- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I have an i7-4790K in a GA-Z97MX Gaming 5 mobo latest F4 BIOS.
I used a Noctua NH-L12 cooler rated at 95W for cooling the processor rated at 88W in a Lian Li PC V354 with 4 fans, 2 in 2 out, case closed and case open, ambient is 27-30°C.
Temperatures in BIOS and memtest86+ were high so I decided to try stress testing and in Prime95 small FFT cores 1&2 overheated to 100°C using Core Temp.
I tried reseating the heatsink and renewing the NT-H1 TIM and opening the case but it made no difference. I have a photo of the contact pattern here.
When I tested using the OCCT benchmark I was unable to complete a test due to the processor overheating so I underclocked the processor to 3.6 GHz, disabled turbo and manually set vCore to 1.1v.
With an underclocked processor I was able to get a heating and cooling curve using the OCCT auto capture, to enable me to study the problem.
Even when underclocked the processor was reaching high temperatures, rapid fluctuations in temperature with work load suggest a bottleneck in the thermal pathway. When I tested with the intel retail cooler which came with the CPU the cooling was much less effective than the NH-L12 (even when underclocked taking just over a minute of OCCT to reach the 85°C cut off point see below) indicating the NH-L12 was doing a good job of removing heat, which meant the processor was making the heat or the source of the bottleneck.
I have discussed it http://forums.hexus.net/cpus/327593-4790k-overheating-nh-l12.html elsewhere. Advice was to contact Intel due to an absence of information relating to my retailer's testing procedures. I have asked about these but am still waiting for a reply.
So my question is how should I proceed from here? Does this qualify for an RMA? If so is it possible to negotiate this with Intel direct or do I have to go through my retailer?
I have done my best to make sure I am not doing anything wrong and I would be grateful for any pointers to any mistakes I may be making.
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Bulldog,
Your most recent post got erased (I think), but no, I haven't given up on you. I'm just really confused on how to help you. More than anything I don't think you understand what I'm saying because I keep telling you to leave everything alone and you keep changing a setting, or 2, or 3... I honestly don't think there is anything wrong with your CPU. Something isn't right about your settings, however. What I'm trying to have you do is just put everything back to stock, but either 1) you're not doing it right, or 2) something is wrong with the way your clear CMOS/dual BIOS feature works. I have to admit that I have very little experience working with dual BIOS systems, but I don't think it would be that different from single BIOS systems.
"I have reloaded optimized numerous times, which puts everything to auto as you say. I had everything on auto when I entered XTU and set the 3 settings 88W, 160W, 265A ran the benchmark. It gave me a score of 878 which is quite low, but NO throttling. So I enabled XMP. Go do it again, get a 1073 with Power throttling."
- This is what I'm talking about... Can you tell me why you entered XTU to set the 3 settings before running the benchmark? Can you not run the benchmark with auto settings? Will it give you high temperatures? For you to have no throttling without XMP and then have throttling when you turn on XMP, that's telling me you have your settings set too low. I know I keep saying the same thing over and over again, but put everything to auto and leave everything alone. Don't change anything at all and then run benchmark/stress test. Then do the same while ONLY changing XMP. Again, same thing here, but I have no idea why you are changing anything to a static setting. You're not overheating!
As far as your AIDA64 benchmarks you are running, it's the same deal with Prime95. I can run and pass those benchmarks, but I have 100C+ when running them. I think by figuring out what these tests do, people narrowed it down to AVX instructions that were possibly causing this. But as we were discussing earlier in this thread, 99% of what most people do with their computers will never use those instructions, so for all intents and purposes, your CPU is fine if it can pass Intel PDT and the stress tests on Intel XTU. I know it's a stupid answer, and I totally disagree with it as well, but as of now, we gotta go by what Intel said which is to stay away from 3rd party CPU stress tests (AIDA64, Prime95, etc.). I do, however, recommend using the AIDA64 stability test to check your system stability. You can also use OCCT to do this which works fine for me. If you're looking to get what you paid for, you purchased the wrong processor. Like I said earlier, I would put my money on getting a $20 check from Intel 20 years from now to tell us how sorry they are for how this processor is performing. However, if you want decent performance on games, ability to edit videos quickly, and just generally have a good performing, stable PC, then stop messing with it and just run everything on auto. Key word: EVERYTHING. Your computer won't blow up, it won't crash, it will be fine.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Well I am willing to try! So let me just clarify a few things,
-reload optimized, leave everything on auto
+do I turn on XMP so I am actually utilizing my RAM?
+When you open XTU, does it give you these settings
*all cores 44x
*Turbo boost Power Max 3408.250W
*Turbo Boost Short Power Max 4090W
*Turbo Boost Power Time 0.00097656 Sec
*Processor Current Limit 1000W
So I don't change any of these?
+What do you feel is OK for the TDP max to be?
