Mr. Fox

He will need to confirm that the problem is resolved, but I believe moving the chiller or tank higher than his PC fixed the issue. Before adding the tank his 300mm reservoir was the high point. Hopefully so, as it was understandably frustrating.

So, my second 13900KS finally arrived from Germany. It took like 16 days to get here using "priority mail" LOL. E-core SP rating is much higher and P-core VIDs are much lower even though the SP-rating on the P-cores is the same. MC rating is a bit lower, but I doubt that is going to affect anything. I am running it at 8600 with the same memory OC profile and I think the MC SP rating is the same as @tps3443CPU so it should be fine. I will delid it either tomorrow or next weekend. I am running the Optimus block on it for now.

I think I see the problem. It looks like the tubing leading into your CPU block is the highest point in the loop. The point of air accumulation will always end up being the highest point, because air rises to the surface. I don't think you'll ever be able to correct that until the lines leading into the chiller or the top of the water tank end up being elevated above the highest point in the loop so all air accumulates there. In order for it to work the way it is right now it would have to defy the laws of physics. Try reinstalling your pump and reservoir combo on the side of the bench again so that becomes the high point. Only have one pump between the tank and PC and put most of the pump power on the return side of the loop. You can stand the tank on end so the lines are in and out of the top the same as the chiller and that will eliminate any air inside the tank. This will force all of the air to go up to the top of the reservoir. Another alternative would be to stand the tank on the end next to the PC on your desk so that the lines are going in and out of the top of the tank. That will become the highest point in the loop. Put a tube on the outlet fitting inside of the tank and thread it into the inside of the brass fitting like an oil sump pickup tube on your car engine so that the line leaving the tank has the pickup tube always submerged in water.

All of them do really sleazy things. Before we purchased our current home, we owned a house about 1.5 miles from here in a different subdivision. There was no exclusive cable provider kickback scam with Cox like our current HOA. I had a great package with Cox and after a year they decided to jack the price to double, so I switched to Mediacom and got my cable internet at twice the speed and the same price as Cox before they doubled the price. Before making the switch I called the Cox customer-retention department and told them what I was about to do with Mediacom in case they wanted to try and retain my business. The idiot "helping" me just kept reciting a mantra, "I'm sorry Mr. Fox, your package was a time-limited offer and a cannot reinstate it." So, I suggested she find a new line of work because she was failing at her job in customer retention. I pulled the plug and moved to Mediacom. I never looked back and Mediacom was great to deal with. Fast-forward about a year later, we sold that house, purchased this one and when I tried to move my cable internet with Mediacom to the new address they told me "Unfortunately, we are not allowed to provide service to your subdivision due to a contract between Cox and the HOA" and I was like WTH? That's just so wrong and messed up. So, back to Cox with a fantastic deal. Everything was great. It worked well and the price was good. It expired a year later, the price doubled and they again refused to "retain me" after contact with their customer-retention department. Same mantra, different zombie customer-retention rep. So, I switched to T-Mobile 5G and it was horrible here due to a poor signal. It is awesome on my phone, but the home internet was poor at my address. One mile up the road is a strong 500 Mbps down, 50 Mbps up connection. At my home address there was a frequent signal loss, even with an outdoor 5G antenna. At that time, Centurylink 20 Mbps DSL was the only available alternative, so not an acceptable one. I struggled with T-Mobile 5G for 2 months and Cox ran an ad for a special package for Gigabit internet with unlimited data for $50/month. I tried to sign up and they said, "Sorry, this is for new customers only, it has been less than a year since you left so you're not a 'new' customer" so I had my wife call Cox and use her name. They gave it to us. Same last name, same address... she was considered a "new customer" even though I was not, LOL. Stupid. One year later the price goes to $70. I cringed, but bit the bullet due to no acceptable alternative. One year after that (last month) the price goes from $70 to $108... for internet only. Insane and not worth it. Centurylink began putting in fiber lines 6 months ago and the service just went live last month. I have been on the waiting list for 3 months (when they started taking pre-orders) so it happened just in time. Less than a month of Cox service at the jacked up $108 Mafia scalper rate. Buh-bye Cox. Their name is a fitting and accurate facsimile of the vulgar word describing the part of the male anatomy that is phonetically the same pronunciation. I do Netflix, Paramount+, Discovery+ and all of the free streaming channels. We use our kids accounts for a couple of additional streaming services and they use ours for a couple. So far none of the providers have stopped it. I cancelled Disney+ when they suddenly got extra-stupid with their woke cancel-culture nonsense. I think it was the campaign against Gina Carano when I told them to go mate with themselves. Now that they are remaking all of their movies with politically-correct liberation theology/equality white privilege propaganda I will never subscribe again and plan to boycott all Disney movies going forward. I canceled my Planet Fitness membership at the same time and moved to EOS because Planet started belligerently spamming my phone and email multiple times per day with gay crap, and plastering their walls at the gym with disgusting propaganda posters during "Pride Month" and I told them to shove their woke garbage up their limp-wristed lover's tailpipe. I think I will use the money saved on the fiber internet and cancelled woke businesses to level-up on my Second Amendment load-out and stockpile some extra ammo... just in case, LOL. Is your tank higher than the chiller? If the air in the tank is not the highest point in the loop that could be somehow contributing to the issue. It will be interesting to see if I run into a similar situation when I reconfigure my system with the water tank. Anything is possible, but I have been using QD3 fittings for years and haven't experience them allowing air into my loop or leaking while they are connected. I have had the female fittings stick open after disconnecting them, but that was due to the black anodize finish on the male part getting scratched and scraped and the particles from that jamming up the female portion. I stopped buying black and have been replacing them with chrome and the chrome fittings never stick open like the black ones. I think it may be something related to location of the pumps in the loop or something else related to the addition of the tank if that is when it started. Finding the source of something like this can be tedious and frustrating. Have the tiny bubbles affected your temperatures? If not and everything is working correctly other than the visible bubbles, then it might not be something to be overly worried about.

