Justin_

Hummm, even at 1.0X8 I am getting the same (expected) FPS. Maybe I am reading the data wrong? (**edit** yeah, they are working at gen3, they were just in like a power saving mode idling at gen1) Would love if my onboard NIC worked, my USB bus did not have issues, and I could add USB controller cards to this system tho. Prolly just need a new Mobo after all.

I also went in and tried setting the PCIEX16 bifurcation to 8x8 rather than auto, but that just gets the card running at 1.0x8 rather than 1.0x16, so thats no help. I swear, that was not even the initial issue I was chasing.

M.2 moved to secondary slot and GPU drivers reinstalled and they are still running at 1.0X16 and 1.0X4

Yes, so, TL;DR GTX 1070 in PCIEX16 (CPU PCI-e x16) is running at Gen1 by 16 lanes instead of Gen3 GTX 1050 in PCIEX4 (chipset PCI-e x4) is running at Gen1 by 4 lanes instead of Gen3 Cant get system to pick up additional PCI-e based USB controller cards System has issues with the buffer on the USB bus Onboard NIC does not work Im starting to feel more like this might be a PCI-E lane availability issue. Think ill try moving the M.2 to the secondary slot so that its on the chipset rather than the CPU and see what happens to the USB controller then (it should have 4 lanes for the M.2, 8 lanes for the dGPU, and 8 lanes for the iGPU, but this /should/ work (I think))

Went into the BIOS and did the following Disabled the iGPU Disabled the onboard DAC (never use it anyway) Enabled Above 4G Decoding (why not) Set Resizable Bar to Auto (dont think it will do anything with a 1070, but whatever) Turned off Wake on LAN Turned off onboard NIC Turned off SATA Hotplug Set "PCIEX16 Slot Config" to Gen3 (this should be the 1070) Set "PCIe Slot Config" to Gen 3 (think this should be the rest of the expansion bus) however I have no notable changes to report. Im going to try reinstalling the GPU driver here and see if that fixes the link speed issues.

Hey everyone! I currently have a system with an AMD 5600G (w/onboard GPU) and an AORIS X570 Elite mobo (manual linked here) and I am having PCI-e device issues and I am not totally sure if its a mobo flaw, a PCI-E lane bottleneck, or something else and was hoping to get some help On top of the onboard hardware (USB root controllers, STAT controllers, etc) I have a GTX 1070 (gaming card), a GTX 1050 (this is just used to output to more than 3 displays), and a Realtech 2.5Gb network adapter as the onboard networking is not working right. As mentioned, the onboard Intel I211 Gb nic does not work, but I also have issues with my USB audio interface dropping out. Additionally, if I try any PCI-E USB 3 controller cards in the PCI-e slots they refuse to work at all. I have tried 2 cards (one with a VLI VL805-Q6 root controller with VL813-Q7 hub, and another with a Renesas D720201 root controller), attempted to install drivers automatically and manually and none work. With Black Friday sales coming up and other sales already happening I am looking at getting a new CPU or Mobo to try and address this issue.I dont want to go for a more expensive platform upgrade currently due to limited budget (yay being back at school :P) Notes on the audio interface: I use a Focusrite 2i2 gen2 audio interface. Occasionally it acts up and gives me corrupted audio. Its more likely to work if it is isolated alone on its own USB controller, but that is not a guarantee. The issues feel like some sort of buffer issue as once corrupted the audio starts to hitch and be delayed (both coming in from the microphone and out to the speakers). Additionally, if I go into the 2i2 settings and crank the sample rate to 192000 and the buffer size to 1024 (highest setting on both) so that there is a bigger buffer while still keeping the latency down then I dont have audio issues near as often. Its been good for over a week rather than just the few hours it would hold up for between resets before. Notes on the Intel I211 NIC: This NIC never worked. I bought this mobo used and the seller told me it did not work for them either. Beyond drivers, I didnt poke at it too much tbh. Possible Issue 1, Motherboard is partially DOA or failed: First thought is that the Mobo is damaged. This could be cause from being marginal from the factory, something the last owner did, lighting strikes/ESD event, or some other such. Maybe one of the PCI-e lane trace pairs was not etched properly or maybe a chip on the bus is faulty. Not sure, but this was my first thought. Possible issue 2, 5600G does not have enough PCI-e lanes to back this system, maybe the past owner had the same issue? (dont know what chip they ran, too late to find out now) From what I read, the 5600G, like the 5600 has 24 lanes of PCI-e, however, it is laid out as 8+8+4+4 lanes. The first 8 lanes reserved for the onboard GPU. The next 8 lanes for the "first" PCI-e slot (GTX 1070+audio interface). Another 4 lanes is for the first M.2 slot (boot SSD). Lastly 4 lanes for the X570 chipset that in turn provides 8+8 lanes for other things, PCI-e 4X (GTX 1050+audio interface), PCI-e 1x (unused, under 1050 cooler), PCI-e 1x (2.5Gb NIC), second M.2 (Unused), what looks like 4 SATA controllers (2 ports per controller including M.2's, controller 1 port 0 used for 2TB SSD), Seemingly 4 USB root hub, etc. Looking at PCI-e lane assignment in PowerShell using this script I can see; (red text is my interpretation of what im looking at, correct me if wrong) The faulty NIC is using 1 lane of gen 1 PCI-e, tho it reports this as its max gen and number of lanes (intels website backs this up) The NVME SSD Controller is linked at 3.0X4, this is good GTX 1070 is reportedly running at 1.0X16, tho it can link at 3.0X16 and its connected audio controller is linked at 1.0X16. Even when I launch a game it stays at a 1.0X16 link, this could explain issues launching certain games recently (will reboot, check BIOS and report back once I submit this thread, still runs TF2 at over 120 FPS at all times peaking ~300). GTX 1050 is currently linked at 3.0X4 (this is as expected as the lane is 3.0X4 electrically) 2.5Gb NIC is linked at 2.0X1. The slot is electrically a 1 lane slot and the documentation I can find lists this as a 2.1X1 link, so sounds right It seems that 2 of the USB host controllers are linked at PCI-e 4.0X16 and 2 are linked at 3.0X16. So what im wondering here is as follows; A) Should I suspect the Motherboard as faulty and replace it? B) Should I poke at PCI-e links? Is there a recommended way/guide of doing this? C) Should I replace the CPU with one that does not have an onboard GPU taking up 8 PCI-e lanes and expect that to fix my issues?

