<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis" -->

---
title: My monitor was cutting my download speed by 75% | daily.dev
description: A USB-C Ethernet adapter plugged into a Studio Display XDR was quietly limiting download speed to 600 Mbps on a 2.5 Gbps symmetrical line. The cause traced to...
canonical: https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: My monitor was cutting my download speed by 75% | daily.dev
og:description: A USB-C Ethernet adapter plugged into a Studio Display XDR was quietly limiting download speed to 600 Mbps on a 2.5 Gbps symmetrical line. The cause traced to...
og:url: https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis
og:image: https://api.daily.dev/og/posts/yIgZ5imIs.png
og:image:alt: My monitor was cutting my download speed by 75%
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# My monitor was cutting my download speed by 75%

**[Hacker News](https://daily.dev/sources/hn)** · 9 min read · 4 upvotes · 0 comments

## Summary

A USB-C Ethernet adapter plugged into a Studio Display XDR was quietly limiting download speed to 600 Mbps on a 2.5 Gbps symmetrical line. The cause traced to the Realtek RTL8156B chip's tiny USB burst size of 4 packets, which capped the drain rate to the Mac below the 2.5 Gbps coming in, overflowing its buffer and dropping frames. Installing Realtek's own driver nearly doubled speed to 1.1 Gbps, and enabling Ethernet flow control pushed it to about 2 Gbps. Switching to a newer UGREEN adapter with the RTL8157 chip, which bursts 8 packets, fixed it completely with zero dropped frames up to 2.4 Gbps.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://blog.tymscar.com/posts/studiodisplayethernet>

## Questions this post answers

### Why is my USB-C Ethernet adapter limiting download speed but not upload speed on a 2.5 Gbps connection?

The bottleneck is often the USB burst size the adapter's chip uses when transferring data to the computer, not the Ethernet link itself. A Realtek RTL8156B adapter sending only 4 packets per burst (bMaxBurst 3) drained data to a Mac slower than the 2.5 Gbps coming in from the router, overflowing its internal buffer and dropping frames, which caused heavy download loss while uploads stayed fine since the Mac's own USB burst handling wasn't the constraint.

_Debugging an asymmetric speed drop like this is easier when you can find write-ups others on daily.dev have already worked through._

### Does enabling Ethernet flow control help with dropped frames on a USB Ethernet adapter?

Yes, turning on flow control on a router or switch can recover much of the lost throughput. Enabling it on a UniFi console pushed download speed from around 1.1 Gbps up to about 2 Gbps on an RTL8156B-based adapter, because the adapter can send a pause frame telling the sender to buffer data itself instead of dropping it, though it can add latency to other devices like gaming PCs.

_Developers tuning home or office network performance can compare flow control trade-offs like this on daily.dev._

### What is the difference between the RTL8156B and RTL8157 USB Ethernet chips in terms of burst size?

The RTL8156B uses a USB burst size of 4 packets (bMaxBurst value of 3, which counts packets after the first), while the RTL8157 bursts 8 packets (bMaxBurst value of 7). Combined with a 2x faster link speed, the RTL8157 reached a theoretical adapter-to-host speed of about 4.5 Gbps, comfortably above a 2.5 Gbps input rate, eliminating buffer overflow and dropped frames entirely up to 2.4 Gbps tested.

_Choosing between Ethernet adapter chipsets gets easier when you can see real benchmark comparisons like this on daily.dev._

## Community take

How the wider developer community reacted, aggregated from 3 discussions and 322 comments across hackernews, lobsters (as of 2026-11-22).

**TL;DR:** Most of the discussion drifts away from the actual USB-C/Ethernet bug into tangents about io_uring throttling a game's shader compilation and broader debates about hardware reliability, with only a small side-thread specifically praising the networking writeup's diagnostic depth.

**Sentiment:** 20% positive · 65% mixed · 15% skeptical

**The case for**

- The writeup's diagrams and inclusion of Python code were appreciated as genuinely illustrative.
- Some found the debugging narrative satisfying and well-documented compared to typical posts.

**The pushback**

- One commenter wished the post showed exactly how findings were confirmed (which logs, which commands) rather than asserting conclusions.
- A side discussion accused a commenter's writing style of looking AI-generated simply due to em-dash usage, which was called out as an unfair knee-jerk reaction.

**By community**

- hackernews (mixed): The top comments barely engage with the actual adapter/driver bug and instead spin off into a long, heated tangent about io_uring, shader compilation throttling, and whether software should work around flaky hardware.
- lobsters (mixed): A thoughtful request for more rigorous documentation of the debugging process is followed by an unrelated spat over whether a comment's phrasing seemed AI-written.

**Hottest debate:** Whether software should proactively throttle itself to protect against questionable hardware/OS metrics, versus blaming and fixing the underlying flaky hardware or kernel accounting instead.

**Open questions**

- What exact logs or commands let the author confirm dropped-frame counts and rule out other causes?
- Is there a reliable replacement for the iowait metric that io_uring doesn't break?

