2600K @ 5GHZ, How common is it really???

[cytg111]

Finally some one out there understands like me

fish much? shhh

 

[grkM3]

people dont know that almost every chip will do 5+ghz it just takes a lot of volts to push them up there.

I can bench at 5200 mhz but I dont like jaming 1.55 volts into this chip but it will be stable given enough volts.

people get scared when they see 70c temps even tho the chip wont throttle until 99c.

I can almost bet that most chips with the right board will hit 5ghz at 1.5 volts

 

[OILFIELDTRASH]

What 4.5ghz ain't enough?

I remember when 4ghz was the holy grail. Now with a couple keyboard strokes you can get 4+ without even trying.

 

[LagunaX]

Preliminary Ivy Bridge 3770k testing shows ~10% performance gain over the 2600k @ the same speed.
However, they were "only" able to get it to 4.8ghz - not 5ghz - for cinebench and a coupla other benchmarks.
However, the IMC can handle speeds of ddr3 2400 and pci-e 3.0 whatever small advantage it has if any over the current pci-e.

Undoubtedly the cpu it will be more expensive too.

Bottom line just go ahead and pick up a 2500k or 2600k and a z68 motherboard that will be able to upgrade to Ivy Bridge with a bios update in the future.

Then the only thing you'll be swapping out would be the cpu in the future - and you would still be able to sell the cpu.

 

[T101]

I am running my 2600k at 4.5Ghz (1.31V, 4.6 required 1.40, and I saw no point in going that high). It completely smokes my Q6600 clocked at 3.6Ghz in everything I do. It even smoked it on stock speeds.

The Q6600 is certainly an ok cpu, but the 2600k is just plain better. Games that stuttered on my Q6600 runs flawlessly smooth on my 2600k. And that is using the same GPU (470 GTX).

 

[MrX8503]

5ghz is very uncommon. Even 4.8ghz is uncommon. Expect a 4.5ghz, with a 4.7ghz being very good.

 

[exar333]

5ghz is very uncommon. Even 4.8ghz is uncommon. Expect a 4.5ghz, with a 4.7ghz being very good.

It is important to distinguish between what's 'possible' and what people are running 24/7. 5ghz is pretty much possible on almost all 2600Ks, but most people don't want the volts needed to keep it stable at those speeds. Thats why you see most folks at the 4.4-4.8ghz speeds.

 

[Idontcare]

It is important to distinguish between what's 'possible' and what people are running 24/7. 5ghz is pretty much possible on almost all 2600Ks, but most people don't want the volts needed to keep it stable at those speeds. Thats why you see most folks at the 4.4-4.8ghz speeds.

Enter the PTPP!

 

[slayernine]

Put together a system for my friend with a good Asus mobo and i5 2500k a few months ago. Asked him if he wanted to see how well it overclocks for fun. Messed around with settings in bios and after a few reboots I had it up to 4.8GHz and ran Prime95 for about an hour no issues. After that I reduced it to a much milder overclock and left it at that. If that was my system I would have easily hit 5Ghz with a bit more time at home.

I say it is largely a matter of luck at getting a good OC'ing processor as I've had other processors that were "good" at overclocking and could never get clocks as good as other folk.

 

[tweakboy]

I'm running my i5 2500K at the top multiplier (57). It runs at 5.7GHz all day at 1.55v and the temps never exceed 95 Celsius under 100% load.

WOW thats a sick OC my friend.
1.55v ,, I hope your under water or have a 60 dollars HSF making noise to cool tha thing down. 90c means lesser life span.

Your temps are way too high to put the CPU voltage where it is. I know at certain degree the comp will shut off by itself to save the CPU.

Im surprised you took her to 90c You should be getting 60's c low seventies, when you OC

Everyone can raise a number for a bigger OC but you have to take consideration in cooling and voltage setting. The only reason your stable @ 5.7ghz is the crazy OC ,, but if you don't have her under water or a expensive HSF like True 120 or whatever. Lapped too!!!

