Overclocking Escapades: socket LGA1155 Intel Core i5 2400

[Silver_Pharaoh]

No, the server board had no overclocking options whatsoever. I couldn't even change voltage on that thing.

The overclocking and undervolting was done on my Asus P8Z77-V Pro.

The server board this came with was just what this processor was originally run on (for the rest of its life up to now).

I thought you were still using the server board.

 

[wingman99]

The CPU's are tested at the factory for optimum conditions like cold and heat and stress.

Can you elaborate on that? As written, I have no idea what you mean. At all.

What is done though is testing a cpu to fit within speed, voltage, and power envelopes. Not sure where 'heat and cold' come into play...

I have seen a videos of the process packaging and they test for heat, stress and selecting voltage. I cant remember if they test for cold at that time.

High-performance packaging

After fabrication, it's time for packaging. The wafer is cut into individual pieces called die. The die is packaged between a substrate and a heat spreader to form a completed processor. The package protects the die and delivers critical power and electrical connections when placed directly into a computer circuit board or mobile device, such as a smartphone or tablet.

Intel makes chips that have many different applications and use a variety of packaging technologies. Intel packages undergo final testing for functionality, performance and power. Chips are electrically coded, visually inspected, packaged in protective shipping material for shipment to Intel customers and retail

http://www.intel.com/content/www/us/en/history/museum-making-silicon.html

I'm still looking for the video on packaging it is so cool and vary secret nowadays, because I can hardly find anything.

Here is one with heat testing shown. How the Intel Processor is Made.

 

[EarthDog]

What you copy/pasted above shows performance, functionality, and power...

The yt link shows they sit in an oven, unpowered. How is that heat testing when it's notbpowered on?

I guess we have different meanings of 'heat testing'.

 

[Tech Tweaker]

Wonder why they gave it so much extra vcore if it's stable at 1.14 volts?

Just to ensure that nothing goes wrong ever when the customer gets it I guess.

This is not the first time I've seen a CPU run with considerably less than stock voltage. I had an AMD socket 939 Athlon 64 4000+ Clawhammer that would run stock frequency at 1.35v, when the stock voltage was actually 1.5v. A link to that thread here.

Anyway, I think this processor is fully stable at 1.13v. I ran Prime95 small fft's test on it for six hours last night and it didn't have any errors or warnings, and the system did not lock up or blue screen.

Under my normal stability testing if I can get a system to run Prime95 (under any test) for three hours or more I consider it to be stable.

 

[Silver_Pharaoh]

Well done!

 

[PolRoger]

I can't believe it either. That's why I had to post pics of it, because I figured no one would believe it otherwise.

Well, 1.13V is the minimum voltage for this particular CPU to be stable at 3.8GHz. I ran Prime95 blend test with it for a little over two hours, before shutting it down to move on to the next test.

I tried 1.12V next, but it couldn't get into the OS, just blue screened when it hit the Windows load screen.

Very impressive, I must say. It would likely require more voltage if running under worse cooling (i.e. stock or otherwise) due to thermal resistance, but I really don't think it could run on any lover voltage with a custom watercooling loop or something of that sort (since it's running so cool as it is with air cooling).

A note to others with same model CPU who may see this: Your Mileage May Vary, as in it may require more or less voltage to keep your CPU stable at the same speed. Or perhaps your CPU may not be able to run at the same speed as this one at all, or it may be able to run at a higher speed. Just the luck of the draw as to how good the silicon in the chip is and what it can do.

Hey -TT,

I like your "overclocking escapades" threads.

This one actually made me curious about my SB cpu? I've tested quite a few... Mostly 2600K... but also a couple of 2500K. However I've never tried running/testing any of my samples at really low voltages/lower overclocks? I was always more interested in 44x/45x/46x(+)...

I've got a nice clocking 2500K and after seeing your run/test(s) on the i5-2400... I decided to try and see how my sample would fare?

In BIOS vcore was set to 1.100v but it BSOD before an hour had passed while stressing with Prime blend... Bumped up the vcore to 1.115v and it seemed to stabilize the overclock.

4.0GHz with ~1.1v (stress load):

 

[Silver_Pharaoh]

Hmmm so it looks like Intel gave some extra vcore for all sorts of chips then.

Nice job PolRoger!

 

[wingman99]

I have seen a videos of the process packaging and they test for heat, stress and selecting voltage. I cant remember if they test for cold at that time.

High-performance packaging

After fabrication, it's time for packaging. The wafer is cut into individual pieces called die. The die is packaged between a substrate and a heat spreader to form a completed processor. The package protects the die and delivers critical power and electrical connections when placed directly into a computer circuit board or mobile device, such as a smartphone or tablet.

Intel makes chips that have many different applications and use a variety of packaging technologies. Intel packages undergo final testing for functionality, performance and power. Chips are electrically coded, visually inspected, packaged in protective shipping material for shipment to Intel customers and retail

http://www.intel.com/content/www/us/en/history/museum-making-silicon.html

I'm still looking for the video on packaging it is so cool and vary secret nowadays, because I can hardly find anything.

Here is one with heat testing shown. How the Intel Processor is Made.