The Aquacomputer Cuplex Kryos Next water block ís also temperature only. Yours might be the same. I was disappointed to discover that as well. Caveat emptor. The version with both is a lot more expensive and grossly overpriced, poor value. You can get a TT TF2 and enjoy both with no need for crappy software. It works even with no operating system, and starts displaying data as soon as you press the power button. It has an alarm that can be adjusted to go off with a user-defined limit on flow rate and that can be set with buttons on the device. It has no features that rely on the presence of an operating system, or a USB header, which makes it awesome. I suspect the same thing is going to happen here. Cox has had a long-time monopoly with no competition other than Mediacom, whom they screwed, along with would-be subscribers, by doing sleazy exclusive-access cable vendor schemes with various housing developers and homeowner's associations. Back in the day when their Mafia deals were struck, Centurylink/Qwest offered only crummy landline, dial-up and DSL. Now that Centurylink (Quantum Fiber) is a threat, and T-Mobile 5G Home Internet are also helping to bypass their monopoly, they are starting to freak out. It makes me very happy to see it taking place. Not only do they no longer have the fastest options available, they now have to compete with a no data limit provider that is offering a no-contract product that is faster and $30-$70/month cheaper, which varies depending on your street address. Those who were blocked from choosing Mediacom due to sleazy business deals will enjoy even greater financial benefits. I love seeing ruthless and brutal justice executed upon the wicked with such merciless precision. I will be giving Cox the nasty-finger eat poop and die salute tomorrow. An interesting challenge with the fiber was that my upload was only like 38 Gbps (same as Cox cable was) and fiber is supposed to be 1Gbps both up and down. The problem turned out to be my Linksys WRT32X router. On WiFi 6 and wireless AC I was getting proper speeds up and down, but wired LAN upload was not. Connecting the Cat6 cable direct to the fiber modem the speeds were correct. So, I ditched the router and will use the fiber modem's internal router for that rather than using it in bridge mode. I replaced it with a 8-port switch and *BOOM* Gigabit both ways. So, the router was the bottleneck on upload speed and I am not sure why. The WRT32X has an internal speed test feature and it was correctly showing up/down both at Gigabit speeds, but the LAN ports were limiting upload to 38 Mbps maximum for some reason. Installing the switch in place of the router was an instant fix. The TP-Link 8-port Gigabit switch was only $21, LOL. The 1 Gbps upload is going to make uploading a large YouTube video a lot faster than <40 Mbps was. At least I would think it should be a whole lost faster.

I am amongst the first to get it so it might go down a bit once more users are consuming bandwidth.

Got new internet... finally... long overdue competition for Cox Cable. It was the only option other than crappy 20 Mbps DSL, but now we have fiber. Same price as Cox Cable Gigabit, but faster both ways. Quantum Fiber is the new ISP.

