Batman Returns

[E-curbi]

NEVER

 

[yotano211]

I AM BATMAN!!!!

 

[E-curbi]

I AM BATMAN!!!!

Now you know, there's not room enough for two Batmans in Gotham City!

----------

Grabbed some new memory sticks for the Batman build, they is insane.

The highest speed on EVGA's Z390 DARK QVL list 4600Mhz CL18. Memory kits do not perform on a linear scale with dollars spent. Not even close, that graph of performance/dollars is so freaking exponential man, and not to mention totally unfair imho.

Planned to just grab the 4400Mhz C18 kit buuuuut after speaking with a few super ddr4 memory enthusiasts, those super memory gents that spend 80+ hours tweaking their 2ndary and tertiary timings to perfection, they convinced me if I pay another $100 oh geez, I'd have a better overall chance at the highest speed/tightest primary timings so reluctantly I went for the Gskill F4-4600C18D-16GTRS dual kit.

I was completely satisfied with my 3200/14 TZ kit and waiting for ddr5 or optane memory next year, but then I plugged into a free Z390 Dark OC Daily LUUMI Bios Preset that pushed my 3200/14 Gskill Samsung B-die kit to 4300Mhz 17CL and IT FELT AMAZING! , yet wasn't so stable...

I was suddenly so hooked on memory overclocking after 7years into the hobby and only using simple XMP 2.0.

Still tweaking with primary timings so far my favorite profile is 4400Mhz 17-17-17-34. The 4600Mhz 18-22-22-42 XMP 2.0 does post just fine, but does not feel as responsive. Have not even tried the entire 4500Mhz range OR the 4000Mhz super-low latency range so lots more yet to discover.

That's the only upgrade since January, although the Seasonic TX-700watt fanless announced at Computex is planned to launch this month August 2019, so there's that to look forward to. Will pick one up from Newegg.

___________

Some bad news, the Z390 Dark had some terrible issues, very weird intermittent behaviors and unexplained electrical phenomenon, actually had to get on the phone with EVGA and troubleshoot-RMA the board, my first time sending a motherboard back - 7years into the hobby.

The 2nd replacement Z390 Dark has been perfect, not a single hiccup, flawless activity, although the Z390 Dark boards really behave very strangely compared to the ROG boards, you just hold your breath a few times hoping the board is not DEAD and eventually after all the double memory training triple reboots, the board comes back to life - the boards, even when 100% healthy will simply "Go Dark" for 3-4, 5 full minutes, only a single white LED still lit up that will not power off even if you switch off the PSU, even if you unplug the pwr from the wall, it's absolutely unnerving man, this board is possessed.

Then the Z390 Dark finally powers off completely then boots normally into bios and a message appears on the main screen that says:

Why you sweating so much? I was just messin' with you man.

signed,

your dark friend

 

[Space Lynx]

I can always tell a Noctua heatsink's metal from other heatsinks. That is a good thing, LONG LIVE NOCTUA!!!

 

[E-curbi]

I can always tell a Noctua heatsink's metal from other heatsinks. That is a good thing, LONG LIVE NOCTUA!!!

I agree. 100%

I started way back in 2012 with a Noctua U12 single tower cooler, then a dual tower NH-D14 > Corsair H100 > Corsair H100i > (4) custom water loops using EK, AquaComputer, Monsoon, Bitspower Hardware Labs (with a Noctua NH-D15 for testing prior to completing the loops) then returned back to an NH-C14S for this build, worked so well I disassembled my last water loop of 2016 to 2018 last year and grabbed a 2nd NH-C14S. lol

Nothing short of astonishment, the cooling performance of the C14S. Reviews place it about 2C higher at load that the very bulky NH-D15. I had the D15, yet for this Batman open air build using no case fans at all for VRM memory chipset cooling, I really love the downfiring C14S's ability to also push at least a little air onto and over those components. Even a small flow of air makes a HUGE difference vs no air at all.

