NickKz

Over the past few days I have been looking at UEFI capsule releases from a few different vendors, trying to extract various thunderbolt 4 firmware files embedded within so I can compare them I am using UEFITool (https://github.com/LongSoft/UEFITool), and I am able to pretty easily understand HP and Asus's image formats (even though the way HP stores the firmware in the capsule and updates it is different than how Asus does it) However for MSI, I am stumped...even for an identical chipset board that Asus uses, a Z690 board, with the same TB4 maple ridge controller...I just can not figure out where in the cap MSI are storing the TB firmware image or how they are updating it in the UEFI code. With Asus: Updater app is TbtNvmDrvShellUpdate (5EDCCC1A-F4CA-42F4-8D3E-CA4C0B396BF7) Actual update is performed by an EFI_FIRMWARE_MANAGEMENT_PROTOCOL (86C77A67-0B97-4633-A187-49104D0685C7) instance Protocol instance for Thunderbolt installed by driver CCF23F50-F7C1-4F00-8E70-13643C37E8B0 NVM firmware update loaded from EED54281-1C11-4358-BF5A-F64995FBF11B on most (all?) ASUS boards with built-in Maple Ridge / TB4 With HP: DXE updater at EED54281-1C11-4358-BF5A-F64995FBF11B Actual TB4 firmware stored uncompressed in padding section at the end of the capsule, alongside several other binary firmware packages for additional I/O chips, USB-PD etc. If anyone wants to take a stab at this, https://www.msi.com/Motherboard/MEG-Z690-ACE/support I'm going based off of the assumption that original release (7D27v10 https://download.msi.com/bos_exe/mb/7D27v10.zip) has _some_ TB4 firmware in it...I checked the other releases as well, but couldn't find anything there either. The other possibility is, MSI updates the TB4 firmware some other way, and it's not stored in the UEFI capsule at all...but I don't believe this to be true, as Intel have said as a policy with TB4 specifically that the way OEMs are supposed to handle updates is through UEFI caps. MSI tech support of course are no help at all, but that's not surprising.

Can you share the source if you still have it? I’m gonna try to work on this again. This computer has been a nightmare, the mobo died ugh… edit; I figured out the main issue — you actually can’t edit STL files directly…I didn’t know that. I spent an hour trying to do something that’s not even possible lmao.

Just got an email today, their BEST estimate is early April for a new motherboard Just fkn shoot me if I ever think about buying another HP product ever again

This actually happened before that, but I had already placed the order for the drives . They won’t fit anyway, had to return them. Heatsinks were wayyy too tall.

Hoping there are some HP experts that can chime in with some tips… Is there anything else I can do here? Support wants to send an onsite tech to repair it but the motherboard is out of stock, could take months to get here… I have disconnected everything except a USB keyboard. I tried both taking the OEM ram out and replacing it with something else, no change. Tried removing all the RAM as well, nothing. No beep codes. I do not have a spare CPU to try, but the CPU heatsink is warm to the touch, if that means anything. I’ve also tried long pressing (30 seconds) the power button, no change. https://drive.google.com/file/d/1W2gFysOYIZRzcAAZK47F3DZroQzs1NR-/view?usp=drivesdk

From the reviews I have read it seems like the random read and write as well as the IOPs are around the same as the previous gen4 (with these crucial specifically), but with large block sequential R&W is where the improvement is. So, the answer is —- maybe It also looks like the Phison controller is better in the gen5 drives, but produces more heat. The integrated heatsink version of the drive is HUGE! So big that it won’t even fit in my case, lol. So I guess for now the decision has been made for me. I think I’m going to return them and get the t500 instead.

I have an HP Z2 Mini G9 It uses a W680 chipset but for whatever reason on this device HP does not give you any PCIe Gen 5 enabled lanes at all Both M.2 slots are Gen4x4 and the GPU riser is Gen4x8 (x16 mechanical). First question is why would HP do this? Is there some kind of significant cost or component upgrade fee on their end to enable Gen5 support? What specifically “enables” gen5 pcie lanes on W680? Second question is, is there any benefit at all to running two Gen5 drives in RAID1 like the newer Crucial gen5 t700, over two of the gen4 t500 drives in the same raid1 config? Might as well also ask about the heatsinks…these SSD mfgs seem to charge quite a bit extra for the heatsink equipped versions. I’m guessing probably a better idea to just get a third party heatsink and buy the bare versions?