Same. I just want white, and only white. If the companies that sold RGB products would make static white the default and let those that want something else install software the world would be a much better place. The default rainbow puke is disgusting and makes me hate buying products with RGB. OpenRGB works fine for things other than RAM. I use it for my keyboard and GPU without issue. Beware of using it for memory RGB. Should be fine for peripherals. It works great for RAM as well... until it doesn't. There's no easy way to fix it since Thaiphoon Burner does not work for DDR5. I had Vitality Jungle (Thaiphoon Burner developer) confirm the SPD corruption. He could fix it, but there is no way to flash it correctly. The unfortunate part is that OpenRGB can accidentally write to areas of the memory that have nothing to do with the RGB space on the memory modules. The safest option for memory RGB is to use the RAM manufacturer's RGB software. G.SKILL's RGB utility works for TeamGroup and other brands of memory. But, it doesn't permanently save the changes to the memory modules. It applies the color scheme only while running, which is why it is safe. Nothing gets written to the memory modules.

OpenRGB bricked three RGB memory kits for me. Two TeamGroup and one G.SKILL. It is a known issue. One stick the SPD was completely wiped and the second and third kits had the SPD corrupted so that some timings were erased, the product name and brand were corrupted and did not display correctly in AIDA64 or CPU-Z. It is a dangerous option for DDR5 users. If you enable SPD write in the BIOS almost guaranteed you will end up bricked, but it happened to me with it disabled twice. The first time when it wiped the SPD it was enabled. OpenRGB says it is the memory manufacturers fault for not protecting the modules. Might be the case, and I believe it probably is, but it is still the end user that ends up getting screwed and it directly resulted from using OpenRGB.

Look in your BIOS for both MyASUS and Armory Crate. If you find them, disable both of them. If you need to use Armory Crate for your laptop for crap like RGB and fan control, you can install it. It DOES NOT need to be enabled in the BIOS to install it. MyASUS is the automatic BIOS update filth and if Armory Crate is enabled in the BIOS you will be prompted to install it when installing Windows. It will install LAN/WiFi drivers from the UEFI and connect to the internet without user intervention. @D2ultima - screenshot If you do not have these options available in the BIOS, they are likely enabled by default with no option to turn them off. Especially likely with a turdbook. You never really own a modern laptop. You merely rent it and the OEM and Redmond Retards decide what happens to it.

The MyAsus toggle in the BIOS allows the motherboard to identify itself as open to firmware updates from Micro$lop. This is common on modern turdbooks and it is disgusting that we see this dung surfacing in desktops. I guess we shouldn't have surprised looks on our faces considering the mentality of the zombie sheeple herd. Armory Crate is the ASUS equivalent of garbage like Clevo Control Center, MSI Dragon Center, etc. but the ASUS software is the most extreme bloatware garbage I have ever seen. It installs around a dozen services and TSR applications, about 1,000 registry keys and has a severely negative impact on system resources. It is frequently broken and often will not even function. ASUS has bundled Aura into it and you cannot control your RGB without installing the entire malware payload. You can use OpenRGB, but that places you at risk of the memory SPD getting corrupted if you have RGB memory. Find both of those automated defecation switches in your BIOS and set them to disabled. They are on the same menu as EZ-Flash and BIOS user profile management if memory serves me correctly. I will grab a screenshot for you if you need it. Bummer. I did not know that. When you look at how many Strix motherboards exist it is very confusing and difficult to fully understand what kind of compromises come with a lower price tag. ASUS is not very transparent with their specifications and feature versus compromise comparison among the crazy number of options. They probably assume, and it may be accurate, that most of the kiddos buying their crap shop based on price only and don't really care about things they do not understand. There is definitely truth to the cliché that ignorance is bliss.

You should not be allowing Windows to know whether or not the firmware update is available. That is none of their business and access should be blocked for the best outcome. I would recommend disabling the BIOS options for MyAsus and Armory Crate. I do not recommend updating the BIOS. There are numerous examples of people who are not happy with the results. Memory overclocking has been impaired for some people with the latest cancer firmware updates on their Asus motherboards. You'll probably have to downgrade to an older firmware to regain the loss of function brought through updates. @Falkentyne cannot even boot his system with memory overclock settings previously stable across the previous 5 or 6 firmware versions. Many of the so-called exploit mitigations impair functionality and hinder performance. You are wise to use the approach of not "fixing" anything that is not broken. There has never been a security patch or exploit mitigation worth sacrificing performance or function. Asus actively blocks Intel ME downgrades and ME updates should be avoided. Intel ME provides zero consumer functionality and its existence is a vulnerability. Asus *should* provide an option in the BIOS to disable Intel ME but they do not. It is a built-in backdoor point of entry allowing remote system access for "business" (government and hacker access) that should not even exist on consumer motherboards.

