i7 7700K worth buying after all those Over Heats Dilemma?

[trparky]

I'm not sure why anyone would want to buy a "K" SKU from Intel

Because even if you don't overclock it the K chips do have a higher base clock than their non-K cousins.

For instance, the Core i7 7700 has a base clock of 3.6 GHz whereas the 7700K part has a base clock of 4.2 GHz. So even if I'm not going to overclock the chip the 7700K chip is better since it offers up a higher base clock with no effort involved.

 

[EarthDog]

I mean, cars/bikes are governed, their top speed. They arent designed to operate safely above those speeds...instead of limiting multiplier like SB, they used crap TIM.

 

[Seba_82]

Hi everybody!, my 7700K run with 90% load at 55-60 º with a EVGA WC 280... so i gues its luck...

 

[EarthDog]

90% load.....what is the load, actually? Most here talk about worst case/stress testing, not gaming. My 7900X at 4.5 hit 60C while gaming.............................................let's put some perspective on things, ehh?

 

[Seba_82]

I talk about playing, stress tests do not consider them at all a real-life scenario. If they want when I get to work I upload a screen capture, sorry for my bad english...

 

[hapkiman]

As an owner of the i7-7700k which I've had since January, 2017 - I can honestly say that I have had zero heat issues with it, and this is an incredibly fast processor (which I have overclocked many - many times, but have it currently running at 4.5GHz across all four cores). Mine tops out at 5.0 GHz BTW, and is very stable.

Yes I realize that many have had heat issues with their i7-7700k's (primarily when OC'ing), and I admit that I do indeed have a better than average AIO (a Corsair H115i), and great case air flow - which of course helps, but I would recommend getting a very good cooler for this processor anyway - to avoid any such problems with heat. Nip it in the bud so to speak. The key things to think about if you want to get this proc, you should have a good case with very good air-flow, and a good cooler (you don't have to get a H115i, or a NZXT X62, but I wouldn't get a Hyper 212 EVO and then complain about high temps). And hopefully you have a room with a decent ambient temp, I do - and this also helps. No question, this processor deserves a better than average cooler, especially considering all the talk about "heat issues." Better safe than sorry. Mine idles in the 20's(C) BTW, and hits mid 40's to 50C after hours of gaming. That is fine by me.

Now, if you want to run PRIME95 then yes, you might see some pretty hefty temps, but seriously - PRIME95 is unnecessary IMHO! There was a time when PRIME was the go to test and OC'ers would run it for hours and hours on end just to reassure themselves their systems and OC's were stable. I would submit to you however, that in 2017 PRIME95 it not only no longer necessary, but is way too punishing on modern procs. Yes of course we need programs like this - to test for stability, but PRIME95 is so intense, and places unrealistically heavy loads. Again, before I am attacked furiously by the hardcore OC'ing crowd and PRIME fans, this is only MHO- you guys that love PRIME95, knock yourselves out running it all night. I'm totally cool with that.

Just not for me and my 7700k. After having this hobby (buying, building, tinkering, reading with/about computers. etc..) for 20+years, I feel pretty confident that I know when my system is stable/unstable. And yes, I do occasionally stress test my rig (CPU, RAM, and GPU) usually with AIDA64 Extreme. It's just that I seldom do it anymore, usually just after making a new build or changing a major component, or if troubleshooting a friends/family's rig who is having issues (just ran MEMTEST86 overnight a few days ago for my sister's rig as an example).

But I digress, and to answer the OP's question: Q: "i7 7700K worth buying after all those Over Heats Dilemma?"
A: Yes, it is worth it. And the "heat dilemma" is way exaggerated and a non-issue for most of us. This is a great and cool running proc for me. I would even say it has been fantastic.

Do I wish Intel had used solder on KabyLake's IHS/die? OF COURSE I DO. Who wouldn't want a 2017 version of Sandy Bridge? It would be freakin' awesome. But Intel has apparently, in their infinite wisdom, given up on soldering the IHS for the non HEDT crowd.