What you copy/pasted above shows performance, functionality, and power...

The yt link shows they sit in an oven, unpowered. How is that heat testing when it's notbpowered on?

I guess we have different meanings of 'heat testing'.

She says in the video QUOTE: last stop quality control testing, up to 12 hours in a oven heated to 140 degrees c.

Also the Quote I posted above "final testing for functionality, performance and power". The power would be electrical because they already tested for performance.

 

[Tech Tweaker]

Hey -TT,

I like your "overclocking escapades" threads.

This one actually made me curious about my SB cpu? I've tested quite a few... Mostly 2600K... but also a couple of 2500K. However I've never tried running/testing any of my samples at really low voltages/lower overclocks? I was always more interested in 44x/45x/46x(+)...

I've got a nice clocking 2500K and after seeing your run/test(s) on the i5-2400... I decided to try and see how my sample would fare?

In BIOS vcore was set to 1.100v but it BSOD before an hour had passed while stressing with Prime blend... Bumped up the vcore to 1.115v and it seemed to stabilize the overclock.

4.0GHz with ~1.1v (stress load):

Thanks, good to hear someone likes these threads.

Looks like you had a good result there on your undervolted-overclock.

Lower than I would have thought possible for that level of overclock.

 

[EarthDog]

She says in the video QUOTE: last stop quality control testing, up to 12 hours in a oven heated to 140 degrees c.

Also the Quote I posted above "final testing for functionality, performance and power". The power would be electrical because they already tested for performance.

The yt link shows they sit in an oven, unpowered. How is that heat testing when it's not powered on?

....Where is the cold testing you mentioned? What exactly is that? Is it done powered on?

But we are drifting off topic.. I'm just not sure I understand what you meant by "cold" and "heat" testing. I didn''t think baking them in an oven unpowered was testing for anything relevant for a powered on CPU.

 

[wingman99]

She says in the video QUOTE: last stop quality control testing, up to 12 hours in a oven heated to 140 degrees c.

Also the Quote I posted above "final testing for functionality, performance and power". The power would be electrical because they already tested for performance.

But we are drifting off topic.. I'm just not sure I understand what you meant by "cold" and "heat" testing. I didn''t think baking them in an oven unpowered was testing for anything relevant for a powered on CPU.

The CPU's are not just sitting in a oven, what is the point in that? She says quality control testing, up to 12 hours in a oven.

In the video I can see all the electoral connections in the oven.

So anyway the point of all this is why can you under volt so much, with close to stock CPU clocking that OP is doing?

Well I'm done with this, I cant find the Intel video that shows all the testing procedures with heat and calibration, so the stock CPU's can operate fine in extreme conditions as per specifications.

 

[Silver_Pharaoh]

Been running my i5 2300 @ 107 Mhz BCLK without issues for the past 8 hours...
Prime stable for the 1.5 hours I ran it.

Wonder if the system will be stable at 108 or 109 Mhz BCLK...
Guess we'll find out

 

[Tech Tweaker]

Been running my i5 2300 @ 107 Mhz BCLK without issues for the past 8 hours...
Prime stable for the 1.5 hours I ran it.

Wonder if the system will be stable at 108 or 109 Mhz BCLK...
Guess we'll find out

Couldn't say.

Seems to vary depending on the quality of the silicon in the CPU and what the board and other components can handle. Mine topped out at 105.92 BCLK, then strangely stopped being stable at that clock and I had to back it down and it wouldn't be stable past 105.6 IIRC.

Yours seems to be more tolerant of BCLK increases. Good luck with it.

 

[Silver_Pharaoh]

Couldn't say.

Seems to vary depending on the quality of the silicon in the CPU and what the board and other components can handle. Mine topped out at 105.92 BCLK, then strangely stopped being stable at that clock and I had to back it down and it wouldn't be stable past 105.6 IIRC.

Yours seems to be more tolerant of BCLK increases. Good luck with it.

108 was stable, then 30 minutes into prime95 it just froze.
I'm wondering if bumping up all the voltages in the BIOS by a notch would help?

 

[Silver_Pharaoh]

And there we go:



Nearly 5 hours of Prime95 blend. Stable


It won't post anything over 108Mhz, so that's as high as the chip and mobo will go.

I'd say that's a nice overclock considering the turbo was only 2900Mhz stock



Also, I apologize for the thread necro

 

[Tech Tweaker]

And there we go:

View attachment 148480

Nearly 5 hours of Prime95 blend. Stable


It won't post anything over 108Mhz, so that's as high as the chip and mobo will go.

I'd say that's a nice overclock considering the turbo was only 2900Mhz stock

Also, I apologize for the thread necro

Wonder why your temps are so much higher than mine (~20°C). What kind of cooling are you using Pharaoh?

 

[Silver_Pharaoh]

Wonder why your temps are so much higher than mine (~20°C). What kind of cooling are you using Pharaoh?

Stock with that crappy *** loud fan with speedfan controlling it's RPM's to no more than 1500 RPM.
Any higher, and it's like a jet engine.

I was thinking maybe a H60i or a better air cooler, or maybe just a better fan.