Dual Monitor Studio Display XDR fails on MacBook Pro M3/M4 Pro

TL;DR — for anyone skimming:

If you have two Studio Display XDRs and only one lights up at a time on your M4 MacBook, the evidence so far points to a bug in macOS Tahoe 26.4.1 and/or the display firmware that shipped with it, not your hardware or setup. The one-at-a-time behavior appears to have started right after the display firmware update that runs during the Tahoe install. (Caveat: this read comes from an AI-assisted log dive — it's the most likely cause based on what the kernel log shows, but it hasn't been confirmed by Apple.)

Probably won't help (and here's why):

- Taking displays to the Apple Store / swapping them — both units work individually, and others are hitting the same behavior with different serial numbers, so a swap would likely just give you a new pair with the same symptom.

- Reinstalling macOS — if the bug is in macOS itself (likely), reinstalling installs the same bug. A fresh user account was tested first (same behavior), which is basically a preview of a reinstall.

- Buying a new Mac / upgrading to M5 — if this is an OS-level issue, a newer chip running the same Tahoe would probably hit the same wall.

- Cables / ports / docks / dongles / resolution / refresh rate / color settings — none of the usual knobs seem to move this; the failure appears to happen before the OS reaches those settings.

What's worth trying:

1. File a Feedback Assistant report at https://feedbackassistant.apple.com — volume drives priority. Mention it's the dual-Studio-Display-XDR one-at-a-time issue on Tahoe 26.4.1. If you want a specific string to cite, AppleT603XDisplayCrossbar + allocated: 0 are the log phrases that came out of my diagnostic; Apple's Displays team can tell you if that's the right pointer.

2. Use one display for now. Both work individually. Swap the active one when you need to.

3. Watch for macOS 26.5 / a firmware update. This has the shape of a point-release fix rather than something a hardware revision would address — so a new Mac or new display is unlikely to get you around it.

TLDR on the TLDR: this looks like an Apple-side bug based on the evidence, so returning gear or reinstalling probably won't help. Best bet is filing feedback and waiting for a software fix. Not 100% confirmed — but the workaround path (buy new stuff / reinstall) is a lot of effort for a low expected payoff.

I can confirm that I am seeing the same as well. 2 Studio Display XDRs, apple supplied T5 cables and a 2023 MacBook Pro with M3 Pro. It can not display to both monitors at the same time.

I opened a support case and here is what we found. According to everything online and in Apple's own internal documentation, the M3 Pro can run displays up to 2 6K displays at 60 Hz. In my testing, even taking the XDRs down to 60 Hz manually, I can still only output to 1.

They had me bring my Mac to the Apple Store and they plugged into the only XDR they had and a standard Studio Display. It worked fine until they unplugged one of the displays, then the system crashed. They ended up replacing the logic board. I just picked it up today, get home, and same issue. I can not output to 2 displays at the same time.

I think that their documentation is incorrect. The M3 Pro will not run 2 Studio Display XDRs simultaneously. I am starting to see a lot of complaints across forums for people having the same issue. I have an M4 Max MBP and it works fine.

Their next step was that they wanted me to bring both displays in with my MBP. I am not going through that. It's not worth the hassle and possibility of damaging $7k in displays in transit. They will give me $1900 for my current MBP so I am going to trade it in. The M5 will support the 120Hz anyway.

Apple needs to fix their documentation and specify that the M3Pro and possibly the M4Pro can NOT run 2 Studio Display XDRs but it can run 2 standard 5K whether it's the standard Studio Display or a non-apple monitor

The original Apple Pro Display XDR used a trick to connect with one Thunderbolt-3 or -4 cable. It actually consumes TWO display-generators when connected on a Thunderbolt cable.

XDR display

The 6K Apple Pro Display XDR, which macOS allows to connect using two HBR3 connections to a Mac, doesn't support Display Stream Compression (DSC). That would be 51.84 Gbit/s, impossible for Thunderbolt 3, but it works because the two 3008×3384 10bpc 60 Hz 648.91 MHz signals of the XDR display only require 38.9 Gbit/s total and Thunderbolt does not transmit the DisplayPort stuffing symbols used to fill the HBR3 bandwidth.

https://en.wikipedia.org/wiki/Thunderbolt_(interface)

This means the M3 and M4 PRO processor models of MacBook Pro you are testing with do not have enough display-generators to support TWO Apple Pro Display XDR 32-in, and likely the same issue with the new 27-in Apple Studio Display XDR.

