Feeling stuck with the ASUS 970 Pro gaming setup?
Feeling stuck with the ASUS 970 Pro gaming setup?
Hey everyone, I've checked all the guides and watched every tutorial available. I've been talking on some forums too and I'm pretty sure they're full of mistakes—or maybe I'm just seeing things. Do you have any advice on getting this rig to run faster? I've tried almost everything over the past ten years, but nothing seems to work. I'm almost ready to switch to a new Dell. Could someone please assist me?
Current specs:
- 970 Pro Gaming Aura FX 6350 chip
- Hyper Evo 212 cooler
- 212GB RAM (2 sticks)
- HyperX Savage 2 500GB HDD in RAID array
- Radeon R9 270x GPU
- 6.4mm fans, 3 intake, 3 exhaust fans
- Front, bottom, and middle intakes
- One fan on the back side of the socket
- Rear fans: top, top rear, top middle
- Power supply: 1500W Rosewill Blackhawk 117 case
- Running Windows 10
- Monitoring with Prime95 and aida64
- Stress testing via hwinfo
Let me know if you need anything else!
Hey there! Welcome to the discussion. I really appreciate the FX series of chips. I've pushed several of them overclocked, even the 6300. There are a few points to cover here, nothing too big but they’ll guide you correctly. Let's start with the basics: FX chips have cooling and motherboard as their main constraints. Sadly, your cooling setup isn’t much better than what’s on the board, but a 990FX would be ideal. We can make do with what you have, that’s the plan. On the Hyper 212, I suggest adding some powerful fans in push-pull mode. Airflow is key, and ambient temperature matters a lot. Keep it as low as possible since cooling performance is crucial. The chipset isn’t voltage-limited, so you can run high LLC and v-core settings. Aim for temps around 50-60°C under load. With FX chips, keeping things cool becomes tricky once you hit voltages above 1.45V-1.500V. It’s pretty straightforward—just target around 5GHz to 1.500V, with a bit of margin. These values are just estimates; aim for under 60°C load for the best results. They appreciate higher voltage and better cooling, which means more airflow.
I value your response. You consistently followed the advice you received, achieving similar outcomes—primarily excessive heat and insufficient performance gains. With my 120mm and 140mm fans positioned roughly four inches apart, I’ve observed only a marginal improvement, around a 1% boost in cooler efficiency. I’m consistently reaching approximately 4.5Ghz at about 1.5V, maintaining temperatures near 77°C without any throttling. I’ve reviewed videos and discussions where others reported significantly higher speeds with identical setups, yet my results remain stable but underwhelming. Despite this, I’m open to further tweaks: increasing the multiplier to 21.5 for a 4300Mhz clock, aligning the D.O.C.P. profile with my RAM speed, and turning on the high power setting as you recommended. After adjusting these settings, the system passed a four-hour stability test. While the gains aren’t dramatic, they’re functional. I’m hoping for a few more frames per second in ESO or a slightly quicker boot time—though web pages still load a bit slower and game performance isn’t noticeably better. In desperation, I considered pushing the multiplier higher, but I’m cautious about overheating. Regarding cooling upgrades, you mentioned you’re limited by budget; would a mid-range 990fx chipset be a practical choice for someone on a tighter budget? I’m willing to invest up to $100 next month if it means better results. Working with what I have is the most realistic path forward. Adding more heatsinks or liquid cooling might help, but I’d prefer something affordable. For someone in your situation, a solid 75°C max temperature seems reasonable, and I’d avoid pushing beyond that unless necessary. I appreciate your guidance and will keep sharing updates if you’re open to it.
The voltage looks elevated, but high temperatures increase leakage issues. Transistors require a specific current to function properly. Two possible outcomes exist. Either a shutdown occurs—also known as a low-voltage hang—where enough transistors cease operation and the screen becomes unresponsive, preventing the caps lock from working. The issue stems from needing more cooling than the board provides. I’ve shared several designs using M5A78 L MX Plus boards with reduced core counts to maintain 2.0V while still handling current demands. Due to high temperatures at the socket, performance suffers significantly. For better results, consider using a cooler rather than just upgrading the board. Your setup is unique; others may have more favorable ambient conditions. Comparing temperatures relative to your system is key. You might try adding a small fan near the VRM area on the left of the CPU to help lower temperatures. That’s the primary goal—reduce heat. You’re running too hot. I’m sorry for the update.
