All Activity

Some sponsored reviews have appeared by English speaking Youtubers. I will not post the links because I find them very mediocre and prejudiced, repeating the LG silly marketing points (i.e. local AI with Gram chat, etc.). They present a Gram 16 with the X7 processor with the b390 iGPU and Gram 17 with the RTX5050. I am very disappointed by LG, there is no real progress, more marketing and less substance. They continue to claim of the lightest 17in laptop which is not a big feat, most laptop makers have turned to 18in for their gaming/desktop replacements and at 16in for light gaming laptops so the new Gram at more than 1.7kg!!!!, more than 300gr than the 2025 remains the "king"... If they added weight for a beefier cooling (apparently processor but also the graphics performance is better than the 2025 version, they feed more power) they could have at least went with a RTX5070ti with 12GB VRAM and be more useful for local LLM usage. But the mess does not stop here, still the same 1600p vertical resolution IPS panel since the launch of the 17incher, they should propose by now a 1800p panel, the dpi is pretty weak for a 17in screen. And GLOSSY!!!, a touchscreen but glossy. The 16in keeps the crown as super light but they stick to a 77Whr battery and only 2 speakers. It has a matte 1800p OLED panel that looks good, but we don't get any battery tests from the reviewers. Performance of the PantherLake compared to last year's ArrowLake is significantly better but.. my 2025 model with RTX5050 has better speakers, better graphics performance and more than sufficient battery life. The main problem is that it has an MSRP above $3000, so the 17in with the dGPU becomes the value proposition. The other thing that is very irritating is that most of the configs on LG's US site are underpowered U processors or some Ryzen ones with quite high MSRP. These models like the 355U are barely faster than a LunarLake, so basically you will be better off to get a 2025 model, LunarLake if you want batterylife, or Arrowlake for performance. The Aerominum is just marketing crap, it looks uglier (ok less of a fingerprint magnet) but there is no serious gain in weight, especially since they have to stick to 77Whr battery and only 2 speakers to save on weight.

Hi! Unfortunately, my hardware ID for the 1070 is DEV_1BE1, and I noticed in the Nvidia settings that I could choose some options for G-Sync. I don't have enough experience with BIOS to freely change and modify it, but I believe it works with other laptops and older versions that have LVDS communication. My GTX 1070 works via DP but not via VGA connector, which leads me to think that the problem is the screen which is from the time when VGA was used. When i connect the external VGA display, it detects but remains black... I have to ask a friend of mine who has a computer store if he has experience with GPU BIOS and also if he has a recent 40-pin laptop screen to try before buying one, risking it not working. The heatsink for the GTX 1070 will be easier because it has all the memories on the top side, different from the 980 which has on both sides. I'm Portuguese and for now i have to enjoy summer, we don't know when Russia will use a bomb under the sea to flood and drown us all. I don't believe Nvidia made a driver for this hardware combination. Note that Nvidia released the last official driver for GTX which greatly increases performance in newer games and dx12. Note that this CPU does not have AVX2. More and more recent games require this. It won't be worth having an RTX in this laptop because some games won't open because of the 3940xm.

Unfortunately no, since non DC screen go through the LVDS port of the gpu and all pascal cards and above (including 1070m) don't have LVDS pins, no vbios can fix that. What it should do is make the 1070m work on DC screen since it wasn't possible before due to eDP placement mismatch in vbios. The 8570w doesn't use integrated graphics and the internal lcd goes directly to the dGPU if I am not mistaken.

If you want, I can tell you what changes to make in RU.EFI so that the settings you apply in RU.EFI won’t reset anymore and will only be reset by clearing the BIOS/CMOS.

@JadeRover so if i understand/read that correctly, that Vbios build makes 1070M work on a Non-DC enabled 8570W? cool news if it does. hadnt stumbled on that post before i got my 980M (d'oh) time to search for a 1070M then 😉 for the next upgrade stage. Until the cable saga is resolved the RTX cards will have to wait.

unfortunately i cant use native cables, still rocking 2x8pin on the PSU side 😄 but this sounds like a nifty lil feature!