That's strange. The Strix Z690-E that I used to have had it. Maybe they removed it in a BIOS update downgrade. You're correct in doubting it. Sapphire was an ATi AIB before they were purchased by AMD back in the day before NVIDIA displaced them as the world's premier GPU ODM. My last ATi (pre-AMD) Radeon All-in-Wonder Pro and Rage Pro (circa 1996-1997) were made by Sapphire. They have always been ATi/AMD-only. PCIe did not even exist back then. The cutting edge GPU slot back then was AGP and the VRAM was measured in MB instead of GB.

All is does is reduce fluctuation of BCLK. Most modern desktop motherboards have a search feature where you can search for key word and it will show all settings with the resulting matches. If you search for "spread" it will show it to you and if you click on the item at the bottom of the screen it will tell you the path. In this case, you will find it in your ASUS BIOS on the Extreme Tweaker page under Tweaker's Paradise. There are three Spread Spectrums and they all do basically the same thing (reduce fluctuation in frequency to make it more static). You want BCLK Spread Spectrum. There is also a VRM and PCIe/DMI Spread Spectrum. Having them enabled is intended to reduce electromagnetic interference. I think the worst thing that might happen is you could encounter for very minor audio artifacts on your speakers, but even that is not very likely. @Premaoften sets it to disabled or 0% in his Clevo BIOS mods. Some of the modern turdbooks do not allow modifying it, even with a professional firmware mod.

Look in your BIOS and find the BCLK Spread Spectrum setting and disable it. That will mostly eliminate it and setting it for like 100.05 will take care of the 0.05 MHz variance below your multiplier. If you can't locate it, let me know and I can tell you what screen it is on. That is also kind of dangerous. If it was bloating the tank like that you run the risk of one of the hoses getting pushed out of the compression fitting. That's too much pressure and the pressure is not beneficial for cooling. I've actually had that happen before and it made a huge mess. Luckily it did not short any of my parts or cause any damage to anything other than my laminate flooring that got drenched in water. You can actually purchase a pressure vent for a custom loop system to keep that from happening.

I misunderstood that you were getting air in the lines and not additional air accumulating in the tank. In that case it is probably sucking air in at some location. I once had some tubing that was getting pulled out of the compression fitting and it was making bubbles sucking air in and not leaking water out. It was strange. I did not notice the tubing until I started looking for the source of the bubbles. If that is happening it is most likely at the inlet side of one of the pumps. If it were on the outlet side you would be dripping water. The reason it would not leak water is likely because the tank fill port is open and no pressure is building. If you were to plug the fill port you would likely start to see a water leak at the point it is taking on air because it would pressurize the system with with the fill port being closed. If you can find it and stop it, you are better off without the loop being pressurized. It is less likely to encounter any coolant leaks if the system is not pressurized.

Air can be pulled in if there is a leak on the suction side if your pumps are arranged in a manner that creates a high side and low side like an HVAC system. The only way you can continue taking on air is if the water is leaking out somewhere. The only explanation I can think of is that the air ended up in the tank from other places in the loop. If you keep adding more water there will be no space for the air once all of the water displaces all of the air. That being said, water in the tank or reservoir is of no consequence and that is where you would want it to end up. The only place the air is a problem is inside of a waterblock. If the tank is the highest point in the loop, it would be totally normal for the air to end up in the tank. It should stay there. Whenever I drain one of my loops it takes topping it off several times until no more air appears in the reservoirs. I can fill them to the brim and within a day there is an inch or so of air, and I have to repeat that 2 or 3 times or the course of as many days before no more air appears in the reservoirs. Yes, that is exactly right. The best cooling system in the world is limited by the ability of the water block to capture and remove the heat from the processor (CPU or GPU). Increasing flow rate won't help if the water block is already pulling the most it is capable of pulling. Same thing with air cooling. Cranking the fan up to highest possible speeds and moving more air over the radiator fins won't do anything but create extra noise if the cold plate and heat pipes have reached their peak efficiency and ability to wick the heat from the processor and radiate that heat through the fins. In either case you reach a plateau where nothing changes and the only additional steps are using different TIM and using colder air or colder coolant.