You captured the key points well. The temperatures seem linked together, and you're adjusting fans to match. Adding new fans could help, especially if you mount them externally or add vent holes in the case. Your current setup with a 140 mm fan on the motherboard and CPU socket is a good start. Considering a water block might not be enough for extra cooling needs.
Updated: I shifted my case and earned an extra 10C! Excited! Now here are the questions—research indicates that once temperatures reach or surpass 72°C, issues arise. Studies also confirm 90°C triggers thermal throttling on my CPU. AMD states the safe upper limit is 61°C. I understand heat damages components and reduces their lifespan, but how much does it actually affect it? If I push to 63 or 65°C during overclocking for any brief period, what impact will that have? The same applies at 75°C. My core temperature readings show a maximum of 90°C, but which one is accurate? Currently, I'm testing at 4.5Ghz with a temperature max of 62.4°C. I plan to lower the CPU voltage until core errors appear.
Hey there! Your last message got lost in the chat. Would you like me to like your post or quote me so I can notify you?
The TCase Max is around 61°C. With stock cooler and default settings, the CPU could run at about 72°C under sustained loads. For best performance, aim for a max of 90°C with continuous stress.
At that temperature, VRM cooling becomes crucial—heat spreads through the board and affects the CPU. Many users add a fan on the back of the motherboard to help lower temps. Some even modify the case or install extra cooling if space allows. A slight drop in VRM temps can really improve stability.
If you knew about this earlier, I’d have chosen a model like the H100i with more cooling options. Otherwise, active VRM cooling is still key for high performance.
I've been thinking about getting an H100i for some time. Swapped in better case fans, saving more than $40. The thermal performance drops noticeably above 4.5Ghz. At that point, I might consider it a good deal. Since the system feels outdated, I don't expect it to deliver extreme results anymore. If I push it to 63 or 65°C during an overclock, just how much will it affect its lifespan? Also, what's the meaning of 'delidding'? Are you someone who overclocks competitively? Do you have any videos to share? I assume this isn't something I'd tackle on my own machine.
Many useful questions here. I don’t know how to measure a CPU’s lifespan from an overclock. No one really can. But as I mentioned from the beginning, your target should be under 60°C. Generally, the higher the clock, the cooler you want it to be. For this CPU, I’d lower the voltage perhaps to 1.400-1.45V and try to hit around 4.6 or even 4.7GHz. Yes, boosting speeds usually brings temps down without extra voltage. Keep an eye on that. The rest will give better answers when I’m at my desk with a proper keyboard. Phone typing isn’t my style lol.
The process involves taking off the copper sheet that's bonded, revealing all the chips and solder connections on the silicon or CPU. In older Intel models such as the 8700K, the chips aren't soldered, making it straightforward to simply cut the adhesive and "remove" the component. Swapping out the original thermal paste beneath the IHS plate significantly lowered temperatures. But with soldered chips, it becomes less useful since solder typically offers better heat transfer than liquid metal, even if they're similar. I've worked with nearly every processor I've owned since around 2009, regardless of whether it was soldered or not. I'm definitely into it, though I don't consider myself a professional overclocker like Der8auer. Instead, I enjoy competing in events. Right now I'm actively participating and having a great time. It's more of a hobby than a career. Modifications bring me joy. I recently upgraded my first graphics card by increasing the v-core, which performed well on a 7900GS. Besides that, I've made a few BIOS tweaks on VGAs. It can be risky to brick a good card if you're not careful. I don't have many videos—just a couple—and those are mostly benchmark results from HWBot. My records go up to the submissions I've filed there, which helps me track my progress and compare scores later. These entries are stored on HWBot, where I have around 700 submissions. I know some people submit thousands more, so my screenshots aren't huge. Regarding your last point, this isn't something you should try on a regular PC—just because you've overclocked for years doesn't guarantee success in the competition. If you're interested in trying, I can send you a link to a friend who set up the event at their site. Just remember, it's not about advertising here; it's about sharing knowledge. And don't worry, I'm not pushing our site too hard since there are many experienced users on HWBot.