This is a small form factor computer and it only has one low profile PCIe slot, but it uses non standard brackets...Don't get me started on that. I'll rant for days...haha. If I could find a low profile thunderbolt 2 (or 3) card, it's possible I could just run it without the bracket, however I am actually using the PCIe slot for something else right now, and it would suck to have to give it up. If anyone knows of a low profile TB2/3 card though, link it and I'll check it out. Another concern though -- a lot of those thunderbolt AICs are system specific. IE, you put the card in one computer, and it won't work at all, due to firmware / bios incompatibilities between vendors, etc. Some of them require GPIO pins from the motherboard in order to work properly too, and I know for a fact this computer doesn't have such pins to connect to. I've got quite a few thunderbolt 2 devices...a couple Promise storage arrays, 8 and 16Gb fiber channel adapters, 10Gb ethernet, as well as some UAD apollo audio interfaces. From time to time I also use the Apple thunderbolt to firewire adapter to connect older hard drives and firewire audio / video stuff. RE: The presonus audio interface -- I've read some reports of placing a thunderbolt 3 (or 4?) hub in front and then putting the TB2 device behind that, and it will work, but some people also said they tried it more recently and it does not work...so ::shrug:: EDIT: I have been told specifically that thunderbolt add in cards will absolutely *not* work. The TB add in cards that you see available are vendor specific and require customized firmware and platform / bios support to actually work, as well as GPIO pins from motherboard to provide power to bus devices.

I recently updated the system BIOS firmware for my HP Z2 Mini G9, and the update contained within it a newer thunderbolt controller nvm firmware image. This computer has a Maple Ridge / JHL8340 / 1134 Thunderbolt FlexIO module (HP markets it as “Thunderbolt 3 / USB4", Intel says it’s Thunderbolt 4, which is confusing to say the least). Originally, the thunderbolt module was running nvm firmware 31.0. On this firmware, older thunderbolt version 1 and 2 devices worked, for the most part. Sometimes I would have issues with hot plugging, but a cold boot always fixed it. On this new version 39.3 firmware, now all thunderbolt 1 & 2 devices will not function at all. Even worse, there was no warning at all in the bios update. No notes, nothing. There are similar reports for other OEMs. Asus at least warns their users that if they update, they will no longer be able to use legacy thunderbolt devices. HP doesn't tell you jack. To be perfectly clear, these older devices have worked mostly fine, for years now. They also continue to work fine even on the latest M3 Macs. Apple made it clear from the beginning that they would do whatever they could to maintain compatibility going forward, and they (for now) have lived up to that promise. Intel on the other hand seems to have deliberately pushed this update for the specific purpose of disabling older device compatibility, and without telling anyone. I imagine anyone who is familiar with Intel is not surprised at all by this move. The way Intel have structured the Thunderbolt group, there is no way to actually contact anyone there. They don’t even allow posts about Thunderbolt chips on their community support forums. They direct you to go through your OEM. HP support, as you probably already surmised, are useless. I went through three bottom level techs until I finally got one who agreed to escalate my case to a level 2. I’m not hopeful that I will be able to get anywhere via this channel. On my own, I have discovered a few things. 1.) The newer thunderbolt controller firmware is actually embedded within the bios system firmware image. This is rather unfortunate, as both the USB controller and the PD controller have separate firmware images that can easily be changed. 2.) With HP, they post the current bios, as well as the previous two releases on their public website. None of them actually contain the older v 31.0 thunderbolt controller nvm firmware image. 3.) I suspect the thunderbolt controllers came flashed from the factory and there was no need to package the older firmware image into the system bios OR there is an older system bios firmware image that was pulled from their website that contains the older thunderbolt controller firmware image. I believe the situation is exactly the same for Asus users. I have read reports of people trying to revert their BIOS, but it doesn’t actually revert the thunderbolt controller chip firmware. It’s the same for me. There is one report from an Asus user who asked Asus support to package a custom flash for them, and they actually did it. Wow. Go Asus. https://answers.presonus.com/67591/can-quantum-be-connected-to-a-thunderbolt-4-port?show=78370#a78370 At this point, the only other thing I can think of trying is ordering a new thunderbolt module, and hoping that it ships with the older firmware image on it. At that point, I can put it in, and load up Ubuntu, and then dump the firmware image to a file. I’m not entirely certain how I would go about loading it onto another module (dumping it is easy, flashing it is something else…some systems can be flashed easily, this one can’t…there is no “nvm_non_active0” node in sysfs for this particular controller! https://docs.kernel.org/admin-guide/thunderbolt.html). I managed to find a Windows flashing utility that Intel has publicly available, but it is for NUCs, and it’s possible that they could be using a different firmware image format. At the very least, I could give it a try I suppose. The worst that could happen is I would brick the controller, and just return it. Oh well. Sucks for the reseller I bought it from. They can thank Intel for that. Wondering if anyone else has seen anything like this and possibly has any ideas of what is actually going on, and how to reverse it. I’ve read some posts on reddit saying this was all about “security” but I don’t believe that to be true. Even with Kernel DMA protection turned off and SL0 (no security) mode enabled on the TB controller, older Thunderbolt 1&2 devices still do not work at all. So Intel deliberately did something to disable legacy device signaling, for some specific reason not at all related to security. If anyone is thinking about buying into the thunderbolt 5 ecosystem, I would think twice about that. If they did it before, they can do it again. Poof, your devices are now all bricks. Just go buy the newer models! Thunderbolt 6 comes out in 2027! Lol... Hopefully going forward, with USB4, and USB4 v2, there will not be these kinds of shenanigans, as Intel will not be at the helm asserting full control over everything. It's really a wonder that USB devices from 20 years ago still work fine on newer computers. That should be the standard.