Some of my thermals recorded at full 100% load all 6cores at 5.5Ghz peaking out at 65C 67C a single core recording at 72C, I'm absolutely hooked on Noctua coolers especially the single tower single bank of fins on the C14S. I've tried to consider maybe the extremely long length of the phase change tubes, longer than any other Noctua sku might contribute to it's amazing thermal efficiency, but honestly I just don't know. For $74, the C14S is really an amazing value, imho.

Although I did swap out the maroon and beige stock 140mm fan in the box and replaced with a Noctua 140mm PWM Industrial which was close to $30 with tax, ouch. And that fan arrived with the dark brown corner noise absorption pads and swapped those out for the all black corner pads around $7.99 for (16) of them, enough for (2) fans. So yea total cost of the CPU cooling solution was closer to $115, but the Aquaero 6 really loves the Noctua Industrial black PWM fans, allowing almost infinite adjustability.

I never could get the Corsair H100 and H100i to go completely inaudible, the pump noise and vibration, that low frequency hum I could not isolate, even tried placing rubber o-rings at the base of my motherboard standoffs (most likely not electrically very wise) lol, that darn AIO hum drove me crazy. Did reach inaudibility with (4) the custom loops isolating the pumps with Sorbothane and the fans using various silicon pull-through mounts, and fans of light to medium total mass. Overall, I'd say the Noctua NH-C14S cooler with a 140mm fan attached using only a simple gravity mount is the simplest and easiest cooling subassembly to move into the inaudible range, by far.

So Go Noctua!

I have tried the newer Noctua thermal paste, the NT-H2 and found it runny, difficult to apply, and my thermals were horrendous. Maybe it was just me, but after 4 mountings on my 2ndary rig, I just gave up and went back to the NT-H1 paste, which I get amazing results with the "in line with die" application method.


This pic below is the Noctua NT-H1 paste "In line with die" application, then I add separate (4) dots of Thermal Grizzly Kryonaut (1) dot about 10mm from each corner and that's it.

My secret sauce.

Now the secret is out and everyone is going to have amazing thermals, just kidding.

Love it or hate it, right or wrong, I get solid thermal results at high clocks using this crazy off beat wacky eccentric mix of (2) thermal pastes from different manufacturers. lol

 

[E-curbi]

"Don't laugh too hard at my build, it's an open air test bench with a horizontally mounted motherboard, rolling around on four wheels and five radiators that don't actually do anything. lololol "

 

[E-curbi]

Also, anyone waiting for the Seasonic 700watt fanless to launch,

Seasonic Pushes the Fanless Envelope to 700W, Including 80 Plus Titanium

Each year we look forward to the next big fanless PSU. Seasonic did not disappoint this year, with the new Prime Fanless TX-700. This replaces the Prime Titanium Fanless 600W as the company's most powerful PSU of its kind. Its connectivity is updated to include two 4+4 pin EPS connectors, up to...

www.techpowerup.com

The titanium certification was posted only 48hrs ago at plugloadsolutions.com. So availability should be within the next 2-3weeks.

https://clearesult5.sharepoint.com/:b:/s/PLS/EfdiTYTNRHJFvSkqWY2HJZUBaP7h43WCJfqFoP377TIJhA?e=UZ2CSa

When the Prime Titanium 600watt fanless launched back in October 2017, Superbiiz.com had stock 3.5weeks before Newegg, Amazon etc, so I took a chance and ordered from Superbiiz.com. Got a good deal about $10 lower than the Newegg price, but no idea if that pattern will repeat with the 700watt Titanium fanless.

I had zero issues purchasing from Superbiiz, but this time with my build feeling more complete, feeling more patient, I'm planning on buying the TX-700 fanless from Newegg since the returns seem just as painless as Amazon nowadays and I really don't want to pay the Amazon 6.5-7% state sales tax - still no tax at Newegg.

 

[E-curbi]

He creates so much possibility.

Thank you simple Albert.

 

[E-curbi]

I keep returning to this website and wondering to myself.

PassMark CPU Benchmarks - Single Thread Performance

Benchmarks of the single thread performance of CPUs. This chart comparing CPUs single thread performance is made using thousands of PerformanceTest benchmark results and is updated daily.

www.cpubenchmark.net

Am I the new world record holder for PassMark CPU Single Thread Benchmark using my simple yet beloved Noctua air?