As far as the Dual BIOS, I don't actually think it is an issue, since the nature of the feature is that it only uses the backup if something in the primary is corrupted, which isn't the case here. I am just running Memtest86+ at the moment as I had left my computer to just idle for a few hours and I came back to a rebooted comp. So I looked in the minidump, then found a program to READ the minidump, then did a google search for what the heck it was! haha it said something like Memory Management, or something, and that it was ntoskrnl.exe issue, so now I'm trying to figure out what the heck that means! Yup, this whole 'build your own computer' thing SURE seems like a great idea! haha
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
No, don't change any of those. Run Intel PDT and stress test in Intel XTU. Do you pass with OK temperatures?
As for your TDP, Intel says your TDP should be 88W at stock settings. However, that doesn't mean it's dangerous for it to go over 88W. As we discussed earlier, TDP is just a measurement Intel gives so you have an idea of what kind of cooler you're going to need to properly cool your CPU. For example, for TDP of 88W, a stock fan may be sufficient in keeping it cool whereas a liquid cooling system may be needed to cool down a TDP of 200W. So quick answer is that there is no answer to your question "What do you feel is OK for the TDP max to be?". All that matters is: when you're running at 100% load, is the cooling system you have sufficient in keeping your temperatures in safe range? If you're not overheating, TDP is not something you need to be concerned with.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Forgot to mention, try first with XMP off. Then try again with XMP on. You should be passing both Intel PDT and stress test in XTU at acceptable temperatures with XMP on. Remember, don't change ANYTHING else. NOTHING else.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
so don't change the...... haha just kidding! OK, I just never actually knew what TDP was, all I knew was that It was supposed to be 88W, so when I saw it going over it worried me. After I'm done running the RAM I will try. So far all is well, have to start trying different slots on motherboard
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
How many sticks of RAM do you have? Make sure you read your motherboard's manual on which slots to put the RAM in. If you don't put them by manufacturer's suggestions, your RAM will not run optimally. For me, my Maximus VII Formula suggests that for 2 sticks of RAM, they need to go in slots 2 (B1 DIMM) and 4 (B2 DIMM) (counting from the CPU). This is the usual configuration for 2 sticks, but again, it can be different per manufacturer so make sure to check your manual.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Ok, here we go! So I did like you said, running Intel Stress test and IPDT and passed everything, stress was 71*C or so. Now, I ALSO ran an XTU benchmark, and here is where some big differences are.
Total default
Core Volt Max 1.262
TDP 106W
Temp 72*C
score 884
Default with XMP
Core volt 1.262
TDP 126W
Temp 84*C
Score 1106
My previous Settings I had Core volt set to 1.150, LLC to Extreme, XMP
Core volt 1.150
TDP 81W
Temp 65*C
Score 1113
So not sure what to think.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Oh, and I didn't touch anything in the actual XTU program, just changed the couple things in BIOS and left everything maxed in XTU sampark1980
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
OK! Now, I know you were throttling when you had static voltage of 1.150 with XMP. Are you throttling with all your settings maxed out?
Also, why are you changing load line calibration to extreme? Do you know what you're doing there?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I swear man, every single time, I say don't touch anything and your response is: "I did exactly as you said.... except I touched these 1 or 2 other things"
What's the deal??
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
NO! When I ran the tests like you said, DEFAULT, I didn't touch ANYTHING! Then I ran it and only changed XMP. I was merely stating that when I ran MY settings that I had used previously (1.150,XMP,LLC) that I didn't touch anything in XTU, I changed in BIOS. So the tests I ran were EXACTLY what you said to do. And no I wasn't throttling anything at 1.150 since the default settings give XTU the max setting on all the options. I changed the calibration to extreme because, well this will sound stupid, but, that's what the guy in the video said to do. hahaha, I was watching a video on You Tube called Intel 1155 Overclocking Guide, by OC3D TV. It came up when I was doing a search about setting up a Gigabyte motherboard, so I watched it. The guy does a good job of giving some explanation on what some of the different settings are in the Gigabyte BIOS.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I just read back what I typed earlier and I can see how you would think that I changed something. Unfortunately there is no EDIT button, or I would have included my second response (about not changing in XTU) to my first post under where I put my 'previous settings' items. I just thought I would add it so that you would know that in my previous settings that all XTU options were left at default, which is maxed out on everything. Sorry about that.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Just to clarify what I'm trying to do, I'm trying to help you do what you said your goal is, which is to have your CPU function the way it's supposed to function. You don't care about bragging rights on benchmarks, etc. etc. One of the things to determine if your CPU is functioning correctly is to make sure there is no throttling of any kind. So then:
"ok, this is getting silly. So I ran the Benchmark in XTU and had no Thermal Throttle, Power Limit Throttle, or Current Limit Throttle. so I thought ok, I'll put my memory back to XMP mode (G.Skill Trident X 2400 32G) and then I get Power Limit Throttle again! What the heck. difference in Benchmark score (XTU) was 878 without, 1073 with XMP"
This is what you said a couple of posts ago, hence why I thought you are throttling with VCore set and 1.150 and XMP turned on. Now you're saying you WEREN'T throttling... So then what are we doing?
