026 Earnings Analysis
カートのアイテムが多すぎます
ご購入は五十タイトルがカートに入っている場合のみです。
カートに追加できませんでした。
しばらく経ってから再度お試しください。
ウィッシュリストに追加できませんでした。
しばらく経ってから再度お試しください。
ほしい物リストの削除に失敗しました。
しばらく経ってから再度お試しください。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
-
ナレーター:
-
著者:
概要
**ALEX**: Welcome to Beta Finch, your AI-powered earnings breakdown. I'm Alex, and joining me as always is Jordan. Today we're diving into Qualcomm's second quarter 2026 results, and wow, there's a lot to unpack here.
**JORDAN**: Absolutely, Alex. But before we jump in, let me remind our listeners: This podcast is AI-generated content for educational and entertainment purposes only. Nothing we discuss should be considered investment advice. Always do your own research and consult a qualified financial advisor before making any investment decisions.
**ALEX**: Thanks, Jordan. Now, let's talk numbers. Qualcomm delivered $10.6 billion in revenue with non-GAAP earnings per share of $2.65, hitting the high end of their guidance. But the real story here isn't just the headline numbers—it's this massive pivot toward what CEO Cristiano Amon calls "agentic AI" and their diversification strategy.
**JORDAN**: Right, and let's break down those business segments because they tell an interesting story. QCT, their chip business, brought in $9.1 billion, while licensing pulled in $1.4 billion. But here's what caught my attention—automotive hit another record at $1.3 billion, up 38% year-over-year. They're now at a $5 billion annualized run rate and expect to exit fiscal 2026 above $6 billion.
**ALEX**: That automotive growth is impressive, but I want to talk about this elephant in the room—the China handset situation. They're dealing with what they call "memory industry dynamics" that are causing handset OEMs, particularly in China, to be super cautious with their build plans.
**JORDAN**: Exactly. CFO Akash Palkhiwala was pretty candid about this. He said their China Android shipments are "meaningfully below the scale of end consumer handset demand" because OEMs are drawing down channel inventory due to memory supply issues and price increases. But here's the key—they believe Q3 will be the bottom, with sequential growth expected after that.
**ALEX**: So basically, people are still buying phones, but manufacturers aren't ordering as many chips because they're worried about memory costs. It's like a supply chain traffic jam. But what really fascinated me was Amon's vision for where AI is heading. He's talking about this shift from basic AI inference to what he calls "agentic AI"—AI that can orchestrate multi-step tasks and run continuously in the background.
**JORDAN**: And this is where Qualcomm thinks they have a competitive advantage. Amon argued that agent orchestration is predominantly CPU-bound, and he claims Qualcomm has "the world's best performing CPU across smartphones, PCs, auto, and soon the data center." That's a bold claim, but they're backing it up with some interesting product launches.
**ALEX**: Speaking of bold claims, let's talk about their data center ambitions. This was probably the biggest surprise in the call. They announced they're starting shipments to a "leading hyperscaler" in December for a custom silicon engagement. When pressed for details, Amon was pretty tight-lipped but called it a "multi-generation engagement."
**JORDAN**: The timing on that is interesting because it suggests they've been working on this longer than many people realized. Remember, they acquired AlphaWave earlier, which gives them custom ASIC capabilities. But Amon mentioned they've been talking to data center customers for several quarters even before that acquisition.
**ALEX**: One analyst asked a great question about the competitive landscape, especially with ARM now trying to vertically integrate and NVIDIA focusing on inference. Amon's response was fascinating—he basically laid out how the AI market is evolving from training-focused to inference-focused to now this new phase of "agentic" experiences.
**JORDAN**: Right, and his argument is that as AI becomes more about generating demand for tokens rather than just generating the tokens themselves, you need different types of
This episode includes AI-generated content.
まだレビューはありません