I mean the Performance Benchmark clearly states (within the testing window) that the World Maximum Score for Single Thread is 3333 and I just scored 3389 at 5.6Ghz soooo what gives, where's my World Champion overclocking trophy?

When my 8086K scored 3327 after a successful 5.5Ghz attempt I contacted - emailed the Siliconlottery guy, I was frantic saying "Hey brother, I'm about to break a World Record" "What voltage should I set for a 5.6Ghz run?"

He didn't respond.

Here's how the current CPUs scored at stock settings, this chart below, and the last screenshot is my 8086K breaking that 3333 World Record. I guess, lol Really? I don't know....

 

[tabascosauz]

Seasonic, Noctua and Caselabs in one place. Somebody sure knows the definition of quality and reliability (and how to spend money ).

I love this build. All hail Noctua.at. While you're out there pushing the boundaries of Noctua's 140mm big boys, I've been here pushing their small minions to the limit. U9B SE2, L9i, L9x65 and D9L. Trying to figure out what to do with the complete 4790K system that just got retired, so a new D9L or U12A may be in my future.

 

[E-curbi]

Seasonic, Noctua and Caselabs in one place. Somebody sure knows the definition of quality and reliability (and how to spend money ).

I love this build. All hail Noctua.at. While you're out there pushing the boundaries of Noctua's 140mm big boys, I've been here pushing their small minions to the limit. U9B SE2, L9i, L9x65 and D9L. Trying to figure out what to do with the complete 4790K system that just got retired, so a new D9L or U12A may be in my future.

I'd like to try some of those smaller extremely adorable pieces of brilliant engineering from Noctua, have a feeling they would do just fine at higher clock speeds. Really, it's all about delidding and a very thin die distance for the heat to travel and the pressure you add when mounting - giving that CPU exothermic energy an extremely time-efficient exit route. Whaddid I say?

Noctua

Having received more than 6000 awards and recommendations from leading hardware websites and magazines, Noctua's heatsinks are serving hundreds of thousands of satisfied customers around the globe.

noctua.at

So many amazing components within this hobby, unfortunately I only get to sample (and play with) a very small number.

When Noctua launches the successor to the C14S C-class cooler (seen at Computex 2018), I'll definitely try it out along with the NF-A12 superfan when it arrives in 140mm form factor and an all black industrial sku.

Also, have that extremely large Noctua passive cooler shown at Computex 2019 on my list for this build for 2020. My plan is to run it actively with a 140mm fan (just laying on top) turned way down into the inaudible range.

I mean, fanless vs a fan you cannot hear is the same experience right? You could argue added dust as a factor, but sonically - it's the same experience.

Sorry if I went off on a tangent, I apologize.

Thanks for stopping by, I really LOVE that you did.

 

[tabascosauz]

I'd like to try some of those smaller extremely adorable pieces of brilliant engineering from Noctua, have a feeling they would do just fine at higher clock speeds. Really, it's all about delidding and a very thin die distance for the heat to travel and the pressure you add when mounting - giving that CPU exothermic energy an extremely time-efficient exit route. Whaddid I say?

So many amazing components within this hobby, unfortunately I only get to sample (and play with) a very small number.

When Noctua launches the successor to the C14S C-class cooler (seen at Computex 2018), I'll definitely try it out along with the NF-A12 superfan when it arrives in 140mm form factor and an all black industrial sku.

Also, have that extremely large Noctua passive cooler shown at Computex 2019 on my list for this build for 2020. My plan is to run it actively with a 140mm fan (just laying on top) turned way down into the inaudible range.

I mean, fanless vs a fan you cannot hear is the same experience right? You could argue added dust as a factor, but sonically - it's the same experience.

Sorry if I went off on a tangent, I apologize.

Thanks for stopping by, I really LOVE that you did.

No, I love this build! It might not be the form factor for me, but it has all the quality I could ever want in one. It's too bad that Caselabs went out of business; I was looking to acquire a Mercury S3 (which I've been drooling over for 6 years) once I settled down from my highly nomadic lifestyle.

