Issues with overclocking bundles
Issues with overclocking bundles
I received an overclocking package from Overclockers Co UK, featuring an Asus ROG Strix Z490.
Gaming, 10900k and DDR-3600 RAM. I’d share their forum posts, but the community there is quite toxic. The bundle includes 5GHz SSE and 4.9GHz AVX speeds. Occasionally it emits a strong, burning odor—possibly due to heat or a malfunction. Additionally, I experience sudden bursts of extremely hot air escaping from the top. This is typical behavior.
I also have an excellent water cooling system that maintains temperatures between 350-400 watts. The airflow around the VRMs is substantial.
There have been some problems with my graphics card, one linked to overclocking (data corruption and crashes). Running Prime95 causes the VRM temperatures to rise to over 110°C (I haven’t run it long enough to risk damage). 280 watts escalates to around 330 watts, sometimes reaching 350. With strong airflow, I’ve lowered the voltage from 1.385V to 1.28V, which leads to severe VRM heating. The hot air blast from the top is noticeable. This issue occurs with 7 fans. Without adequate cooling, the board risks overheating. Most stress tests pass, but only Prime95 reliably shows stability. After a short period, it becomes apparent. I haven’t used Prime95 for extended periods to avoid potential damage, but customer support indicated difficulty in resolving warranty claims if the bundle fails. If the product malfunctions, they may claim I’m overclocking it excessively. They won’t upgrade the motherboard to a superior model after stating this board is prone to such issues even with replacement. That means I’ll need to handle all repairs myself once I get a better board and the bundle arrives.
I already have a much improved motherboard—Asus ROG Maximus XIII Hero.
Do any of you have experience with the Asus ROG Strix Z490?
Gaming motherboard? It seems not built for heavy overclocking. The VRMs handle up to 150°C, but under heavy loads they behave like a BBQ. In a compact form factor, 110°C+ VRMs suggests something is amiss.
Increase core voltage while lowering load line calibration. The LLC is probably forcing the VRMs to operate at their maximum settings. AVX should be reduced further if it's not being used much. Gaming and similar tasks won't require it as much. Intel has already removed it from their upcoming 12th generation chips.
I attempted to lower the LLC and boost the vcore. When running prime95, the VRMs turned into a BBQ. They reached at least 109c, which caused me to halt the stress test. If they had stayed at 100c or below, I wouldn’t have been concerned. The real issue was hitting 110c, indicating excessive heat.
From reviews, it seems that even at stock settings, the CPU would reach 300 watts under heavy load. Overclocking would require a significant power draw. I could push the overclock to 5.2GHz, but prime95 generated intense heat for the VRMs. The maximum wattage for the CPU only slightly increased—from 330W at 4.9GHz to 350W at 5.1GHz AVX. At 4.9GHz with AVX and prime95 at FM3, there was no further improvement possible, except stock performance.
I envisioned a compact case with 110c VRMs and one failing fan. Hopefully, a better overclocking board could handle it. In one review for the ASUS ROG Strix Z490, they mentioned it could manage prime95 with low VRM temperatures. That’s exactly what I’m after for a small form factor.
In short, I can’t blame overclockers—the board still functions and the overclock works. If this board were in a compact case and I ran prime95, it would likely exceed 110c on the VRMs. Eventually, that could cause problems and damage the component.
I’m probably getting strange odors due to the heat. It feels like a furnace when I run prime95. So I’d probably just test in OCCT. Hours of OCCT don’t usually reveal unstable cache overclocks. The same applies to aida64 and Intel Extreme Tuning Utility—it works fine, but prime95 catches it within an hour. That’s why prime95 is the best option.
Just wondering if anyone has experienced this with the Asus ROG Strix Z490.
We all understood that the 10900k tends to overheat when you try to keep it fully overclocked, which is why some $400 Z490/Z590 boards are available. You might want to consider a water-cooled solution if you really aim to push the CPU further.
It's possible your chip is leaking and needs more power. My i7-4770k had similar issues—it could reach 4.5Ghz, but its power consumption was extremely high, so I reduced it to 4.3Ghz.
Even my 10900F operates at 192W at its stock 4.6Ghz boost (power limits removed). Intel specifies a 65W TDP for this chip.
The overclock was a package from enthusiasts. They crafted it and verified it using prime95. All power caps were stripped away, with a 1.385 volts Vcore and an LLC7 that essentially allowed the VRMs to operate freely. Then they chose a rather weak 10900k chip. With the compact motherboard size and CPU, they ought to have selected a processor with the lowest feasible Vcore for the target frequency, paired with a component that consumes less power.