Then again, looking at the actual per-process usage I don't see anything using it I can see maybe 4GB of usage here, not just shy of 10GB

Actually im not 100% sure that is the answer either. Resource Monitor is telling me that only 262MB are hardware reserved. the iGPU is allowed to steal upto 50% of system RAM, so that is where the number 7.9 is coming from. Not sure why it wouldn't be useable then though.

So yeah, there is 7.9G of shared memory, but the dGPU still has the 4GB onboard memory

As far as normal paste goes, it could vary by like 1 or 2 degrees, so not much. Specific thermal pastes making a difference are mostly just for extreme usages like sub zero or LN2 overclocks. A gallium based paste (like conductonaut) would get a lower temp, but comes with some downsides, like how it will literally dissolve anything made of aluminium and short things out if it drips. After doing some googling it seems that the 5960X is just a fkin hot chip. Apparently those temps are kinda normal. So I guess it might be worth researching some gallium paste then.

iGPU dGPU Top RAM usage in Task Manager

You can get thermal paste form best buy for cheap, heck even toothpaste should do better than that. I ran toothpaste for 3 months once and it didn't even dry out. The one other thing you can check is to see if the pump power cable coming off the block is plugged in. if it is a 4 pin fan plug you should see RPMs reported back. Can't remember if a 3 pin plug will do that. If the pump is not pumping than it will heat up fast, but if it's instant that still sounds like a paste issue.

///somewhere/// on the internet there are extra files you can throw into the handbreak directory that will allow it to copy off DVDs, I imagine its the same with BluRay. Its easy to find, but don't know if I would get in trouble if I linked to it. Its technically a pirating technique.

Look at bitcoin mining gear. Lots of people used those there for that exact reason. A lot of them are 12 volts only, so you might need a pico PSU for the motherboard and any SSDs you might use. Something along the lines of the 2 following things. https://www.amazon.com/dp/B08WCNGMP3 https://www.amazon.com/dp/B09156HP1B The thing is unless you have free server PSUs it will probably end up being the same price and not mount to your case.

As far as those temps go, I would pull your AIO block back off and make sure you have good thermal paste coverage and that you removed any stickers. It should take a good while to get to temperature and I would expect it to be quite a bit lower. Once you fix the temp problems your fans should chill out (no pun intended) As for RGB software, just go to your motherboard support page downloads section and download there RGB software from there.