Initial impressions of this cable are very positive. It is very soft and pliable and appears to be very high quality. The pin balance is good. I honestly have no room to complain about the original cable that came with the old Toughpower GF3 1650W PSU. It survived a 3090 Ti, 4090 and my 5090 up until yesterday, when it began to demonstrate a 2-3A variance between pins. It is probably still entirely within the safe zone, and well within the Thermal Grizzly default range, but it did suddenly change and I thought it not worth tempting fate. It has served me very well. It has no visible evidence of degradation or having been hot at the GPU end or PSU end. The Corsair cable has a tight range of variance.

I have a Corsair ThermalProtect 12V-2x6 cable being delivered tonight from Amazon. I had ordered one of those overhyped new ASUS cables and cancelled it after reading frequent reports that they're not nearly as good as the price tag suggests. (Twice the cost of the Corsair and no better pin balancing, plus no thermal protection.) Incoming from NewEgg for the B850MPOWER living room PC is a black with walnut wood front Lian Li A3 and a Super Flower Combat SFX FP 1000W 80+ Platinum ATX3.1/PCIe 5.1 PSU. PSU was on sale for $149 so my price match in-store credit on the Z890 Apex made it essentially free. End up getting both items for under $100 out-of-pocket grand total.

I hear you, it’s rough out there. That thread really nails how hollow these platforms feel. I focus on small, real connections instead of chasing likes. Makes a huge difference.

Tried the force speed settings with no change. 0x993 = 03 0x997 = 03 0x99B = 00 again only improved with power cycling the whole AGA once booted to windows. I do like where you are going with the debug. To me there are better dB settings for the PCIe Gen3 lanes assigned at Win10 vs BIOS cold boot. If I can query them somehow via windows and then hard code them for every boot I would be golden.

nope havent tested those yet. ive actually just reseated the cable i was using before (Seasonic) and the current imbalance is gone, crazy how that works. i got 3 our of the 5 ordered backup cables on hand, 2 still on the way. 3 of them Seasonic, 2 from Asus.

Thank you for the suggestions. I set 0x425 = 00 with no changes in poor link status on cold boot. Perhaps it will maintain the link when laptop is suspended overnight which would be a huge improvement. I hate having to set all my programs up yet again every day.

I literally told you why I need an unlocked bios in the post, But I guess if it helps CPU upgrade to I7-8700. I have Nvidia GTX 980M. also, why are people telling me to contact prema despite you saying "prema went obscure years ago"?

Are you using the new Corsair cables with the thermal switch, or have you tested them? The original cable that came with one of my PSUs finally started having current imbalance and I will be replacing it. It's within Thermal Grizzly's default range and only throwing imbalance errors because I set lower thresholds for amp (2.0A) and percentage (25%) variance manually. It doesn't have any errors using the WireView Pro 2 default thresholds (6.0A or 40%).

That is due to a vbios mismatch with the mxm slot regarding eDP placement on the 8570w. What he needs is a custom vbios that has eDP on DP_C position of the card as that's where it is located on the mxm slot of the laptop. Most laptops have eDP on position DP_D of the mxm slot so most vbios have it set to that. He can use this Macrumors vbios for 1070m that has eDP on DP_C (that luckily exists) : https://forums.macrumors.com/threads/imac-2011-maxwell-and-pascal-gpu-upgrade.2300989/page-10?post=30799977#post-30799977 I discovered this when my p3000 worked with the internal display of my zbook 17 g3 (eDP on position DP_C) once I used another "Mac compatible" vbios

Full copper GPU sink would be the way to go. Ive seen some machined ones, but never any OEM. They seem to like the Alu / berylium things we see in our machines, which are a pig to use LMP solder too. A couple of my heatsinks dissasembled at points i didnt want to undo before the point i needed came apart. KBX is working on a 1070m but hes hitting DC issues too, card works but only via DP. That said his H/sink is different to the 980M as chip heights so i dont know if there is another plate available, unless you scour Ali/Google for a 1070 sink from another manafacturer and decouple the GPU. 9xx is best for 8570W / 8770W ( well was in OEM spec) our crazy minds have gone a little bit further than that here

Hi. I bought Lenovo P51 in used conditions and I have not used it at all for long period of time. I bought it online from user, I remember I did power it on just to check if everything is okay, and it was. Yesterday I powered it on, and noticed this vertical line, which made me worried a lot. Will it worsen with time? What would be cause of it?