**Highlights**

> I enjoyed this. I wish articles like these would show exactly how they confirmed their findings. "there were no dropped frame errors in macOS’s logs at all." What logs? Where? "Then I can check the chip’s own counter of dropped frames" How did you do that? Is it another log? A command? Anyway, this isn't meant to be a condemnation of a post which is worthy on its own—they made a cool diagram! They showed (some) python code. But I would love a book or course that walks through each networking layer step-by-step and shows how to test various pieces and understand the output. Not an LLM, but real knowledge.
> — [edwardloveall on lobsters · 35 points, 1 comments](https://lobste.rs/s/qzerls/my_monitor_was_cutting_my_download_speed#c_bfglzs)

> This reminds me of a (completely different, but felt equally) bizzare problem with ghostty (the terminal emulator) on Linux and Steam, specifically Baldur's Gate 3. At the start of the game after an update, BG3 compiles the (Vulkan) shaders. To avoid hammering the computer, it checks the load, including I/O pressure, and then throttles itself. It turns out that Ghostty uses io_uring, and even if it's not doing anything this somehow counts as pressure. This was made worse by the fact that I usually have a dozen or so Ghostty windows. End result, BG3 compiles shaders using 1 core, throttled, instead of all available 20 cores. Instead of a minute or so, it'd take several hours. (For closure: I configured ghostty to use epoll instead. Probably not optimal but works around the problem.)
> — [senko on hackernews · 9 comments](https://news.ycombinator.com/item?id=50589046)

> Interesting to see the "io_uring vs io pressure measurement" issue come up here; it's turned up on qemu-devel recently too: https://lore.kernel.org/qemu-devel/4EF13FE7-E6F3-42E2-83F9-6... If I am reading the thread correctly, the io_uring maintainer's position is that applications trying to use the metric that "poll via io_uring" handles differently from "poll via epoll or select" are fundamentally misguided and measuring something that doesn't make sense, so they don't want to slow down io_uring to placate those apps.
> — [pm215 on hackernews · 1 comments](https://news.ycombinator.com/item?id=50591173)

**Source threads**

- [hackernews](https://news.ycombinator.com/item?id=50587875) · 664 points · 313 comments
- [hackernews](https://news.ycombinator.com/item?id=50508373) · 7 points · 0 comments
- [lobsters](https://lobste.rs/s/qzerls/my_monitor_was_cutting_my_download_speed) · 114 points · 9 comments

## Similar posts on daily.dev

- [Monitors That Throttle Downloads](https://daily.dev/posts/monitors-that-throttle-downloads-wkg22wwxe) · Hackaday · 0 upvotes · 0 comments
- [Framework's 10G Ethernet module exposes USB-C's complexity](https://daily.dev/posts/framework-s-10g-ethernet-module-exposes-usb-c-s-complexity-ip92umde8) · Jeff Geerling · 2 upvotes · 1 comments
- [Debugging my new network, when 10 Gigabit Ethernet Runs at 300 Megabits](https://daily.dev/posts/debugging-my-new-network-when-10-gigabit-ethernet-runs-at-300-megabits-intohpmzt) · Scott Hanselman · 3 upvotes · 2 comments
- [This $10 fix made my expensive TV's internet way faster](https://daily.dev/posts/this-10-fix-made-my-expensive-tv-s-internet-way-faster-av9ptbnic) · XDA Developers · 0 upvotes · 0 comments
- [New 10 GbE USB adapters are cooler, smaller, cheaper](https://daily.dev/posts/new-10-gbe-usb-adapters-are-cooler-smaller-cheaper-9q5yp1afn) · Jeff Geerling · 0 upvotes · 0 comments

---

Tags: [#networking](https://daily.dev/tags/networking), [#mac](https://daily.dev/tags/mac)

[View this post on daily.dev](https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"My monitor was cutting my download speed by 75%","url":"https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis"},"datePublished":"2026-11-22T08:49:37.408Z","dateModified":"2026-11-22T18:22:04.768Z","description":"A USB-C Ethernet adapter plugged into a Studio Display XDR was quietly limiting download speed to 600 Mbps on a 2.5 Gbps symmetrical line. The cause traced to...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/bd319bb8ef04af83dabccba4a0c47ab9?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/bd319bb8ef04af83dabccba4a0c47ab9?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":4},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"networking,mac","timeRequired":"PT9M"}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"My monitor was cutting my download speed by 75%"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/my-monitor-was-cutting-my-download-speed-by-75--yigz5imis#faq","mainEntity":[{"@type":"Question","name":"Why is my USB-C Ethernet adapter limiting download speed but not upload speed on a 2.5 Gbps connection?","acceptedAnswer":{"@type":"Answer","text":"The bottleneck is often the USB burst size the adapter's chip uses when transferring data to the computer, not the Ethernet link itself. A Realtek RTL8156B adapter sending only 4 packets per burst (bMaxBurst 3) drained data to a Mac slower than the 2.5 Gbps coming in from the router, overflowing its internal buffer and dropping frames, which caused heavy download loss while uploads stayed fine since the Mac's own USB burst handling wasn't the constraint. Debugging an asymmetric speed drop like this is easier when you can find write-ups others on daily.dev have already worked through."}},{"@type":"Question","name":"Does enabling Ethernet flow control help with dropped frames on a USB Ethernet adapter?","acceptedAnswer":{"@type":"Answer","text":"Yes, turning on flow control on a router or switch can recover much of the lost throughput. Enabling it on a UniFi console pushed download speed from around 1.1 Gbps up to about 2 Gbps on an RTL8156B-based adapter, because the adapter can send a pause frame telling the sender to buffer data itself instead of dropping it, though it can add latency to other devices like gaming PCs. Developers tuning home or office network performance can compare flow control trade-offs like this on daily.dev."}},{"@type":"Question","name":"What is the difference between the RTL8156B and RTL8157 USB Ethernet chips in terms of burst size?","acceptedAnswer":{"@type":"Answer","text":"The RTL8156B uses a USB burst size of 4 packets (bMaxBurst value of 3, which counts packets after the first), while the RTL8157 bursts 8 packets (bMaxBurst value of 7). Combined with a 2x faster link speed, the RTL8157 reached a theoretical adapter-to-host speed of about 4.5 Gbps, comfortably above a 2.5 Gbps input rate, eliminating buffer overflow and dropped frames entirely up to 2.4 Gbps tested. Choosing between Ethernet adapter chipsets gets easier when you can see real benchmark comparisons like this on daily.dev."}}]}
```