The 92mm towers from Noctua are decent, but with an extra fan (or even two in case of the D9L) they turn into very hard hitters. With 2 x B9s, the U9B/S rivals 120mm towers, and I reckon the D9L can beat them with a complement of 3 x A9s. The L9x65 is an admirably hard worker for its footprint, but there are barriers that cannot be overcome with just innovation. The L9i/a are just marginally better stock coolers, lmao.

The best things about Noctua are the attention to detail, fantastic customer service, and SecuFirm2. Everything builds on the same foundation, except the L9i/a.

I'm thinking might actually get my first 120mm Noctua with the U12A. That NF-A12 looks inspired by Gentle Typhoon and looks like a monster. I like my NF-F12, but I wouldn't say it's excellent at what it does.

 

[E-curbi]

No, I love this build! It might not be the form factor for me, but it has all the quality I could ever want in one. It's too bad that Caselabs went out of business; I was looking to acquire a Mercury S3 (which I've been drooling over for 6 years) once I settled down from my highly nomadic lifestyle.

The 92mm towers from Noctua are decent, but with an extra fan (or even two in case of the D9L) they turn into very hard hitters. With 2 x B9s, the U9B/S rivals 120mm towers, and I reckon the D9L can beat them with a complement of 3 x A9s. The L9x65 is an admirably hard worker for its footprint, but there are barriers that cannot be overcome with just innovation. The L9i/a are just marginally better stock coolers, lmao.

The best things about Noctua are the attention to detail, fantastic customer service, and SecuFirm2. Everything builds on the same foundation, except the L9i/a.

I'm thinking might actually get my first 120mm Noctua with the U12A. That NF-A12 looks inspired by Gentle Typhoon and looks like a monster. I like my NF-F12, but I wouldn't say it's excellent at what it does.

For this build, the Mercury S8 test bench was once a CaseLabs Mercury S8S with side and top panels, fully enclosed. I kept the bottom panel, the midchassis, the right side motherboard support and ATX motherboard tray and added the test bench shorter front and rear panels in gunmetal and test bench specific front cover in black, only three additional parts.

CaseLabs used to sell a conversion kit from Mercury S8 or S8S to test bench for $79.99. But that was back in 2014, and by the time of my conversion in 2017, they didn't even list the test bench parts any longer, had to pull the schematic out of mothballs just to make three panels for me, lol, and they charged me significantly more than $79 in doing so, since at that point the test benches were for all practical purposes out of production. When they were in production, well I got my first Mercury S8 test bench in all gunmetal in 2015, and it was only $225 complete, a really great value I thought at the time, still using that chassis today as my secondary work rig - research build.

Yea, I miss CaseLabs so much, really wish they would return to the industry. All my CL parts look brand new, exactly like the day I unboxed them, such amazing quality... a few scratches near the PCIe card slot area from AMD and Intel and ASUS add-in-cards not fitting so well, you know, that wedge-effect with applied force, but overall the powdercoat has really held up in pristine condition.

The fan blades on the NF-A12 do look like a Gentle Typhoon facsimile and I believe Noctua's technical selling point is the extremely low blade gap between blade and frame housing for that sku. Again, I'm wanting to try it out in 140mm industrial all black, until then I'll keep waiting patiently, Noctua parts seem to really take quite a long time to move from prototype to production.

I think I could boot at 5.5Ghz 5.6Ghz all core all thread with no issues using some of the smaller Noctua coolers you mention, surface area is important of course, but thermal efficiency from die to cold plate I believe plays a larger role. Of course, I could not run a heavily stressing benchmark for any length of time with those smaller coolers without hitting 100C, but I still believe I could boot at high clocks and run my work apps with no issues at all, only slightly higher CPU temps.

Maybe I should order one and run some tests, you know, for the sake of science and boundless curiosity. lol

 

[tabascosauz]

Yea, I miss CaseLabs so much, really wish they would return to the industry. All my CL parts look brand new, exactly like the day I unboxed them, such amazing quality... a few scratches near the PCIe card slot area from AMD and Intel and ASUS add-in-cards not fitting so well, you know, that wedge-effect with applied force, but overall the powdercoat has really held up in pristine condition.

