Is Qualcomm actually capable of building a $15B+ data-center business, or are we underestimating Nvidia/Broadcom/hyperscaler competition?

Qualcomm is making a pretty aggressive bet on data centers.

Management is targeting **>$15B of Data Center revenue by FY2029**, with the business spanning server CPUs, AI accelerators, custom silicon, high-bandwidth compute and connectivity.

And this isn’t just a PowerPoint announcement anymore:
Meta has a multigeneration agreement around Qualcomm’s Dragonfly server CPU.

Microsoft is involved with Qualcomm’s High Bandwidth Compute architecture.

Amazon just announced a multigeneration collaboration with Qualcomm covering custom AI inference silicon and optical connectivity.

Qualcomm is targeting production of its C1000 server CPU in 2028.

But here’s what I genuinely don’t understand:
**Can Qualcomm actually scale this into a major data-center business against Nvidia, AMD, Broadcom/Marvell and the hyperscalers’ own silicon?**

Nvidia has the obvious advantage in AI accelerators and software/CUDA.

AMD already has server CPUs + GPUs.

Broadcom has enormous custom-ASIC and networking experience.

Amazon, Google, Microsoft and Meta are all developing their own silicon.

And Qualcomm itself previously attempted the server CPU market and eventually pulled back.

On the other hand, Qualcomm has some potentially interesting advantages:
Strong ARM CPU/IP expertise

Very strong power-efficiency/system-level design background

Custom silicon experience

Connectivity/SerDes/optical capabilities

Existing relationships with major hyperscalers
A strategy that isn’t simply “replace Nvidia GPUs” — it also targets inference, custom silicon and rack-level connectivity

So I’m not asking whether QCOM will “destroy Nvidia.”

I’m asking something more realistic:
**Can Qualcomm establish a durable $10–15B+ annual data-center business by FY2029?**
And what is the biggest thing that could prevent it?

**Nvidia’s software moat?**
**Hyperscalers designing internally?**
**AMD/Broadcom competition?**
**Qualcomm execution?**
**Power/performance not being good enough at scale?**

**Or is the AI infrastructure market simply large enough for another major player?**

I’d especially like to hear from people who understand data-center hardware rather than just QCOM/NVDA stock bulls.

**Bull case vs bear case — what am I missing?**