Take her back down to 1.31v would let u go to 4.2Ghz stable,, for 5Ghz you need I would say,,, 1.41v gl

 

[exar333]

WOW thats a sick OC my friend.
1.55v ,, I hope your under water or have a 60 dollars HSF making noise to cool tha thing down. 90c means lesser life span.

Your temps are way too high to put the CPU voltage where it is. I know at certain degree the comp will shut off by itself to save the CPU.

Im surprised you took her to 90c You should be getting 60's c low seventies, when you OC

Everyone can raise a number for a bigger OC but you have to take consideration in cooling and voltage setting. The only reason your stable @ 5.7ghz is the crazy OC ,, but if you don't have her under water or a expensive HSF like True 120 or whatever. Lapped too!!!

Take her back down to 1.31v would let u go to 4.2Ghz stable,, for 5Ghz you need I would say,,, 1.41v gl

This brings-up a good point about ambient temps. If you are riding the line at 90C, is that with just the CPU stressed, or with other components at 100%? When I stress-test my CPU, I usually have my GPU running at 80-100% as well so my ambients represent the worst-case temps.

It's one thing to just have your CPU maxed-out, but you don't want crashing or throttling when other components are also stressed. This may not apply to all, but plenty of folks also are stressing multiple HDDs and/or GPUs as well as the CPU. This goes back to good case construction and airflow, but those can only do so much with X amount of heat.

 

[MGF Derp]

It is known that Sandy Bridge is good to go up till around 1.52V if your temps are ok. Everyone in here that is scared to go above 1.35 needs to put on there big boy shorts and just do it. Or get better cooling. Corsair H series, water loop, DH14. Stop running 212's and get higher clocks. And chips can be dialed to 5Ghz as low as 1.4V with enough know how and practice.

 

[DrBoss]

I run my 2600K @ 4.4GHz 24/7
http://forums.anandtech.com/showpost.php?p=31551449&postcount=78
core voltage +1.2
offset voltage -0.05
additional turbo voltage +0.02

I've taken the chip to 5Ghz, 12 hours stable on Prime & Everest. Didn't break 70C... but i don't have the voltage numbers i was running on me at the moment.

If you plan on running the chip on air, i have nothing but great things to say about the Thermalright Archon heatsink.
My build: http://forums.anandtech.com/showthread.php?t=2155564

Sandybridge was my first chip since Conroe... needless to say, it was a huge leap forward for me... very impressed with intel this gen... though i tend to agree with those saying your current chip is pretty damn good. It's certainly got some life left in it. If i were you i'd wait a year.

 

[grkM3]

I'm running my i5 2500K at the top multiplier (57). It runs at 5.7GHz all day at 1.55v and the temps never exceed 95 Celsius under 100% load.

sure you are

Ill paypal 100 bucks if you make a video of your chip booting on a 57x multy into windows and running load with just 1.55.

I dont think there has ever been a 57x multy chip ever seen with liquid nitrogen cooling and 1.7 volts

some are lucky and have 55x chips and run the bclk at 104 for 5.7ghz runs

 

[Grooveriding]

It is known that Sandy Bridge is good to go up till around 1.52V if your temps are ok. Everyone in here that is scared to go above 1.35 needs to put on there big boy shorts and just do it. Or get better cooling. Corsair H series, water loop, DH14. Stop running 212's and get higher clocks. And chips can be dialed to 5Ghz as low as 1.4V with enough know how and practice.

Best of luck to all running their SB at 1.5V on air!

 

[njdevilsfan87]

It is known that Sandy Bridge is good to go up till around 1.52V if your temps are ok. Everyone in here that is scared to go above 1.35 needs to put on there big boy shorts and just do it. Or get better cooling. Corsair H series, water loop, DH14. Stop running 212's and get higher clocks. And chips can be dialed to 5Ghz as low as 1.4V with enough know how and practice.

I'd rather not lose my hearing due to running SB at 5ghz. I can push 5ghz, but that requires my turning of the H100 from low to high. Low is quiet, medium is loud, and the high setting is just retarded loud. Maybe I need different fans, but until then, I greatly value the peace and quiet.