I’d suggest tightening your RAM timings a bit. Since your RAM is already running at tighter timings than its standard settings, I think CL18 should be easily achievable. This should noticeably reduce memory latency, and you’ll definitely see the difference in benchmarks. Going from CL21 to CL18 could make a significant difference.

0x43F3E One Of: Enable Root Port, VarStoreInfo (VarOffset/VarName): 0x98F, VarStore: 0x1, QuestionId: 0x4C6, Size: 1, Min: 0x0, Max 0x2, Step: 0x0 {05 91 CE 08 D1 08 C6 04 01 00 8F 09 10 10 00 02 00} 0x43F4F One Of Option: Disabled, Value (8 bit): 0x0 {09 07 8F 00 00 00 00} 0x43F56 One Of Option: Enabled, Value (8 bit): 0x1 (default) {09 07 8E 00 30 00 01} 0x43F5D One Of Option: Auto, Value (8 bit): 0x2 {09 07 8D 00 00 00 02} 0x43F64 End One Of {29 02} 0x43F66 One Of: Max Link Speed, VarStoreInfo (VarOffset/VarName): 0x993, VarStore: 0x1, QuestionId: 0x4C7, Size: 1, Min: 0x0, Max 0x3, Step: 0x0 {05 91 D4 08 D5 08 C7 04 01 00 93 09 10 10 00 03 00} 0x43F77 One Of Option: Auto, Value (8 bit): 0x0 (default) {09 07 8D 00 30 00 00} 0x43F7E One Of Option: Gen1, Value (8 bit): 0x1 {09 07 C6 09 00 00 01} 0x43F85 One Of Option: Gen2, Value (8 bit): 0x2 {09 07 C7 09 00 00 02} 0x43F8C One Of Option: Gen3, Value (8 bit): 0x3 {09 07 C8 09 00 00 03} 0x43F93 End One Of {29 02} 0x43F95 Suppress If {0A 82} 0x43F97 QuestionId: 0xE80 equals value 0x0 {12 06 80 0E 00 00} 0x43F9D One Of: Max Link Width, VarStoreInfo (VarOffset/VarName): 0x997, VarStore: 0x1, QuestionId: 0x4C8, Size: 1, Min: 0x0, Max 0x3, Step: 0x0 {05 91 9F 06 A2 06 C8 04 01 00 97 09 10 10 00 03 00} 0x43FAE One Of Option: Auto, Value (8 bit): 0x0 (default) {09 07 8D 00 30 00 00} 0x43FB5 One Of Option: Force X1, Value (8 bit): 0x1 {09 07 A3 06 00 00 01} 0x43FBC One Of Option: Force X2, Value (8 bit): 0x2 {09 07 A4 06 00 00 02} 0x43FC3 One Of Option: Force X4, Value (8 bit): 0x3 {09 07 A5 06 00 00 03} 0x43FCA End One Of {29 02} 0x43FCC One Of: Power Down Unused Lanes, VarStoreInfo (VarOffset/VarName): 0x99B, VarStore: 0x1, QuestionId: 0x4C9, Size: 1, Min: 0x0, Max 0x1, Step: 0x0 {05 91 75 06 76 06 C9 04 01 00 9B 09 10 10 00 01 00} 0x43FDD One Of Option: Disabled, Value (8 bit): 0x0 {09 07 8F 00 00 00 00} 0x43FE4 One Of Option: Auto, Value (8 bit): 0x1 (default) {09 07 8D 00 30 00 01} 0x43FEB End One Of {29 02} We are going to do this now: lock the AGA to PCIe Gen3 and x4 so the link speed and width cannot drop, ensuring it operates at x4 from the beginning. We will also disable Power Down Unused Lanes so the unused PCIe lanes are not powered down. 0x993 = 03 0x997 = 03 0x99B = 00 Apply these changes, then check whether the errors are reduced or not. Let me know the results so we can investigate further. By the way, are you sure the AGA cable is definitely working properly?