Between the initial reddit post and the present day, Apple released display firmware update v 26.4. It is thought to address the un-necessary dual-pipe issue directly, likely by providing a more nuanced EDID report.

Be sure you have applied that firmware update to the display before proceeding with your plan. Applying the firmware update is done from the Mac, and does not appear to require a picture on the display to proceed.

Also, only ONE such display on a cable can be properly updated.

Yes, it appears to be the case that the displays need to be connected AT STARTUP. I was just quoting what the Apple support tech told me. At no point did they mention I would need to restart the computer, that was my discovery.

It does make sense that the computer needs to see all the displays available to negotiate a refresh rate configuration that will work. It would be nice if Apple can to support renegotiation of all displays when plugging in additional displays, since it would be a much better experience for MacBook users. I have no idea if this actually possible. I'm just very pleased they were at least able to support it on boot. I wish I had discovered that sooner. 😅

Dual Studio Display XDR — only one active at a time on M4 Pro / macOS Tahoe 26.4.1 — kernel-level failure isolated to AppleT603XDisplayCrossbar

Same symptom many of us are seeing: both XDRs work individually, only one at a time when both connected. Firmware 26.4 (Build 23E246) post-Tahoe.

Disclaimer up front: the analysis below was produced by Claude Code (Anthropic, Opus 4.7 w/ high thinking) doing a diagnostic deep-dive on my logs. AI interpretation of kernel traces can be wrong. Treat it as a lead, not a conclusion — but the raw strings it identified are real and should be verifiable by anyone else hitting this.

TL;DR finding: the failure happens below WindowServer, in the M4-family kernel display-routing MUX. Studio Display XDR drives its 5K panel in DisplayPort tiled mode (two DPIN streams per display). When the second XDR requests pipes, the crossbar reports three of four UFPs (upstream-facing ports) as noconnect to the second DFP, so the allocator can't form a pair:

AppleT603XDisplayCrossbar::makeInactiveUFPAvailable:

searching for inactive UFP with domain-mask 1...

checking ufp(100000843) -> dfp(1000003c2)... (first XDR, busy)

skipping ufp(100000844) -> noconnect...

skipping ufp(100000853) -> noconnect...

skipping ufp(100000855) -> noconnect...

displayAllocationEventOccurred: DFPs processed: 1 allocated: 0 domains: 0x1

Result: second display's pipe request is rejected → black panel. First XDR keeps its allocation and works normally.

Why this rules out everything most of us have tried: it's below user settings, so refresh rate, resolution, color profile (8-bit SDR / sRGB), deleting com.apple.windowserver.displays.plist, fresh user accounts, left-vs-right TB port groups — none can help. Link also trained fine (HBR3, 32.4 Gbps) on USB4 v2 (80 Gbps), so it isn't bandwidth.

Strong circumstantial evidence this is a post-firmware-update regression: during the Tahoe 26.4.1 installer, before the display firmware bump, both displays + built-in ran simultaneously. After the firmware updated, the one-at-a-time behavior started. Per-display UUIDs (derived from EDID) also changed across the firmware update — so the EDID itself changed, which is the most likely culprit for the crossbar now refusing to fan out to both displays.

Reproduce this on your own setup — if you're comfortable with Claude Code running Opus 4.7 (high or xhigh thinking), this prompt will walk it through the same diagnostic:

▎ I have two Apple Studio Display XDR on a MacBook Pro M4 Pro running macOS Tahoe 26.4.1. Both work individually but only one activates at a time when both are connected. Please: (1) run system_profiler SPDisplaysDataType SPThunderboltDataType and summarize the Thunderbolt bus topology and display firmware versions; (2) run log stream --predicate 'process == "WindowServer" OR eventMessage CONTAINS[c] "display"' --level info --style compact 2>&1 | tee ~/Desktop/xdr-diag.txt while I plug the second XDR in, then stop it after 30 seconds; (3) grep the captured log for AppleT603XDisplayCrossbar, getAvailablePeerUFPForDFP, makeInactiveUFPAvailable, noconnect, and displayAllocationEventOccurred; (4) tell me whether you see "allocated: 0" for the second display and how many UFPs reported noconnect. Report only what the log actually shows — don't speculate beyond it.

