I don't have the latest information on z790. Could you check the official sources or recent announcements for updates?
I don't have the latest information on z790. Could you check the official sources or recent announcements for updates?
The 14900K isn't the part that requires updates in the newest BIOS, as it's essentially just a rebranded 13900KS. This is typical for the 14900K. What concerns? It's likely unrelated to the BIOS version at all. If temperature problems were mentioned earlier, those are normal for this chip.
On a few cores I noticed 10-degree spikes along with throttling. Most instances were normal, but some consistently reached 100-103. I’m hoping this was due to bent components. It’s useful to know about BIOS settings. I also experienced many window hardware and PCIe bus errors. It seemed like the issue might be related to the CPU or chipset. I remember needing internet access to install Windows 11, but I didn’t have Wi-Fi or the necessary drivers. Still, it worked out, so I added the drivers from Windows. This might have played a role in the problem. I’m still trying to figure out what’s causing the issues.
It seems like the issue might stem from hardware problems rather than the processor itself. A 10C core to core delta is common in recent chips, especially during idle. The warning in Windows 11 usually means internet access is needed and isn't strictly required. You can likely work around it by testing different chip revisions or running stress tests. If the problem persists, a faulty motherboard or RAM might be the cause, so consider trying another revision or checking your components.
I'm covered under warranty for both units. You have roughly 10 more days. The board has one ATX12 MB power port. Should I use both unless you're overclocking, or just aim for a 6GHz AI boost? Concerned about the board's 7200MHz design with two 24GB Trident Z5 chips—faster RAM can cause more instability. Adjusting speeds might help. I'm expecting a CPU contact plate tomorrow to secure the new chip evenly. If you need to swap the board before removing the bracket, consider doing it now.
The ASUS name for the EPS 8 pin connector is correct. You’re connecting both pins mainly because the 14900K draws a lot of power, even though that’s not the main reason for your problem. This seems more likely to be the cause. 7200MT/s may not work consistently across all CPUs, and most manufacturers focus on the top 1% for maximum frequency validation. It should function on most 14th generation chips, though it’s possible this is what’s affecting you. The 24GB DIMMs need specific BIOS updates, but as long as your system isn’t several months old, it should be supported. You’ll likely need a BIOS revision for those DIMMs. Disable XMP and stick to the stock 5600 CL46 until you identify the exact issue. Once you know what’s wrong, you can boost speed safely. Running it at JEDEC is advisable to rule out speed-related problems. These errors, though uncommon, can still occur.
The RAM should remain stable at stock clock speed, though it might show a different reading. The voltage specs are around 1.45, not 1.35 as mentioned. I recall seeing it at 3200 in hwinfo. Regarding the errors, Event 17 points to PCI bus channel 5, possibly linked to undervolting. Throttling is common when hitting 103°C, and adjusting the thermal settings to 100 seems reasonable.
It should perform well at that velocity. Odd that it displays 3200 instead of 2800MHz or 3600MHz—those numbers convert to different throughput rates, but not 3200 because G.Skill doesn’t bundle several XMP profiles on their memory modules and 6400MHz isn’t officially recognized as a JEDEC speed standard.
I'm more interested in the actual voltage readings rather than the voltages you mentioned. Running the CPU above 1.2-1.25V under load is uncommon at stock levels. At idle, 1.45V can provide a bit of extra boost for single cores, but it should decrease noticeably when temperatures rise. Voltage specs vary between boards, and my Unify-X runs notably lower than others I've tested in Cinebench—230W versus over 300W on the B660M-A Pro. ASUS generally uses higher stock voltages, leading to more heat from the start. I’d likely lower the thermal throttle setting to around 103C for better long-term performance. Undervolting is manageable; I’ve tried several methods and would probably go ahead if I confirmed any system issues first.
It seems the readings might reflect a maximum voltage in hwinfo, but actual performance under load differs. The issues are quite complex—I need to monitor them closely while they run. As for the errors, tracing them through the event viewer and using hwinfo logs can help pinpoint the source, though keeping drivers updated makes diagnosis harder. I’m considering a fresh installation and separating the M.2 and GPU into different channels to see if it improves things.