My friend, I want us to run a few tests together. Go into RU.EFI and check this option. If it is Enabled, disable it: ACPI D3Cold Support = Disabled Current value: 0x425 = 01 Then do the following several times: Perform a complete shutdown. Power on the laptop with the AGA + RTX 5080 connected. Check the PCIe Error Counters in GPU-Z. Run a demanding game or GPU stress test. Repeat the Cold Boot test several times. If the rate of Bad Boots noticeably decreases, that would be a strong clue that the problem is related to PCIe Power State / Power Sequencing. If nothing changes, we can change the setting back to Enabled.

Regarding the performance testing, I have the 2080 enabled and set dedicated to PhysX. Then a soft disable (in device manager) of the internal GPU. I have it documented my cold boot AGA+5080 in the AGA thread linked at bottom: I would really appreciate some help as I got a second AGA but same issue. So it appears a 5080 when used in AGA issue.

Even with the Intel graphics enabled, the laptop still displays an image. What I mean is: once you’re in Windows, open Task Manager and disable the RTX 2080. Also, run the test using an external monitor at a resolution higher than 1080p. If you test at 1080p, the CPU will become a significant bottleneck. Also, tell me exactly what issue you’re having with the RTX 5080 on your system. Maybe I can think of something that will help us fix the problem.