 

[trog100]

in a two horse race when one horse can run faster than the other.. only one horse has to try..

back in the day when i first moved from amd to intel the intel horse came well "underclocked" relative to its real performance.. my first core 2 chip came clocked st 3 ghz.. i was soon benching it at 4.5 ghz..

the underclocked intel chip still pissed all over the amd competition.. its a rigged game in the sense that the team with the best horse never runs it any faster than they have to..

the big so called overclocks of the past were not really overclocks they were just running the chip at its real speed rather than the fake rigged game speed..

the secret of way back when was simply to figure out which mid range chip was simply an underclocked high end chip..

in a perfect world one production line would be used to cover all market sections.. an outboard boat motor example.. the 20 horsepower version is exactly the same as the 30 horsepower version except for one thing.. a restriction plate fitted under the carb..

playing this game is even easier when things are chip controlled.. all market segments have to be filled.. the easiest way to do this is make everything the same except for artificial restrictions..

trog

 

[hapkiman]

I have thought people moved on with this 7700K temps issue thing but unfortunately i was wrong move on people.

You are right. Sorry for my soapbox. Lets move on people.

 

[trparky]

in a two horse race when one horse can run faster than the other.. only one horse has to try..

back in the day when i first moved from amd to intel the intel horse came well "underclocked" relative to its real performance.. my first core 2 chip came clocked st 3 ghz.. i was soon benching it at 4.5 ghz..

the underclocked intel chip still pissed all over the amd competition.. its a rigged game in the sense that the team with the best horse never runs it any faster than they have to..

the big so called overclocks of the past were not really overclocks they were just running the chip at its real speed rather than the fake rigged game speed..

the secret of way back when was simply to figure out which mid range chip was simply an underclocked high end chip..

I agree. The days of overclocking are essentially over.

 

[aiVANHO]

What overheat dilemma? You got to stop listening to people. There are no overheating issues with the 7700K.

Guess somehow you've missed the news. But its real. Notcible number of user have submitted feedback on Intel official forum regarding i7 7700K temp spike issue. Even each & every person i know around me who've bought 7700K, been facing this issue since very first time. Its not like they OC'ed their 7700K insanely or in little margin. The stock 7700K spiked weirdly. Please check the link.
https://www.extremetech.com/computi...g-issue-cluelessly-suggests-stop-overclocking

 

[cadaveca]

You might want to be blaming the motherboard, and not the CPU, just sayin'....

I mean, you could just look at the last paragraph and see the negative hype of that post for what it is. 90c is NOT "very" high.


That said, we did have a news post here about this issue as well, so most of us are aware of that reported issue.

 

[Vya Domus]

I agree. The days of overclocking are essentially over.

"Overclocking" never really existed to begin with. You design a chip that operates within a given set of operating parameters. , you then limit those to whatever target you have for the product you want to make. By overclocking all you do is exceed those default parameters but you are still within what the chip was design to withstand given the manufacturing process. Even when you use LN2 you are still bounded by those initial characteristics.

 

[EarthDog]

You are going to split hairs on the definition of overclocking. While you are correct in a manner of speaking, most would certainly think overclocking is raising the clock speeds over its STOCK (what Intel/AMD/NVIDIA sets) values.

 

[cadaveca]

"Overclocking" never really existed to begin with. You design a chip that operates within a given set of operating parameters. , you then limit those to whatever target you have for the product you want to make. By overclocking all you do is exceed those default parameters but you are still within what the chip was design to withstand given the manufacturing process. Even when you use LN2 you are still bounded by those initial characteristics.

I think you missed it. It being what overclocking used to be. It used to be that anything above those "stock" settings required manual adjustments to voltages, timings, and other things, like GTL settings, etc. This took time and effort, and at least a small idea of how things work. Today, you change just a multiplier, and all the rest of the settings are changed for you automatically.