Filed to Feedback Assistant with the full kernel log attached and the specific failure lines highlighted. If anyone else reproduces the noconnect / allocated: 0 signature in their own log, filing your own Feedback report with the matching strings should help drive prioritization.

Not claiming this is the definitive root cause — only that it's the point at which the kernel gives up, and it's the right place for Apple's Displays team to look first.

The issue is documented in below reddit link. It's tied to the thunderbolt firmware and the handshake between the non MAX chips (MAX chips use a different driver path): https://www.reddit.com/r/mac/comments/1rzafpz/troubleshooting_the_studio_xdr_with_claude_code/

Works fine with a MAX chip but the pro and regular chips do not work. These m4/m5 chips support dual display. I have a regular M5 and did not work either. I was also told to go to the Apple Store and bring my laptop and cables to test. The test was not successful.

At the Apple Store we tried:

1. My M5 MacBook Pro with the studio XDR cable against their Studio XDR and Studio Display non daisy chained <-- did not work
2. Same test as above but daisy chained <-- did not work
3. Same two tests as above with their cables and my M5 MacBook Pro . <-- did not work
4. Plugged in a Mac Studio with M4 Max Chip and it worked (daisy chained). <-- worked
5. Used another customers M4 MAX MacBook Pro and it worked (with my cables daisy chained). <-- worked

found a workaround to the issue

full details in the macrumors post here: https://forums.macrumors.com/threads/studio-display-xdr-impressions.2479042/page-23?post=34527033#post-34527033

TL;DR: use a 4K HDMI dummy plug to make your M4 Pro believe you're connecting an HDMI display, that forces your XDR to 60Hz, then remove the 4K dummy plug and the second display comes to life.

making it abundantly clear this is a handshake bug with Tahoe vs a capability issue.

Funny these kinds of hacks is what I used to associate with our friends in Redmond and their BOSD machines

adding image for proof of the setup working like a charm.

I think "the part left out" is that for many combinations of higher resolutions and higher refresh rates at 10 bits/color, these XDR displays (and possibly the 5K studio displays as well) grab an additional display-generator to achieve those rates over Thunderbolt-3/4.

This allows them to achieve purist, top-notch data rate results without the more typical resorting to Display Stream Compression to get there.

Similarly, if you connect an 8K HDMI display on a Mac that can support one, it will use a similar TWO display generators to get those very high speed results.

I now have my setup working successfully! Hopefully my solution on my M4 Mac Studio will at least be helpful to those of you with similar chips in other Macs, as I think the issues are related, no matter the specific chip or Mac model.

Apple support said the engineers said that all three displays needed to be connected directly to the Mac to work. I had tried this before with no success, but thought it was worth a try again. Of course, initially no luck. But something the support tech said, "All displays need to be connected directly to the Mac for it to configure the maximum refresh rate," got me thinking — maybe I need to restart. That worked! On restart all the displays work! In hindsight I can't believe I hadn't done that before, but I guess I've been spoiled by years of just being able to plug in a monitor and have it work immediately.

If I unplug any monitor and plug it back in then it no longer works. It would be nice if Apple could support that action, re-configuring the monitor refresh rates without a restart when a new monitor is plugged in. That isn't necessary for my Mac Studio set up of course, but it would probably be nice for those of you with MacBooks.

For those of you having issues: try plugging your monitors in, then restarting your Mac. Make sure all monitors have the latest firmware and you're running MacOS 26.4.

I'm having the exact same issue with my employer provided computers. 16" M4 Pro MacBook Pro (Default SKU with 24 GB of RAM 500GB SSD. I've tried two different M4 Pro MacBook Pro's with the same specs and I run into the same problem: It will only display output to one studio display, whether it's daisy chained or both displays are connected directly to the laptop. Using a dummy 4K display to connect temporarily is the only way I can get both Studio Display XDR's to output an image, but even then it's max 60hz.

I also have an M1 Pro MacBook Pro (personal laptop) and it works flawlessly with both studio display XDR's -- it also only runs at 60HZ but that's expected due to the display support specs on the M1 Pro.

Apple, please provide a fix. Spending almost $7,000 on two monitors not working properly is extremely frustrating.

I am not the one you have to convince on this, but I am willing to work with you to see where we can go on this.