The Problem: Many older legacy systems (such as Ivy Bridge-generation laptops like the HP EliteBook 8570w) are completely blocked from receiving the modern Windows Secure Boot 2023/2024 certificate updates. Because these older devices have strict NVRAM variable storage limits, Microsoft programmatically flags them with an internal Known Issue block (KI_8 / Temporarily Paused) in the Event Logs, leaving the system silently unpatched or stuck in infinite update retry loops. Manual attempts often hit a brick wall with authentication or access errors due to payload targeting mismatches. What is special about this recipe: Instead of throwing your hands up or risking a system brick, this recipe bypasses the OS-level blocks by safely splitting the signed payloads and manually appending them to their exact cryptographically matched NVRAM variables (db vs dbx), managing the tight variable space perfectly. Broad Adaptability: While this recipe explicitly covers the HP EliteBook 8570w, the same fundamental principles can be adjusted to unblock other legacy hardware facing similar NVRAM or staging limitations. 100% Native PowerShell: Done entirely within standard Windows PowerShell. Zero Bloat: No third-party tools, no sketchy flashing utilities, and no dangerous BIOS modding required. Validated Cleanup: Includes a built-in verification step and a clean sweep script to unmount the EFI partition and wipe temporary footprints when finished. Below is the complete, field-tested step-by-step recipe to get your machine fully patched, verified, and reporting a clean bill of health in the Event Viewer: RECIPE Test Environment: * Hardware - HP EliteBook 8570w * Firmware - 68IAV Ver. F.71 * Operating System: Windows 11 Requirements / prerequisites: * Latest BIOS version (Ver. F.71 for HP EliteBook 8570w) * Boot mode: UEFI Native (without CSM) * Secure Boot supported and enabled * Disable BitLocker * Make sure there's a '%WINDIR%\System32\SecureBootUpdates' directory * Create directory - C:\Temp - temp one, for simplicity * Copy everything from %WINDIR%\System32\SecureBootUpdates to C:\Temp * Start 'Windows PowerShell' as Administrator (IMPORTANT: use this version only!!!) Step 1. Install SplitDbxContent in PowerShell: - Install-Script -Name SplitDbxContent -Force Step 2. Identify your exact HP KEK update file: - Run this command in PowerShell: Get-SecureBootUEFI -Name PK -Decoded - Look for 'SerialNumber' value - Go to https://github.com/microsoft/secureboot_objects/blob/main/PostSignedObjects/KEK/kek_update_map.json and find KEK file by 'SerialNumber' value above. File specified in "KEKUpdate" (ex. HP/KEKUpdate_HP_PK3.bin) - Find that file on the left and download it to C:\Temp Step 3. Update KEK by running the following commands in PowerShell: - PowerShell.exe -ExecutionPolicy Bypass -Command "cd C:\temp; & '$HOME\Documents\PowerShell\Scripts\SplitDbxContent.ps1' C:\temp\<bin_file_from_step_2>" - Set-SecureBootUefi -AppendWrite -Name KEK -ContentFilePath C:\temp\content.bin -SignedFilePath C:\temp\signature.p7 -Time 2010-03-06T19:17:21Z - On success you should see something like: Name Bytes Attributes ---- ----- ---------- ... {...} NON VOLATILE... Step 4. Update DB by running the following commands in PowerShell: - PowerShell.exe -ExecutionPolicy Bypass -Command "cd C:\temp; & '$HOME\Documents\PowerShell\Scripts\SplitDbxContent.ps1' C:\temp\DBUpdate3P2023.bin" - Set-SecureBootUefi -AppendWrite -Name db -ContentFilePath C:\temp\content.bin -SignedFilePath C:\temp\signature.p7 -Time 2010-03-06T19:17:21Z - On success you should see something like: Name Bytes Attributes ---- ----- ---------- ... {...} NON VOLATILE... - PowerShell.exe -ExecutionPolicy Bypass -Command "cd C:\temp; & '$HOME\Documents\PowerShell\Scripts\SplitDbxContent.ps1' $env:windir\System32\SecureBootUpdates\DBUpdate2024.bin" - Set-SecureBootUefi -AppendWrite -Name db -ContentFilePath C:\temp\content.bin -SignedFilePath C:\temp\signature.p7 -Time 2010-03-06T19:17:21Z - On success you should see something like: Name Bytes Attributes ---- ----- ---------- ... {...} NON VOLATILE... *(Note: If DBUpdate2024.bin fails due to NVRAM space or certificate limits, check for a legacy version or proceed directly to the OROM step below).* - PowerShell.exe -ExecutionPolicy Bypass -Command "cd C:\Temp; & '$HOME\Documents\PowerShell\Scripts\SplitDbxContent.ps1' $env:windir\System32\SecureBootUpdates\DBUpdateOROM2023.bin" - Set-SecureBootUEFI -AppendWrite -Name db -ContentFilePath C:\Temp\content.bin -SignedFilePath C:\Temp\signature.p7 -Time 2010-03-06T19:17:21Z - On success you should see something like: Name Bytes Attributes ---- ----- ---------- ... {...