So yeah. How we define overclocking has changed, it seems. Oh boy. You know, it's like not a skill any more, because you don't need to know squat, and anyone can change 2 to 3 BIOS setting and max out a CPU. Memory OC, on the other hand... has been removed because the top-tier ICs are already screened out into the highest-priced kits.

Because even if you don't overclock it the K chips do have a higher base clock than their non-K cousins.

For instance, the Core i7 7700 has a base clock of 3.6 GHz whereas the 7700K part has a base clock of 4.2 GHz. So even if I'm not going to overclock the chip the 7700K chip is better since it offers up a higher base clock with no effort involved.

I know that, but that added 600 MHz requires nearly a 50% increase in power consumption at times, and yet no one thinks that a 18% increase in clocks equating to a 50% increase in power used hints that a CPU is near it's limit? And then they want to push it further, much further? Well, do I really need to explain how that might not be the most prudent expectation? Seems so....

 

[Vya Domus]

I think you missed it. It being what overclocking used to be. It used to be that anything above those "stock" settings required manual adjustments to voltages, timings, and other things, like GTL settings, etc. This took time and effort, and at least a small idea of how things work. Today, you change just a multiplier, and all the rest of the settings are changed for you automatically.

So yeah. How we define overclocking has changed, it seems. Oh boy. You know, it's like not a skill any more, because you don't need to know squat, and anyone can change 2 to 3 BIOS setting and max out a CPU. Memory OC, on the other hand... has been removed because the top-tier ICs are already screened out into the highest-priced kits.

It changed from the point of view of the user , yes , but under the hood it's still the same thing. In terms of whether or not it requires less skill , I think it's just a matter of how much you're allowing yourself to have the impression that your "cheating/hacking your way" because in the end you would have still hit the same limits. You can't really modify IC's , they have fixed characteristics , you're just asking them nicely to run faster within those characteristics and do your best so they don't blow up

You are going to split hairs on the definition of overclocking. While you are correct in a manner of speaking, most would certainly think overclocking is raising the clock speeds over its STOCK (what Intel/AMD/NVIDIA sets) values.

Eh , of course the idea still needs to be kept simple.

 

[EarthDog]

That's the idea, period. It has nothing to do with the properties of the silicon. Intel/AMD/NVIDIA sets 'stock' clocks, and we push it past that... its called overclocking.

 

[cadaveca]

It changed from the point of view of the user , yes , but under the hood it's still the same thing. In terms of whether or not it requires less skill , I think it's just a matter of how much you're allowing yourself to have the impression that your "cheating/hacking your way" because in the end you would have still hit the same limits. You can't really modify IC's , they have fixed characteristics , you're just asking them nicely to run faster within those characteristics and do your best so they don't blow up

We needed to tweak the BIOS, not the ICs...

For example, I was asking for trfc settings in BIOS before we had them. I started the conversation that lead to the fail B2B tweak on X58 from ASUS (TRC was missing on DDR3, we needed similar I said, fellow staff at the site I was working at at the time took that to ASUS, and ASUS offered the B2B option).

THAT is overclocking. Simply changing settings... meh. My 9-year-old can and does do it.

 

[Vya Domus]

We needed to tweak the BIOS, not the ICs...

For example, I was asking for trfc settings in BIOS before we had them. I started the conversation that lead to the fail B2B tweak on X58 from ASUS (TRC was missing on DDR3, we needed similar I said, fellow staff at the site I was working at at the time took that to ASUS, and ASUS offered the B2B option).

THAT is overclocking. Simply changing settings... meh. My 9-year-old can and does do it.

Exactly , for the most part we tweak the BIOS/software just like in your example , not the IC/hardware. That's my point , overclocking is bounded by how the manufacturer of the IC/hardware designed it , what they allow to be achieved through software is nothing more than their "generosity". Now if you can find some nifty way to bypass all that , that's what I call "true overclocking".