Standard timing without Display Stream Compression gives 105 percent utilization of one Thunderbolt to support a 6K 6144 by 2456 display at 60 Hz at HDR 10 bits/color. So when set to full color space, resolution, and refresh rate, it grabs two display-generators.

Dropping the refresh rate to 48 Hz -OR- dropping HDR 10 bits/color and going back to 8 bits/color gets you around 83 percent. Dropping the resolution should be another way around it, but we need a way to get off two display-generators.

The Mac generally does not allow pre-setting up the display parameters -- it wants to ask the display about its capabilities and "do it myself".

My educated guess about why it does not work is that the Mac sees what that displays capabilities are, sees that it can not make the full resolution, and changes to two half-displays on two display generators before you can intervene to tell it something else.

Perhaps if you set up the first display at a lower refresh rate or a low color palate or lower resolution, then Restarted, the settings would 'stick' [But NOT using two display-generators]. Then maybe the second display could be added.

another possible way to work around would be to start the display when suing safe mode. safe mode takes defaults for most settings, including that display settings are set to very low, safe values.

The "trick" with safe mode is that if you change display settings off the default, those values will 'stick' in regular mode.

If we assume Safe Mode sets the display very low resolution with one display generator, then change it very slightly upward, perhaps it would come up with those same settings on one display-generator in regular mode.

The fundamental issue appears to be:

The new 4K XDR display defaults to dual-pipe operation, whether it is needed or not. If set to 60 Hz, rather than "adaptive" it SHOULD be able to get all the needed data inside single-pipe. This should work on Thunderbolt-3/4 connections without issue.

"dual-pipe" is a fancy name for dividing the display into Left and Right halves, and driving each half separately. the older 5K XDR display used that trick to get just enough more bandwidth to run over a thunderbolt-3/4 connection without having to invoke Display Stream Compression.

The problem is that the Mac display software decides to use dual-pipe, which uses TWO display generators. The M4 Pro has a total of THREE, and using TWO for each of your two displays leaves you one short. it works on the M4 Max because it has FOUR display-generators.

In some cases, users are also finding that the 4K XDR dual-pipe operation is slightly unstable, and crashes unpredictably, on the order of once a day, with one half of the display left in 'snow'

Just confirming, I was able to connect two Studio Display XDRs with my M4 Pro macbook that was running 15.7.3. But ever since upgrading to 26.4 I'm not able to do the same. See attached image as reference.

Clearly looks like a regression introduced in Tahoe.

[Edited by Moderator]

Adding a data point that might help whoever's debugging this... I think I found a workaround, though an unreliable one...

My setup is a MacBook Pro (M4 Pro), 2x Studio Display XDR, same symptom as above, each works individually, only one comes up when both are connected (also when daisy chained).

I went as far as a full macOS reinstall. Partway through that process (before it fully completed), both external displays and the builtin MacBook screen were active at the same time.. all three, no problem! So the hardware and firmware clearly can handle this, at least in whatever reduced display-driver state macOS runs in during installation, ***. But downgraded now to 26.5.2, and still working with the workaround. Even opening and closing the MacBook Pro is working fine, as it should be.

Once the reinstall finished and I was back in the full OS, I tried daisy-chaining (main display TB with the dot into the other's second TB port)... back to "broken" only one active..., and at one point I even got audio on one display and video on the other, no clean handoff either way.

Then I tried something I hadn't seen suggested else where, plugging both TB5 cables into the MacBook Pro at the same time, so NOT one first, then the other.. but both simultaneously, into two separate ports. That worked. Not always on the first attempt, but almost always by the second try.

Not an engineer, but my guess, whichever display connects first grabs the pipe pair it needs, and by the time the second one requests pipes, the crossbar can't find a spare pair for it, hence the "no connect" behavior others have mentioned. Connecting both at once might force the negotiation to consider both requests before locking in an allocation for just one.

First I wanted to return de display, but after I discovered thi, I've since gotten in touch with Apple and they're actively looking into this rather than just processing a return, for what it's worth.. If they cannot fix this with an update, I might consider returning 1/2.

Curious if anyone else can reproduce this with the simultaneous-plug approach, would help to know if it's consistent across setups or specific to mine. It's obviously not a real fix. Still have to do the 1-2x replug dance after every unplug, waking up is working fine when you leave it plugged in.

[Edited by Moderator]