Dude! That looks amazing. Holy crap. Sorry I've been busy with holidays and such. Are you able to get the SFP holes and under bracket screw tab locations from this? https://www.thingiverse.com/thing:4291895/files

The HP blank slot bracket is 0.0325” thick And the Intel LP bracket is 0.0355” thick Thanks!

Yep, I found the LP version for the dual SFP x520. https://www.thingiverse.com/thing:4291895/files And here’s the HP blank bracket I measured with a digital caliper, what I’ve got so far anyway. I wish I could just learn to make the model myself. Maybe I’ll try again this weekend.

I had about an hour to spend on it, and yeah, I watched a couple videos and still could not figure it out, got frustrated and gave up. I couldn’t even figure out how to draw a simple shape to an actual ruler like with true to size 3D dimensions. The way the ruler works, it’s just totally nonsensical for whatever reason. It’s not like Photoshop, I can tell you that much, lol.

Wondering if anyone would be willing to help me out with this, I am hoping to get a custom bracket 3d printed / fab'd / etc. that will work in this system, as I posted about here: I have a digital caliper and I have measured the blank bracket, and I also found a STL of the Intel card online already...If I knew how to actually modify the STL file and make my own model, I would. I tried, but I couldn't figure out how to use Fusion 360, heh. I would be willing to compensate you for your time and help. Let me know, cheers.

Here it is in the manual… “custom rear bulkhead” and “it intent to supported HP certified dGFX card” lol So much for “standard low profile graphics” Seriously though if anyone knows how I could possibly get a bracket made I wouldn’t mind paying a reasonable fee I hope other people who are considering buying this garbage find this post and change their minds! I still have trouble believing that they created their own PCIe bracket format just so they could sell marked up GPUs…I can’t imagine there are even that many people buying these models, there’s almost nothing on the web about them except Linus’s “glowing” review LOL — I could only find a handful of forum posts on Reddit and on HPs own forums; Serve the home, etc…I mean it doesn’t look like these are very popular at all.

Edit: Actually it explicitly does say “standard low profile graphics” which it absolutely isn’t I suspect there are no other low profile gfx cards that would fit in this system besides the custom bracketed models that HP sells... https://www.hp.com/us-en/workstations/z2-mini.html The Intel x520 card itself fits and works just fine, it’s just the bracket that is the issue really... This looks really bad for HP, as if they needed any help in that department lol But honestly for my first PC in ten years this has been a disaster, and all the online trolls who love to yell at Apple for doing Apple things and love to shout “just get a PC” don’t seem to be as loud when it comes to PC companies doing stuff like this I don’t want to make any assumptions here but I would reckon that the sole purpose of using proprietary brackets has nothing to do with space or form factor and has everything to do with HP charging more for their custom bracketed version of the gfx cards they support in here like the A2000.

