I also 😭 CPU > 700%

I too am experiencing a similar issue after upgrading to 15.2.2. After running npm run dev NPM calls the script but nothing happens. Tried upgrading to the Canary version, reinstalling packages, etc.

I am also experiencing the same issue after upgrading to v15.2.2, v15.2.3 or v15.2.4.
I was working fine at version v15.1.6.

I'm getting a variant of this of 15.2.3. It's absolutely insane. Everything will run fine for a little bit, then, randomly, it wall just hang and peg all of my cpu cores indefinitely. Restarted machine, cleared cache, etc. Nothing helps.

For us the issue (stuck in "compiling") is fixed after upgrading to v15.3.0

We've also confirmed it works with version 15.3.0 after upgrading from 15.2.0.

Is there any word on what exactly was causing this?

I also have the same issue just right now. I use "next": "15.2.4" and it just fine/work normally for a few minutes before I restart "npm run dev" command.

I faced same issue on latest version Next 15.4.6, after spending 3-4 days. I managed to resolve this downgraded next version to Next v15.3.0 and it works.

I got same issue on latest version too. After the first compile is done, next page is not compiled.
Any answers for it?

im having the same issue

For me it's caused by .next/postcss.js taking an endless amount of memory (+20GB)

next build --turbopack

▲ Next.js 15.5.3 (Turbopack)

• Experiments (use with caution):
  ✓ useCache
  ✓ inlineCss
  ✓ clientSegmentCache

Creating an optimized production build ...

thread 'tokio-runtime-worker' panicked at turbopack\crates\turbopack-ecmascript\src\lib.rs:2429:13:
The high bits of the position 1365071 are not all 0s or 1s. modules_header_width=11, module=20
note: run with RUST_BACKTRACE=1 environment variable to display a backtrace
✓ Finished writing to disk in 31ms

Build error occurred
[Error: Turbopack build failed with 1 errors:
./app/page.tsx
Code generation for chunk item errored
An error occurred while generating the chunk item [project]/app/page.tsx [app-rsc] (ecmascript)

Caused by:

• The high bits of the position 1365071 are not all 0s or 1s. modules_header_width=11, module=20

Debug info:

• An error occurred while generating the chunk item [project]/app/page.tsx [app-rsc] (ecmascript)
• Execution of *EcmascriptChunkItemContent::module_factory failed
• Execution of ::content_with_async_module_info failed
• Execution of *EcmascriptChunkItemContent::new failed
• Execution of EcmascriptModuleContent::new_merged failed
• The high bits of the position 1365071 are not all 0s or 1s. modules_header_width=11, module=20

]
 ELIFECYCLE  Command failed with exit code 1.

Got the same issue with 16.1.1 so this is quite alarming. CPU, page memory and RAM at 100% bringing the machine to a halt. It seems to be caused by tailwind.css compilation, it spawns an infinite amount of nodejs processes

pnpx next info output:

Operating System:
  Platform: darwin
  Arch: arm64
  Version: Darwin Kernel Version 25.2.0: Tue Nov 18 21:08:48 PST 2025; root:xnu-12377.61.12~1/RELEASE_ARM64_T8132
  Available memory (MB): 16384
  Available CPU cores: 10
Binaries:
  Node: 24.11.1
  npm: 11.6.2
  Yarn: N/A
  pnpm: 10.24.0
Relevant Packages:
  next: 16.1.1 // Latest available version is detected (16.1.1).
  eslint-config-next: N/A
  react: 19.2.3
  react-dom: 19.2.3
  typescript: N/A
Next.js Config:
  output: N/A

I am using pnpm to run my NextJS (app router) project which uses TailwindCSS, TypeScript, and a few Shadcn UI components. My layout.tsx file just has a font loaded from next/font/google and my page.tsx file is extremely simple:

import {
Card,
CardContent,
CardHeader,
CardTitle,
CardDescription,
CardAction,
} from "@/components/ui/card";
import { Minimize2, Maximize2 } from "lucide-react";

export default function Page() {
return (
  <div className="w-full h-screen flex flex-col items-center justify-center">
    <div className="flex gap-16">
      <Card className="w-100 rounded-none">
        <CardHeader>
          <CardTitle>9th Grade</CardTitle>
          <CardDescription>10 credits</CardDescription>
          <CardAction>
            <Minimize2 />
          </CardAction>
        </CardHeader>
        <CardContent className="flex flex-col gap-2">
          <div className="border p-4 w-full text-center">Course</div>
          <div className="border p-4 w-full text-center">Course</div>
          <div className="border p-4 w-full text-center">Course</div>
          <div className="border p-4 w-full text-center">Course</div>
          <div className="border p-4 w-full text-center">Course</div>
          <div className="border p-4 w-full text-center">Course</div>
        </CardContent>
      </Card>
    </div>
  </div>
);
}

Literally no other files. The CPU usage seems high when I run pnpm dev:

Also memory usage currently:

Previously, the memory usage hovered around 3.33 GB.

Not sure if this is relevant but my NextJS developer tools show it stuck at compiling:

However the console seems fine and just shows the regular "GET" and success messages.

I'm also using the Arc browser, in case that helps.

Same here. The project was running smooth with Next 14. But as soon as i upgrade to 15 or 16 with React 19 it get stuck at "Creating an optimized production build..."

So frustrating, been trying different node versions and debugging with ChatGPT. Nothing works, no logs of value etc..

I'm on 16.1.6 and it's happening every time I do a production build. Machine becomes unresponsive; couldn't even ssh into it to check.

I was seeing 500%+ CPU, 10GB+ memory growth, infinite "Compiling…" in the client (mobile safari), requests taking 60+ seconds or infinite. Next.js 16.1.6 + Turbopack. Seems like Next had inferred the wrong workspace root.

On startup I noticed this warning:

⚠ Warning: Next.js inferred your workspace root, but it may not be correct. We detected multiple lockfiles and selected the directory of <some parent>/package-lock.json...

I had a stray package.json + package-lock.json + node_modules/ in a parent directory (left over from an unrelated experiment years ago). Claude thinks Next 16's stricter root inference picked that as the workspace root, so Turbopack tried to watch and resolve modules across the entire parent tree (33GB of unrelated projects in my case). Symptoms: hundreds of Can't resolve 'X' in <parent> errors per startup, runaway CPU/RAM, and dev requests timing out.

Two-part fix that worked for me:

1. Pinned the root explicitly in next.config.js:

   const nextConfig = {
     turbopack: { root: __dirname },
     // ...
   }

2. Deleted (renamed to .bak) the orphan package.json / package-lock.json / node_modules in the ancestor directory — changing turbopack.root alone (fix Added css using aphrodite #1) silenced the warning but module resolution was still walking up to the bad parent until those were gone.

After both changes cold compile dropped from 60s+ timeouts to ~5s, warm requests to ~40ms, idle CPU back to ~0.1%, and memory back to normal.

I've encountered the same error in the Next.js 16.2.6

I've got 100% usage on my mac

the fix for that helped me is

rm -rf node_modules
rm -rf .next