Constant application crashes with my i7 13700k Raptor Lake

[Edafet]

I’m experiencing ongoing and worsening instability with my Intel Core i7-13700K processor. My system crashes during nearly every gaming session, regardless of system load, temperature, or background tasks. The crashes happen across multiple titles and game engines, sometimes within minutes of launching. System temperatures appear to be within normal operating ranges.

System Details

• Processor: Intel Core i7-13700K (24 cores, ~3.4GHz base)

• Motherboard: MSI PRO Z690-A DDR5 (MS-7D96)

• BIOS: A.C0 (UEFI)

• Memory: 64GB DDR5

• GPU: NVIDIA GeForce RTX 4080

• OS: Windows 11 Pro 64-bit (Build 26100)

• Monitor: Acer XV272U V @ 2560x1440 170Hz via DisplayPort

• DirectX: 12, Feature Level 12_2

• Power Supply: Corsair RMx SHIFT SERIES 1000 Watt

Troubleshooting Performed

• Ran Intel Processor Diagnostic Tool (passed)

• Ran MemTest86 (no errors)

• Reinstalled GPU drivers (clean install)

• Verified temperatures under load (normal range)

• BIOS and chipset drivers fully updated

• Reset BIOS to default (no overclock or undervolt)

• Checked Event Viewer (frequent application crashes with no clear hardware fault)

• Crashes occur regardless of GPU load, RAM usage, or thermals

DxDiag Snapshot (summary)

• Time: 6/28/2025, 05:18:16

• Machine Name: DESKTOP-2IROAA4

• Machine ID: {4691A5EE-A246-4095-B76D-50E05B8564D5}

• Display Driver Version: 32.0.15.7680

• GPU Driver Date: 6/11/2025

• DxDiag Version: 10.00.26100.4202 (64-bit Unicode)

At this point, I’m trying to determine if this is a CPU defect or another root cause. I would appreciate your guidance on further steps, stress testing, or RMA options.

Let me know if you need crash logs or additional diagnostic info. I’m getting close to replacing the CPU entirely, as this has been a frustrating experience.

[eanastasov]

Hey! I had almost the exact same issue as you — constant game crashes (some BSODs, some app-only), no thermal issues, passed Intel tests, MemTest, etc. Nothing made sense… until I dug into the core ratios.

Turns out the cause was auto CPU ratio settings letting a few favored P-cores hit x60, which pushed voltage/VRMs too hard during boosts — even with no overclocking.

Your i7-13700K, just like my i9-14900KF, has a few cores that can hit super-high boost ratios (e.g., 57x, 58x) — and MSI motherboards often let those go wild on AUTO. That can cause silent instability, just like you're seeing.

Here's What You Can Try in the BIOS:

1. Look for:

   • CPU Ratio Mode → Set to “Fixed Mode”

   • All Core Ratio or P-Core Ratio → Set all to 56 or 57 (lower than max fused, just make them all same ratio)

2. Alternatively, you can also try:

   • CPU Lite Load → Try Mode 6 or 7 (less aggressive voltage)

   • CPU Flex Ratio Override → Enable, then set max ratio manually

But I think fixing the ratios will be enough.

---

Why It Helps:

• Reduces momentary power spikes

• Keeps VRMs from getting overloaded

• Still gives you 95% performance, with full stability

Since you already passed all hardware diagnostics and temps are good, I strongly recommend trying this before replacing the CPU. It solved all my problems — I've had zero crashes in 4 weeks now, even with hours of heavy gaming, photo/video editing, Cinebench, etc.

Let me know if this helped!

Good luck! 

[Edafet]

I can try setting the CPU multipliers to 49, as I’m currently unable to change them beyond 54. That said, I have some concerns:

• Would manually reducing the multiplier be considered a form of overclocking or modification that could void my warranty?

• Would this change significantly reduce the expected performance of the processor?

When I purchased this CPU, I expected the performance advertised by Intel. If achieving stable operation now requires altering settings that might void the warranty or degrade performance, I would prefer to initiate an RMA process instead.

I’d appreciate any clarification or guidance on how best to proceed.

Thank you.

[eanastasov]

"Would manually reducing the multiplier be considered a form of overclocking or modification that could void my warranty?"

No, reducing the CPU multiplier is not overclocking, and will not void the warranty. In fact:

• Lowering the multiplier is a form of underclocking, which is perfectly safe and often recommended for stability or thermal concerns.

• Intel’s warranty policy is voided only if you manually increase voltage or multipliers above spec to overclock the CPU.

So lowering from 54x to 53x or even 52x is totally fine and warranty-safe.

---

"Would this change significantly reduce the expected performance of the processor?"

Technically yes, but in practice, the impact is very small.

• In my case, with my i9-14900kf, Intel’s advertised 6.0 GHz is for 2 P-cores only, under ideal thermal and power conditions.

• The majority of workloads never use just those 2 cores at max boost — and heavy tasks usually downscale anyway due to thermal limits.

• In Cinebench, for example, the difference between 6.0 GHz and 5.7 GHz is often just 2–4% performance — barely noticeable.

• Stability, responsiveness, and reliability matter way more in day-to-day use than a theoretical 300 MHz gain.

So you keep ~96–98% of performance and no more weeks or months of frustration.

---

If you still prefers RMA.

I totally understand — but here's the catch:

Intel won’t always accept an RMA unless the CPU:

• Fails consistently under default settings, on multiple boards.

• Wasn’t manually overclocked (so reducing multiplier is safe).

• Shows thermal or structural defects (e.g., random crashes at stock clocks even on high-end motherboards).
  
  

So unless you demonstrate that even at lower clocks the CPU still crashes, Intel may reject the RMA, and you'll just waste more time. But also, if we pay for top quality we should have top quality, yes?

Edit: Just to clarify something, you do have 2 P-cores running at higher multiplier then the rest yes? If so, lower just those 2 to match the rest, nothing more and tell me how it goes.