That’ll work! Still kinda jank but better than nothing if I can’t find or make a bracket. I still would love to know if anyone has ever seen this before and if there’s a name for it EDIT: So the press fittings from the SFP cages definitely touch the back of the PCIe tray if wobbled too much. Might have to get some electrical tape too, aside from the zip ties. Or someone leet with the tools could make me a bracket? A man can dream

Additional photos…is there a name for this specific kind of bracket? Maybe I can find a bracket that already made. Surely HP didn’t make this standard for just this pc?!

I bought this hunk of junk from HP called a Z2 Mini G9 and I thought it just used regular low profile PCIe brackets…but apparently not. I tried to install an Intel X540 dual channel 10GbE low profile NIC in here, and the bracket won’t fit. It runs fine without a bracket, but that’s so jank… From what I can tell there is no way I can modify the exiting bracket to fit. Even if I cut off the bottom tab with a dremel (which I did try already lol) the top part still protrudes, and then if I cut off the top part, then there’s no retention mechanism anyway, so it would basically be like running it without a bracket. I would have to cut off the top part and then cut a tiny tab out to get it to align…ugh. Anyone here have the tools / skills to make me a bracket for this? I’ll definitely reimburse you for your time and help Additional photos incoming.

What I ended up doing was installing this app called "betterdisplay" and it allows you to do HDR brightness upscaling. Basically they use the same approach Microsoft does -- they give you a calibration tool and a brightness slider. Tl;dr, I can basically make my Asus ProArt PA32UCX act like a Pro Display XDR...kind of. You still don't get luminance management in the system color profiles. But perhaps there is a way to get that, too, and I just haven't figured it out yet

I think I understand it now. They essentially created their own HDR display technology that they call “EDR” which requires that the display is one of theirs to work properly. On Windows, laptop makers can get access to os APIs that allow them to essentially implement their own version, but even then, if you’re using an external monitor, it won’t work. The brightness control will be locked out / on maximum and SDR content will look trash with HDR enabled. As a workaround Microsoft gives an HDR calibration tool along with a SDR brightness slider, which still won’t look anywhere near as good as a display where the OS has complete control over it, and every dimming zone, and knows exactly how bright it can get, etc. I guess Apple could also do what Microsoft does but then it would essentially mess with their existing and objectively superior tech.

What specifically is Apple doing on their own screens though? With every other display, turning on HDR makes SDR content look like crap, but on Apple’s screens they don’t seem to suffer from this same problem. In fact their HDR enabled screens don’t even have a toggle, it’s just always on, and they are able to display both HDR content and SDR content side by side exactly as you would expect. Asus recommends that you only turn on HDR when actually viewing or editing HDR content, and even then, they suggest making the video full screen, as even simple video controls like the play and pause and timeline won’t look right — which seems kind of ridiculous IMO.

Yeah they look the same to the naked eye but there are very tiny difference in sizes depending on your existing antennas, it's wonderful isn't it...I forgot what I had originally in this micro desktop server case thing. Chances are you won't need to mess with the antennas if you have a modern laptop / desktop. Since it's Intel, I basically trust them to update the firmware of these new wifi modules when the draft is finalized, however, since it was only like $20 anyway, it's really not that big of a deal...I don't even have wifi 7 access points yet, I am using Ubiquiti U6E's Just figured I would get it since it's something new to try and test out, and I didn't actually have any existing 6E module to test...figured since the AX is around the same price, there's no point in getting that when the BE200 was just released and is super cheap.

I have an Asus ProArt PA32UCK and it supports HDR, however I almost never enable HDR when I’m on a Mac, because…well, it looks like hot garbage, unless you’ve got full screen HDR graded content. But the Pro Display XDR (as well as MacBook Pro screens that support HDR) doesn’t have this problem…so what are they doing differently? Is there a way I can get my Asus ProArt monitor to work like a ProDisplay XDR works? On Windows Microsoft at least has a brightness slider for SDR content when you turn on HDR as well as a HDR calibration tool…but I’m not sure how to get to the equivalent on Mac OS?

I just got the new Intel BE200 WiFi 7 module and ran into an issue with the internal antenna connectors. If what you are asking about are the actual antennas and not the module, then this should help: https://www.everythingrf.com/community/what-are-i-pex-mhf-connectors Sounds like you want the actual WiFi module though, as you’ve probably already got some kind of internal or external antenna system in your case.