} NON VOLATILE... Step 5. Update DBX by running the following commands in PowerShell: - PowerShell.exe -ExecutionPolicy Bypass -Command "cd C:\Temp; & '$HOME\Documents\PowerShell\Scripts\SplitDbxContent.ps1' C:\Temp\dbxupdate.bin" - Set-SecureBootUEFI -AppendWrite -Name dbx -ContentFilePath "C:\Temp\content.bin" -SignedFilePath "C:\Temp\signature.p7" -Time "2010-03-06T19:17:21Z" - On success you should see something like: Name Bytes Attributes ---- ----- ---------- ... {...} NON VOLATILE... - PowerShell.exe -ExecutionPolicy Bypass -Command "cd C:\Temp; & '$HOME\Documents\PowerShell\Scripts\SplitDbxContent.ps1' $env:windir\System32\SecureBootUpdates\DBXUpdateSVN.bin" - Set-SecureBootUEFI -AppendWrite -Name dbx -ContentFilePath C:\Temp\content.bin -SignedFilePath C:\Temp\signature.p7 -Time 2010-03-06T19:17:21Z Step 6. Verify updates by running the following commands in PowerShell (each command should return 'True'): - ([System.Text.Encoding]::ASCII.GetString((Get-SecureBootUEFI db).bytes) -match 'Windows UEFI CA 2023') - ([System.Text.Encoding]::ASCII.GetString((Get-SecureBootUEFI kek).bytes) -match 'Microsoft Corporation KEK 2K CA 2023') - ([System.Text.Encoding]::ASCII.GetString((Get-SecureBootUEFI pk).bytes) -match 'Microsoft RSA Third Party PCA 2023') - ([System.Text.Encoding]::ASCII.GetString((Get-SecureBootUEFI pk).bytes) -match 'Hewlett-Packard') Step 7. Verify you have a boot loader copy properly signed with new certificates, by running the following command in PowerShell: - [System.Security.Cryptography.X509Certificates.X509Certificate2]::CreateFromSignedFile("$env:windir\Boot\EFI_EX\bootmgfw_EX.efi").Issuer - You should get back: 'CN=Windows UEFI CA 2023, O=Microsoft Corporation, C=US' Step 8. Update bootloader by running the following commands in PowerShell: - mountvol S: /S - reg save HKLM\BCD00000000 S:\EFI\Microsoft\Boot\BCD.bak /y - Copy-Item "S:\EFI\Microsoft\Boot\bootmgfw.efi" "S:\EFI\Microsoft\Boot\bootmgfw.efi.bak" -Force - Copy-Item "$env:windir\Boot\EFI_EX\bootmgfw_EX.efi" "S:\EFI\Microsoft\Boot\bootmgfw.efi" -Force - mountvol S: /D Step 9. Reboot your machine Step 10. Update the Registry and certificate update task state by running the following commands in PowerShell: - Set-ItemProperty -Path "HKLM:\SYSTEM\CurrentControlSet\Control\SecureBoot" -Name "AvailableUpdates" -Value 16384 -Type DWord - Set-ItemProperty -Path "HKLM:\SYSTEM\CurrentControlSet\Control\SecureBoot\Servicing" -Name "UEFICA2023Status" -Value "Updated" -Type String - Remove-ItemProperty -Path "HKLM:\SYSTEM\CurrentControlSet\Control\SecureBoot\Servicing" -Name "UEFICA2023Error" -ErrorAction SilentlyContinue - Remove-ItemProperty -Path "HKLM:\SYSTEM\CurrentControlSet\Control\SecureBoot\Servicing" -Name "UEFICA2023ErrorEvent" -ErrorAction SilentlyContinue - Stop-ScheduledTask -TaskName "\Microsoft\Windows\PI\Secure-Boot-Update" -ErrorAction SilentlyContinue - Disable-ScheduledTask -TaskName "\Microsoft\Windows\PI\Secure-Boot-Update" -ErrorAction SilentlyContinue - Enable-ScheduledTask -TaskName "\Microsoft\Windows\PI\Secure-Boot-Update" -ErrorAction SilentlyContinue Step 11. Reboot your machine Step 12. If all is successful you'll have the following Information log from TPM-WMI in Event Log: "This device has updated Secure Boot CA/keys. This device signature information is included here. DeviceAttributes: FirmwareManufacturer:Hewlett-Packard;FirmwareVersion:68IAV Ver. F.71;OEMModelBaseBoard:176B;OEMManufacturerName:Hewlett-Packard;OSArchitecture:amd64; BucketId: 54e624d808a5b9fbd4f3c1d0a184eeee8a1d0d5cc7238ce8403ed6dd6053bf29 BucketConfidenceLevel: Temporarily Paused UpdateType: Windows UEFI CA 2023 (DB), Option ROM CA 2023 (DB), 3P UEFI CA 2023 (DB), KEK 2023, Boot Manager (2023) For more information, please see https://go.microsoft.com/fwlink/?linkid=2301018." Step 13. Cleanup the files by running the following commands in PowerShell: - mountvol S: /S - del S:\EFI\Microsoft\Boot\BCD.bak - del S:\EFI\Microsoft\Boot\bootmgfw.efi.bak - mountvol S: /D - rmdir C:\Temp -Recurse -Force Note: The recipe is also attached to this message as a TXT file. HP EliteBook 8570W SecureBoot Certificates Update Steps.txt

I tried disabling them before and did not notice a change (at least in Steel Nomad) bechmarks. I am having an issue where the 5080 in the AGA does not sync up properly. I have to disable the card and unplug the AGA and plug back in once booted to windows. So i need 2080 enabled so the laptop screen works.

thats how i roll 😄 nah man, 100% hands off. WVPII totally provides peace of mind, let it run in the background and do its thing until an alarm gets triggered 🙂