The fan blades on the NF-A12 do look like a Gentle Typhoon facsimile and I believe Noctua's technical selling point is the extremely low blade gap between blade and frame housing for that sku. Again, I'm wanting to try it out in 140mm industrial all black, until then I'll keep waiting patiently, Noctua parts seem to really take quite a long time to move from prototype to production.

Maybe I should order one and run some tests, you know, for the sake of science and boundless curiosity. lol

Ive been kicking myself for not buying an S3. Back then it was barely more than $200 with a couple of options like USB, a window, etc. Now our CAD dollar is about as strong as a Subaru head gasket, Caselabs no longer exists, and I haven't found a substitute for them in terms of quality and design. I absolutely loved the design of the S3, but fortune had it that I needed something sub-15L for pairing with a Pelican 1510.

Call me peculiar but I still prefer the original palette fans. There's something about the beige that doesn't particularly show dust. Back in the day, original Noctua fans did look pretty hideous when populating a case alongside blue motherboards and red graphics cards, but nowadays as everything trends towards matte black I find it a pretty good look, actually, especially with some white/grey/light blue/other light colour Cablemod cables.

Honestly, I wouldn't be surprised if you could manage with a L9x65, as above everything else, that cooler is wholly dependent on case airflow. But save for cases like the SG05 or DAN-A4 where aren't really other options, it's not a particularly good value proposition. A more boujee Wraith Max, if you will.

You can get a U9S for not much more, though. That's a potent one.

I hear you on the long R&D times. A necessary consequence of a thoroughly tested, quality product, methinks. I'll still take it over the half-hearted, half-baked, RGB infested garbage that's all too common these days.

 

[E-curbi]

Ordered the Noctua NH-L9x65 65mm CPU cooler since it's SecuFirm2 mounting is compatible with my current Noctua kit, so an easy and quick install.

My hypothesis is: The tiny 65mm cooler will allow for an ideal boot and stable working environment for my 8086K at 5.5Ghz and 5.6Ghz 6cores 12threads activated and using current working applications, although CPU thermals will be quite exaggerated beyond base and normal.

Will my hypothesis hold water? I'll find out on Tuesday I guess. lol Maybe it will boot, maybe it won't. I need to find out, since it will eventually prove how much of a factor surface area really is when booting into higher clocks. From my pov right now, I'm thinking not so much, and silicon efficiency die thickness thermal paste application delidding - that tight thermal conduit from die to cold plate play a much more important role in successful booting vs blue screen.

Thank you so much for the amazing and beautiful insight tabascosauz bro!

I will post results of the tiny Noctua cooler as soon as testing (possible failure) is complete.

 

[phanbuey]

Ordered the Noctua NH-L9x65 65mm CPU cooler since it's SecuFirm2 mounting is compatible with my current Noctua kit, so an easy and quick install.

My hypothesis is: The tiny 65mm cooler will allow for an ideal boot and stable working environment for my 8086K at 5.5Ghz and 5.6Ghz 6cores 12threads activated and using current working applications, although CPU thermals will be quite exaggerated beyond base and normal.

Will my hypothesis hold water? I'll find out on Tuesday I guess. lol Maybe it will boot, maybe it won't. I need to find out, since it will eventually prove how much of a factor surface area really is when booting into higher clocks. From my pov right now, I'm thinking not so much, and silicon efficiency die thickness thermal paste application delidding - that tight thermal conduit from die to cold plate play a much more important role in successful booting vs blue screen.

Thank you so much for the amazing and beautiful insight tabascosauz bro!

I will post results of the tiny Noctua cooler as soon as testing (possible failure) is complete.

View attachment 129490

It will. But my hypothesis is that all of your scores will increase if you drop volts/clocks.

 

[tabascosauz]

Ordered the Noctua NH-L9x65 65mm CPU cooler since it's SecuFirm2 mounting is compatible with my current Noctua kit, so an easy and quick install.

My hypothesis is: The tiny 65mm cooler will allow for an ideal boot and stable working environment for my 8086K at 5.5Ghz and 5.6Ghz 6cores 12threads activated and using current working applications, although CPU thermals will be quite exaggerated beyond base and normal.