 

[jimpatrick]

It is known that Sandy Bridge is good to go up till around 1.52V if your temps are ok. Everyone in here that is scared to go above 1.35 needs to put on there big boy shorts and just do it. Or get better cooling. Corsair H series, water loop, DH14. Stop running 212's and get higher clocks. And chips can be dialed to 5Ghz as low as 1.4V with enough know how and practice.

yeah and as if they are going to stick to it for years ^_^

 

[The Sauce]

From what I have gathered 4.5-4.8 is a safe bet at reasonable voltages (i.e. 1.4v or less) with high end air. For gaming my upgrade from an i5-750@4.2 to my 2500k @ 4.7 yielded a palpable difference in system and gaming performance. Make sure to pair with the correct mobo or your overclocking will suffer. The best o/c'ers seem to be the higer-end Z68 ASUS boards

 

[Makaveli]

From what I have gathered 4.5-4.8 is a safe bet at reasonable voltages (i.e. 1.4v or less) with high end air. For gaming my upgrade from an i5-750@4.2 to my 2500k @ 4.7 yielded a palpable difference in system and gaming performance. Make sure to pair with the correct mobo or your overclocking will suffer. The best o/c'ers seem to be the higer-end Z68 ASUS boards

You were actually able to feel a difference from 4.2 to 4.7 without looking at benchmarks?

 

[aigomorla]

i honestly have yet to see a 32nm cpu last longer then 4 months without degredation at anything higher then 1.45vcore.

I honestly have yet to see a cpu NOT die in less then 6 months at 1.5vcore, unless ur pulling a sub ambient cooling system on it.

Even then ive seen Extreme Edition cpu's benched only 3 times @ 1.6+ vcore on LN2 end up crapping out and suffering major degradation... thats 3 bench sessions guys.... rougly less then 1 month of actual usage.

 

[gramboh]

i honestly have yet to see a 32nm cpu last longer then 4 months without degredation at anything higher then 1.45vcore.

I honestly have yet to see a cpu NOT die in less then 6 months at 1.5vcore, unless ur pulling a sub ambient cooling system on it.

Even then ive seen Extreme Edition cpu's benched only 3 times @ 1.6+ vcore on LN2 end up crapping out and suffering major degradation... thats 3 bench sessions guys.... rougly less then 1 month of actual usage.

Thanks for sharing your experience, nice to see some reality from one with experience at those levels in here.

Also, i5-750 to 2500k 4.2 to 4.7 with one 5850 made a noticeable difference in real world gaming? Sounds like confirmation bias to me .

 

[jimpatrick]

i honestly have yet to see a 32nm cpu last longer then 4 months without degredation at anything higher then 1.45vcore.

I honestly have yet to see a cpu NOT die in less then 6 months at 1.5vcore, unless ur pulling a sub ambient cooling system on it.

Even then ive seen Extreme Edition cpu's benched only 3 times @ 1.6+ vcore on LN2 end up crapping out and suffering major degradation... thats 3 bench sessions guys.... rougly less then 1 month of actual usage.

then a 22nm ivy will degrade faster?

 

[Grooveriding]

For a quick epeen stroke. 5ghz is stupid fast on a SB-E



I've gotten the voltage down some from this as well after reducing my LLC settings, 5ghz at 1.47 now. Temps are a bit better now too.

 

[LOL_Wut_Axel]

For a quick epeen stroke. 5ghz is stupid fast on a SB-E

I've gotten the voltage down some from this as well after reducing my LLC settings, 5ghz at 1.47 now. Temps are a bit better now too.

Very good, but you know you're doing a lot of damage to your CPU. I wouldn't think in a million years of doing that to one of the world's fastest desktop CPUs, which happens to also cost a healthy $600. You know if you keep running it like that you'll end up with a heavily degraded chip in less than a year, which could mean something like a measly 4GHz will be out of reach.

 

[The Sauce]

You were actually able to feel a difference from 4.2 to 4.7 without looking at benchmarks?

Well, you missed the part about the processor upgrade as well (i5-750 > 2500k). Sure it's subjective. Sure it's likely somewhat biased. "Feels significantly faster" is what I was shooting for when I got the upgrade in the first place