Instead, they opted for the weak CPU, SP 69, and then selected a chip requiring 330 watts at the overclocked settings where power limits were disabled. They then disabled the board’s power cap, set the LLC to 7, and adjusted the Vcore to 1.385 volts. This represents the most aggressive overclock possible.
The outcome was a VRM temperature reaching 110°C in prime95. If the six fans were removed and only one remained, the VRMs would overheat rapidly.
I discovered that lowering the Vcore significantly improved results—potentially down to 1.25 volts with an LLC7 of 7. Adjusting the LLC further could yield 330 watts without exceeding the limits.
In a small form factor setup with just one fan, I realized the build would likely fail under high load. In prime95 I achieved 110°C VRM temps while using overkill airflow. After shutting it down, I noticed the support team avoided addressing the issue, directing customers to sales for assistance. They seemed unwilling to help and suggested future upgrades wouldn’t be permitted.
The core problem appears to be their overclocking configuration—setting frequencies, power caps, and components without proper optimization. It’s not a matter of hardware limitations but rather an improper setup.
It’s not stable to simply adjust settings and browse online without overheating the VRMs. Eventually, I plan to use the 10900k for video encoding tasks in handbrake and upload content to YouTube. Occasionally, such intensive work can strain even robust systems.
The 10900k has a TDP of 125 watts, while the 10900F is 65 watts. With the stock model, users report stable performance at around 4.9GHz with TVB, dropping to 4.8GHz without it after 70°C. This aligns closely with OCUK’s typical overclock results.
My experience in a compact case with only one fan suggests the build would fail under stress. In prime95 I reached 330W VRM temps, but with adequate cooling, temperatures stayed manageable. After shutting it down, I found the support team unresponsive, directing me to sales for further help. They seemed reluctant to acknowledge the issue and hinted that replacements wouldn’t be allowed.
The real concern lies in their overclocking strategy. It was their settings, frequencies, and motherboard choice that prevented full power unlocking and forced them to rely on maximum performance at the cost of stability.
I’m not satisfied with the current approach. I intend to upgrade to a better board, fine-tuning the same overclock, ensuring LLC isn’t 7 and Vcore stays at 1.385 volts.
It’s not a stable setup if you can simply tweak settings and search online without worrying about VRM overheating. Eventually, I’ll use the 10900k for encoding tasks in handbrake and upload content to YouTube. Sometimes that workload pushes even powerful systems to their limits.
The 10900k consumes 125 watts, while the 10900F uses 65 watts. Running it at stock levels with proper cooling achieves a solid core speed of 4.9GHz with TVB, dropping to 4.8GHz without it after 70°C. This is essentially what OCUK’s overclocking aims for.
My VRM temperatures would rise further in a compact case, especially under high loads. If I ran this setup with prime95, the VRMs could exceed safe limits. I’m not comfortable with that outcome.
I need to find a better board and replicate the same overclock carefully. I’ll keep power draw in check but avoid setting the LLC to 7 or the Vcore to 1.385 volts.
It’s not a reliable stable overclock if you’re only adjusting settings and browsing without concern for heat. Eventually, I’m considering upgrading to a different model so I can maintain performance without risking component damage.
So I handle the fix myself. I begin with a 230 watt cap. All cores run at 4.9GHz. The vcore appears elevated, but that's mainly due to standard boosting. For the multi-core section in the CPU-Z vcore setting, it's around 0.8 volts. I'm not sure if this reflects the actual value.
The LLC circuitry adapts to 5 volts.
Prime95 is working well. VRM temperatures are stable at 61°C and 63°C respectively. At 220 watts, the frequency drops in Prime95, preventing VRMs from overheating. This approach lets you control CPU power usage during Prime95 runs—raising power if VRMs perform well, lowering it if they struggle.
I fine-tuned my overclock to 5GHz with SSE and AVX, staying within a 270-watt limit. Vcore set at 1.330 volts for LLC4. With Prime95 stable for about an hour, I increased the vcore from 1.3V to 1.33 to resolve single-thread errors. VRM temperatures remain below 80°C.
My current overclocking configuration: 5GHz SSE, 4.9GHz AVX, 270 watts max. Vcore at 1.330 volts for LLC4. BIOS OCUK settings were adjusted to stop 110°C VRMs. I plan to apply these once on a more capable motherboard, aiming to boost power limits while keeping VRM temps low.
For the Asus ROG STRIX Z490-G GAMING (WI-FI), the maximum CPU overclock reached 5.2GHz with SSE and 5.1GHz AVX. LLC5 operates at 1.33 volts under load, and Vcore was reduced from 1.3V to 1.33 to fix single-thread issues. VRM temperatures stayed under 80°C.
I’ll keep all adjustments in mind for future upgrades.