Will my hypothesis hold water? I'll find out on Tuesday I guess. lol Maybe it will boot, maybe it won't. I need to find out, since it will eventually prove how much of a factor surface area really is when booting into higher clocks. From my pov right now, I'm thinking not so much, and silicon efficiency die thickness thermal paste application delidding - that tight thermal conduit from die to cold plate play a much more important role in successful booting vs blue screen.

Thank you so much for the amazing and beautiful insight tabascosauz bro!

I will post results of the tiny Noctua cooler as soon as testing (possible failure) is complete.

If nothing else, it's a really good looking piece of kit, if I may say so myself if it proves insufficient to tame the beast, maybe you'll have something that you can put on an HTPC in the future. It's a lot of heat to send through that poor A9x14. My 4790K first started to have to come down to 3.5GHz when I first moved into the SG05 and the L9x65, and stayed that way until the day before yesterday when Matisse took its place. Could it have done it at stock clocks in a confined space? Probably. I didn't like seeing potential temps in the 80-90s though.

Think of it as the Wraith Prism for Intel (though I hate that because the L9x65 got here first, gtfo AMD trying to claim credit! ). Will handle everything except the top end chips that are a bit too ambitious for their own good.

Ambient temperature even helps a little when it comes to small coolers that have to work hard. Maybe keep the air a bit cooler? Should be easy as we're coming to the end of this god awful summer.

 

[E-curbi]

If nothing else, it's a really good looking piece of kit, if I may say so myself if it proves insufficient to tame the beast, maybe you'll have something that you can put on an HTPC in the future. It's a lot of heat to send through that poor A9x14. My 4790K first started to have to come down to 3.5GHz when I first moved into the SG05 and the L9x65, and stayed that way until the day before yesterday when Matisse took its place. Could it have done it at stock clocks in a confined space? Probably. I didn't like seeing potential temps in the 80-90s though.

Think of it as the Wraith Prism for Intel (though I hate that because the L9x65 got here first, gtfo AMD trying to claim credit! ). Will handle everything except the top end chips that are a bit too ambitious for their own good.

Ambient temperature even helps a little when it comes to small coolers that have to work hard. Maybe keep the air a bit cooler? Should be easy as we're coming to the end of this god awful summer.

The L9x65 just arrived, and it's adorable.

I'll find a good use for it, umm right after I discolor the shiny nickel-chromium plating during my testing, just kidding - I won't let it get that hot.

May not get to mount it today, but definitely this week, I'll boot first into the Daily OC (5.4Ghz 6c12t 1.370v 4400Mhz 17-17-17-34 1.45v) see how this little guy works out before turning up the heat.

(Left the protective plastic cover on the coldplate for these pics. I'm careful about handling nice parts)

 

[E-curbi]

A cool dood over at the ROG forum says my ring bus of 45 is "dog arse slow".

He also says he's got a 8086K Apex XI Cinebench R15 Single Thread score of 245, which betters my score by 2points. Of course he had no screenshots to show but I was planning a few 5.6Ghz attempts anyway, and I believe I can hit his 245 at 5.5Ghz, so has the gauntlet dropped?

He's right, I was/am running my ring bus at 45 way low from a Daily Work OC profile, I'll increase to 49, and rerun AIDA 64 memory latency benchmark.

...all in the spirit of performance, and good faith.

____________

Things I've learned from mistakes I've made:

-A really good idea for anyone just into overclocking for fun and a nice daily performance improvement: When going for new attempts, do not use your daily work and/or gaming OS drive. Set up Win 10 on another alternate drive, since AIDA 64 runs and Cinebench 64 runs (from my experience) at higher clocks and high memory Mhz - tight timings can add/leave artifacts and corruptions in your OS and spending time reformatting your daily drive gets old very fast - I learned the hard way. lol

I remove the Optane 900P work OS SSD from slot number 4 chassis number 7 and use a tiny 32GB Optane M.2 SSD for all new overclocking past and beyond any known stable daily.

-Another good idea is remove any other internally mounted storage devices unplug the pwr and data cables or dismount from hot swap bays from a mobile rack, along with unnecessary PCIe Add In Cards, sound cards, capture cards, Thunderbolt 3 external storage cards, so they're not part of your boot equation - fan controllers should remain connected.