Hey guys, Im having a bit of an issue with THIS Dell Precision 3561. For whatever reason Windows (or an application?) decided that my Intel UHD graphics need to use the maximum allowable 50% of my system RAM and I don't know how to change that back. I have tried updating graphics drivers and just updated the BIOS with no change. Cant find any settings in the BIOS to set the max or startup shared memory amount either. Looking in Intel Graphics Command Center I also don't see settings there. Even after a fresh startup it is still using 7.9GB. I have the Intel UHD iGPU and the NVIDIA t600 DGPU in this system and both show the same 7.9G shared memory, so I'm not 100% sure its the iGPUs fault. Beyond that I have tried running a few games/apps on the laptop that could use the DGPU, but it always wants to go to the iGPU for some reason. Don't know if that is related in any way.

Win 10 pro i7-8750H 6C 12T CPU 64GB RAM 1TB Crucial NVME SSD - boot 2TB Samsung SATA SSD - bulk RTX 2070 GPU All this is in an Asus ROG Strix SCAR II laptop that I keep plugged in

File explorer keeps breaking for me basically whenever I try to transfer files, even if I use third party software like teracopy. It also wont show thumbnail previews and is doing other strange things. For example, I opened a file explorer window, navigated to my desktop, and deleted some things, but those things are still there on my windows desktop even after restarting the file explorer service. I keep having to kill it in task manager and open it back up again, but its a PITA. Is there an option in windows to verify the integrity of system files (like steam has?) or anything else I can do?

Hey guys. I have an ASUS GL504GW with the 8750H and RTX 2070 that I bought about 18 months ago. It has been running great for me until recently. For the last 3-4 months I have been using it as a desktop replacement I have the lid shut on a cooling pad and have monitors, keyboard and mouse, headset, etc plugged into it and it has been running fine. In the last few weeks I noticed a bit of a dip in performance while gaming, but that happened around the same time the CPU fan started to fail so I didn't think much of it. In the last week I got a new CPU fan and popped it in. The new fan is working good as new, Here is the problem, a few days after I got the new fan in (and after a new install of windows for an unrelated reason) it started to hibernate when the CPU was under load with the lid shut. And this was not just with gaming, this was with heavy web browsing and multitasking as well. Stuff that, while above idle, should not be that bad of a load normally. I opened up HWMonitor and found that the CPU was hitting 100C. I figured that maybe I messed up the thermal paste when I put in the new fan, so I went back in and repasted it all. This laptop is a mess to paste. Not only do you need to paste the CPU and GPU, but also the VRAM and power delivery. All of this with no pads, just paste. Anyway dropped on some MX-4 and popped it all back together. I thought that would fix my problem, but nope, still 100C, still hibernating. If I have the lid open on the other hand it will '''only''' get to upper-80s to upper-90s. It however wont hibernate and crash. Right now I have it running with the lid open, but it does not really fit on my desk like that. This worked fine for month before, and the time before that I was still able to game on it just fine. It seems like a thermal issue, but I just replaced the fan and paste, so I don't know what else it could be there. Any advice on what to look at???

Hey guys. Im looking at my desktop (listed below) and my laptop (below) and i'm starting to wonder if there is any reason to not just use my laptop as my desktop. My laptop should be just as powerful if not more so than my desktop. I figure I can sell the CPU, Mobo, RAM, and case for $800-$900 and free up a lot of space. Is there any reason to not do this? The only real thing that has been holding me back up to this point is 1) I had 8-10 HDDs in my desktop, but this has now been moved to a NAS with dual parity drives. 2) I like my PCI expansion slots. But I really dont use them for much. The occasional capture card, but i've moved to USB cards at this point. Desktop: AMD Ryzen 2600 4*16GB 64GB G.Skill RGB RAM ASUS Crosshair VI Extreme Mobo ASUS STRIX GTX 1070 Corsair RM 650X PSU phanteks enthoo evolv x Countless NVME SSDs, SATA SSDs, and SATA HDDs Laptop: Asus GL504GW Intel i7-8750H CPU RTX 2070 GPU 16GB RAM (will upgrade to 64GB if possible, 32 GB if not) 500GB NVME SSD (Will upgrade to 1TB plus a secondary drive)

IT WORKED! Finally after like 2 hours of screwing with this I can play games at 100hz

Ok, after actually reading the error I found a workaround. The error is that the file does not exist. The workaround was installing the take ownership hack from How-To Geek, navigating to C:\Program Files and using Take Ownership on WindowsApps (hidden folder), then navigating to the NVIDIA folder in the WindowsApps folder, then manually launching nvCplUI.exe. This does not fix the issue on the context menu, but it is a workaround.

I installed the 2004 update and now the installer does not fail, but I still can't launch NVIDIA control panel