My tank, insulation, additional tubing and associated parts will arrive on Monday. The 13900KS I bought from the guy in Germany should be here on Monday as well. It was supposed to be delivered today, but it just sat in New Jersey at US Customs with USPS for four days with no movement. Still no response from Supercool on when/if the missing parts shipped, so I have no idea if they have. Initial response was quick, but I made two requests for a tracking number and got nothing.

And, that was with screen capture. Here is how it looks without the overhead at 4K.

Many things affect flow rate. I have 9 QDC fittings and multiple elbows. This reduces flow rate. So does the amount of tubing in a loop. Likewise, the number of components the coolant has to pass through affects flow rate. Using a manifold or distribution block can help because the flow of coolant occurs in parallel to each component. In this scenario the flow rate of coolant through one component will have little or no affect on the other component(s) because the flow to each is handled independently. The flow rate of one component is not hindered by other components in the series because the coolant doesn't have to pass through one part to reach the next. There could be some impact from turbulence, but it would have to be enough turbulence to impede flow rate to the point of becoming detrimental. Flow rate is important, but so is having a system that is convenient to tinker with. I would prefer to add more pumps to keep flow rate up with the extra QDC fittings than to have to frequently drain and refill my system every time I wish to make a simple change to something in my configuration. You will also reach a point that it doesn't matter any more. In general, if the flow rate is around 175-200 L/H it will perform the same as if the flow rate is 300 L/H. As long as the coolant is moving fast enough to carry the water out of the block before it can heat up and the water in the block is swiftly displaced by the cooler incoming water there is no longer a benefit to higher flow rate. Coolant temperature at that point become more important than flow rate. Coolant volume becomes more important because it takes longer to warm the higher volume of coolant present in the loop. It also takes longer for a higher volume to cool down once it has been fully warmed. You could look at it in much the same way as having a fan blowing on a component to keep it cool. At a certain point making the fan runs faster and push air more accomplishes nothing in terms of cooling the component. At the point of maximum benefit, the only way temperatures can improve is if the air the fan is blowing is colder than ambient temperature. This is the point where moving more air at a higher velocity provides no benefit in terms of the component getting cooler. If the component has reached a normalized temperature it will not go lower than the normalized temperature as it relates to ambient temperature. That is why using chilled air is more effective than pushing more air. It is the same reason that using chilled water works when using more water is no longer beneficial.

I have one of these on each desktop and really like them. Half the cost of what you were looking at. https://www.amazon.com/Thermaltake-Construction-Monitoring-Temperature-CL-W275-CU00SW/dp/B08B3S7L3Q/

Yes. That is one very heavy thermal pad. I think some people would complain about a notebook weighing as much, LOL. Even before GPUs weighed as much as a newborn child, it was very foolish to not support one with a brace when the installation is conventional and gravity is putting stress on the PCIe connector. It is also very bad for the motherboard, not only the GPU. Even using the expanding foam bags inside of a chassis when shipping a pre-built is risky. Better off shipping the GPU in a separate box. But, that would cause some people a great deal of stress because they don't even own a screwdriver and couldn't find their head if it were not permanently attached to their neck. I think it is also kind of risky to have a GPU installed in a manner where it is suspended by the I/O bracket unless it is properly braced with a bracket attached to the chassis. The PCB is not designed to be under lateral stress. Cracking is almost inevitable if there is an upset of the chassis. Horizontal motherboard trays and vertical GPU mounts in a conventional vertical motherboard installation solve most of these problems.

A warranty is only as good as the company looking for lame excuses to shirk their ethical duties. The only one that had anything that resembled ethics and cared about customer experience no longer makes or sells GPUs and motherboards. "The law" is never enough to make bad people or bad companies good, because they are fundamentally bad, dishonest and lawless.

Gigabyte 'quality' with an ASUS price tag. Nice. *sarcasm* Pay more, get less. Gee, thanks ASUS. Interesting comments at 21:50 regarding chiplet design and the associated engineering flaws. Same underlying evil intentions as BGA dung.

Welcome, brother! Glad you joined us. We really missed you, bud. Looking forward to seeing you here often.