-And always use a CPU cooler with an adequate and large amount of physical heat dissipation and surface area.

Example below on the right:

Is that joke getting old, or is it just getting started?

 

[E-curbi]

The Noctua smaller coolers - at least this L9x65, did not arrive with a PWM extender cable, only an in-line signal attenuator, what Noctua calls a low-noise adapter. Won't be needing that since connecting to an Aquaero 6 LT fan controller, grabbed a Noctua 4-pin PWM extender cable from my parts drawer.

It's difficult - impossible to tune the exact rpms when a fan goes inaudible to your ears at normal sitting distance using the low-noise adapters, they basically only give you a 2nd rpm option other than 100%. Not to say they're not useful, my first two or three Noctua coolers before getting a Lamptron 4-fan controller, I daisy chained 5-6 fans directly off the PSU using molex adapters and the Noctua low-noise adapters. Do they grant you infinite granular control one-tenth of a percent adjustability? Well no, but they do a good job, huge improvement towards a sonically comfortable computing experience. I never got into motherboard fan control, tried it briefly the ROG software of 2012 (Maximus V Extreme) was so buggy, you attempt to adjust a single fan and all fans connected would increase rotational speed, and that's when it worked at all.

 

[E-curbi]

The tiny Noctua L9x65 mounts into the socket area like a glove.

Even in an ugly build like mine, nice parts are still so sexy - individually.

 

[E-curbi]

Booted right into the daily oc profile at 5.4Ghz 6c12t 4400Mhz 17-17-17-34, ran my work apps no issues at all, only higher CPU temps at idle 35C 36C 37C (with the C14S cooler 30C 31C 32C, so about 5C higher) and two cores at 57C 58C maximum when working in apps.

All with the very tiny 65mm PWM fan set at 50% pwr - 1315rpms, which is not inaudible, yet barely noticable.

In the Aquasuite software, Fan 1 is the 40mmx25mm Sanyo Denki mounted inside the Supermicro 8-bay 2.5inch SSD moblie rack, it's rated at 12,000rpm yet goes inaudible around 1000rpm, not truly needed since I'm only running (4) SATA SSDs inside the unit they need no active cooling, but had an extra port open on the Aquaero 6 so what the heck cannot hear it at 1000rpm and it adds a very slight air flow over the SSDs pulling front to rear. Fan 2 is the 65mm Noctua brown and beige fan running as above at 50% - 1315rpms.

Will it boot at 5.5Ghz and 5.6Ghz? I'll make some attempts after lunch.

 

[E-curbi]

Now things are warming up at 5.5Ghz 6c12t, still no thermal emergencies.

Thinking I should run at least one stress test, something mild to moderate not Prime 95 or AVX, yet at least pushing the cores to 100% for a brief period and recording temps.

 

[Deleted member 178884]

Now things are warming up at 5.5Ghz 6c12t, still no thermal emergencies.

Amazing CPU from silicon lottery there, it's always amazing me haha.

 

[E-curbi]

Amazing CPU from silicon lottery there, it's always amazing me haha.

Not So Fast!

Hypothesis FAILED!

Surface area does matter when simply booting into higher clock speeds.

-Attempted boot at 5.6Ghz 6c12t 1.495v 4600Mhz 17-17-17-34 > bluescreen and corrupted bios #3.
-Switched to bios #2 set to optimized defaults and only set up the CPU OC, no memory OC > bluescreen again yet bios #2 still usable.
-Disabled CPU Hyperthreading attempted to boot > blackscreen with Z390 Dark post code set to CPU temp reading 66C. (most likely 66.6C)

Estimating 66C is just too high a core temperature to boot...

I can boot at 5.6Ghz 6c12t 1.495v 4600Mhz 17-17-17-34 with hyperthreading using the larger Noctua C14S cooler, and run/complete single-core benchmarks, so there ya go.

Surface area is important after all, not only for using/stressing the processor, but also for the initial boot into OS.

Now, have to go back and re-flash my bios(s), crawls away with his tail between his legs. lol