I feel really bad for everyone else following this thread because it seems it's going down a different path from the intention of the OP.... So let's try to keep this conversation on here minimal...
If you're still with me on this, lets go back to:
Default with XMP
Core volt 1.262
TDP 126W
Temp 84*C
Score 1106
This is great right here except your temperatures are a little bit higher than what we want it to be. From here, we're ONLY going to touch the following settings: Turbo Boost Max - 88W; Turbo Boost Short Max - 110W; Processor Current Limit: 256A, keep power time window at 8 seconds or less. Leave XMP on. DO NOT TOUCH ANYTHING ELSE. Re-run benchmark and see if your temperatures are in the 70's.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Done, benchmark 1101, temp 74 with an average of 15% Power Limit Throttling (98% max) TDP 109. The part that you quoted was when I had Intel Ken's settings, the ones I just replaced back in. In my post where I had LLC at extreme, I left all the XTU options maxed as I stated, so therefore no throttling of any kind.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I'm not sure if I'm reading that right, but you're getting a max power limit throttle of 98%?? ummm... if that's the case then yes, there is something wrong with your processor...
Otherwise, f your average throttle is about 15%, then all you need to do now is increase your Turbo Boost Power Max and Turbo Boost Short Power Max in small increments. Start at 10W each time until there is no more power limit throttle. Remember, leave EVERYTHING else alone.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I just increased it like you said, 10 at a time, but I am at 108 and 130 and still getting Power Limit Throttling and my temp was 85*C, so I really don't see how I can keep adding more, or rather what the point would be since with my Core Voltage sitting in the 1.210-1.262V range the temps are climbing too high. Oh and the 98% was just on XTU when you hover the cursor over Power Limit Throttling it says Recent Max 98%. I'm thinking that if I want to keep the VCore on auto I will have to do a negative offset on it of 0.05-0.07mv to get my temps back down. Is doing it that way better than applying a static VCore?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Oh, and you said leave everything else alone, so I haven't changed the core Multipliers, they are all at 44x. Did you want them to be at 44,44,43,42?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
If it were me, I would adjust offset settings before touching vcore. You should technically have it at 44, 44, 43, 42, which is "true" stock as having all cores set at 44 is in essence "overclocking", albeit a very small overclock.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Can you post a picture of this 98% power limit throttle? I want to see what else is going on when this is happening.
One small confusion:
Default with XMP
Core volt 1.262
TDP 126W
Temp 84*C
Score 1106
This is what you posted when you had completely stock settings, meaning your Turbo Boost and Short Boost max were very, very high - a lot higher than 108 and 130, respectively. You got max temp of 84C. I'm curious why your temperature is higher now that you have it set at 108 and 130.
And also, "I just increased it like you said, 10 at a time, but I am at 108 and 130 and still getting Power Limit Throttling and my temp was 85*C, so I really don't see how I can keep adding more, or rather what the point would be since with my Core Voltage sitting in the 1.210-1.262V range the temps are climbing too high."
- This proves that you CAN and SHOULD add more to your power. You were at maxed out power (lot higher than 108/130) and your temperatures were in the same range as you have now, except you didn't have any throttling. So there IS a point. You should definitely add more until there is no throttling. In fact, you had it up to 160 at one point from another suggestion...
As far as your computer climbing into the mid-80's, this is on the benchmark and not the stress test, so higher temperatures are expected. That's really the whole point of this thread is that Intel recommends 70-78C in temperatures, but obviously it is EASILY going over that, even with their own benchmarks on their own Intel XTU program. We're kind of beating a dead horse here, but with the 4790K, it seems like for most people you either have to cripple the processor for good temperatures, or settle for temperatures a little higher than desired. The key (and this is what I've been trying to help you do this whole time) is finding that sweet-spot where you are JUST within the right temps without crippling your system too much. =/
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Okay, thanks again for your insight. So my question is this, why is it better to leave the Vcore on auto instead of a static setting? I'm asking because I don't know the answer. When I look at the numbers, I see Ken's settings were 1.108..... I don't know if that makes it a static voltage setting, but I think so. When I had 1.150, XMP, and LLC Extreme, I was at all 4 cores 44x, and had a temp of 65 in the stress test, 71 in benchmark with 1113 score. Please keep in mind I'm not being argumentative, I know sometimes things can be taken wrong when they are typed, I am simply trying to better understand why one way may be better than another. Oh and I was able to alter my 88 setting to 91 after I changed the multipliers to 44,44,43,42 at 130w and not have any power throttle.
Thanks
![](/skins/images/872293744008A34B36F8ABF94A46CC66/responsive_peak/images/icon_anonymous_message.png)
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page