Have I got an incredibly good CPU chip (i5 4670k @4.6GHz 55C)?

Kobiat

Reputable
Feb 16, 2014
16
0
4,510
I recently got a new CPU from a friend, and I decided to overclock it. I put the voltage on 1.25V and the clock speed at 4.6GHz. When I run AIDA64 stress test for 1 hour the temps don't go higher then 55C and I only have a cheap cooler (cryorig h7) and a B85 motherboard (Biostar Hi-Fi B85S3).

This is my first time overclocking, are these normal temps?
Is it worth it pushing it to 4.8 GHz and 1.30V or will this decrease the lifespan to much?
 
The Cryorig H7 is a very good air cooler. Don't overlook the cheaper coolers as bad coolers.

Remember that temps (and voltage) go up exponentially as you approach the brick wall. It might only be 1.25v at 4.6ghz, but it might be 1.35v for 4.7ghz stable. This, combined with the fact you've got a dirt cheap motherboard, means you're getting closer and closer to creating a fire hazard in your own home. B85 boards were designed to be cheap in bulk and run with underpowered CPU's, not for overclocking, let alone overvolting.

To actually answer your question, yes your temps are well within spec. To be precise, your CPU temps are well within spec. It's winter in the northern hemisphere, so your temps might be lower than average because of it. I fear your motherboard can't take much more without giving up and potentially roasting your whole system with it. Be safe and get a better motherboard and possibly power supply before you start trying to push limits.
 
I've been using my pc with my i5 4670k overclocked @4.6GHz, no problems so far. CPU doesn't go higher than 50 C (idle 30C) and my motherboard temps stay at an avg 32 C.

I don't get why everyone is saying that 'you are gonna roast your system',...

I've been using my pc for 1 month, 8 hours a day and no problems.
 


The problem lies not with what you can see, but with what you haven't noticed/seen yet.

We aren't talking about your CPU roasting. We're talking about your motherboard mosfets, capacitors, and maybe even the power supply itself. Motherboard temps don't always tell when a poor quality cap is going to blow. EVGA got it lucky when bad caps caused high temps before things exploded. Beyond that, none of us know what power supply you're using, so you're also risking roasting that power supply as well. These things are tested to within a specific load and use case scenario. You're pushing beyond what's recommended and tested.

Biostar isn't known for outstanding anything, let alone power delivery, and no B85 motherboard out there has power delivery components beefy enough to be appropriate overclockers. They're designed to be cheap and run lower-power CPU's to outfit thousands upon thousands of corporate network work pc's and hospitals and those crappy prebuilts.

I've now explained this as properly as I can. If you proceed, remember that no one but yourself is liable for the potential damages. You've long since voided all warranties. Coming from someone who's actually lit a motherboard on fire by overclocking, you've been formally warned.
 


Despite all the negative advice, I continued to use my pc at the same clockspeed for the past year. I kept an eye on the temps regularly and saw no spikes. Gaming performance has been consistent and I never had any weird crashes.

So after a year of using it without creating a 'fire hazard', I'm so curious why it runs so well. Any ideas?
 
Is it worth it pushing it to 4.8 GHz and 1.30V or will this decrease the lifespan to much?

What do you get out of the deal? You OC your CPU and it returns________. If it's bumped up another couple hundred megahertz will that boost justify the extra heat or will it compromise the stability? 200MHz the difference between life and death? Not really.

What about its foundation(aka motherboard), which is but one part of the cooling(more on that later on)? Do you have the necessary cooling and power features necessary to support a bigger OC? Doubtful. You have 1 4pin EPS connector. Your board also lacks the passive coolers where the rubber meets the road(VRM's chokes, mosfets and other things I've not a clue about).

Look to the left of the CPU socket.
b20130611.jpg


Now my board
91Cw0igi5OL._SL1500_.jpg


It has the passive cooler and two CPU connectors. I have a little better cooling and power. On paper that means I should have better OC results but then silicon walks in and muddies the water. Each chip is microscopically different and therefore will OC differently and heat and power will also slightly vary. In short?

I wouldn't push your board any further.

If you want you can run this http://www.majorgeeks.com/files/details/intelburntest.html and see what kind of temps you have. I think that 55C ceiling will disappear. Change the Stress Level to maximum, Post a screenshot of your result.

Temps.jpg


and after the test was done
Intel_Test_completed.jpg
 


It could be your case. Ambient temperature. Fan configuration. CPU cooler. Motherboard. Other reasons I'll think of after coffee.

There's nothing negative about the advice you received. It's wisdom gained through experience. We know what boards are designed for. Our advice has to be in line with recommended use. AND there's an exception to every rule. Maybe you got lucky. A full system parts list and OC voltage would help but not here. This is Kobiat's thread.