New AMD Video on Updated PBO for Ryzen 3000

Pretty short, and light on details, but it is interesting to get at least an inkling of what is coming:



FrgMstr seemed pretty impressed with the last gen of PBO when he reviewed it on the Threadripper and this gen seems improved!

Robert Hallock has quite the mirror imaged drawing skills!


But what he explained was already the same for each version of PBO , what he neglected to explain is how X570 adds another layer nor did he explain if there is a difference between Ryzen 2K and 3K series ...
 
But what he explained was already the same for each version of PBO , what he neglected to explain is how X570 adds another layer nor did he explain if there is a difference between Ryzen 2K and 3K series ...


Yeah, as I said, details are light.

Correct me if I am wrong though, but wasn't this previously only a Threadripper feature? Specifically a second gen Threadripper feature. Now it's coming to Ryzen 3000 series chips.

I knew that PBO used various aspects of chip health to determine overclocking capacity and dynamically and did so pretty well, but I did not know it had a data link to the motherboard to also probe and monitor the VRM's. This was new to me.
 
Last edited:
That had always been a part of PBO.

Thanks for clarifying that.

Well in that case I agree with previous posters. This isn't really any new info.

In the past I never paid much attention to the VRM's when motherboard shopping. I just assumed that if I was buying a quality high end motherboard, it would come with good power regulation. Maybe I should pay more attention going forward.

This blog entry suggests motherboard manufacturers are trying to trick users by putting fake chokes and other components on the boards to make them look like they have more phases than they actually do.

This suggests that access to quality hardware reviews are even more important than ever.

My favorite x570 motherboard this far is the Asus Pro WS x570-ACE. This is in part because of its more sedate old school aesthetics, but also because of its claimed official ECC support 12+2 VRM phases and the fact that I've been using an Asus WS board with a rather extreme high clock and voltage overclock for 8 years and it had been Rock solid all that time.

I'm hoping there is no trickery with the VRM's on this board, and it winds up being as solid as my old P9x79 WS. It, according to reviews from when it launched had an 8 phase design.
 
Thanks for clarifying that.

Well in that case I agree with previous posters. This isn't really any new info.

In the past I never paid much attention to the VRM's when motherboard shopping. Maybe I shouöd more going forward.

This blog entry suggests motherboard manufacturers are trying to trick users by putting fake chokes and other components on the boards to make them look like they have more phases than they actually do.

This suggests that access to quality hardware reviews are even more important than ever.

My favorite x570 motherboard this far is the Asus Pro WS x570-ACE. This is in part because of its more sedate old school aesthetics, but also because of its claimed official ECC support 12+2 VRM phases and the fact that I've been using an Asus WS board with a rather extreme high clock and voltage overclock for 8 years and it had been Rock solid all that time.

I'm hoping there is no trickery with the VRM's on this board, and it winds up being as solid as my old P9x79 WS.

If you want to learn way more than you will ever want to know about VRMs check out the Youtube channel "Actually Hardcore Overclocking". A professional overclocker named Buildzoid talks A LOT about VRMs when he goes over motherboards. He likes to ramble a bit so be ready for that. Gamers Nexus also hires him to do motherboard analysis for them. Those videos are slightly more length controlled with slightly less rambling.
 
If you want to learn way more than you will ever want to know about VRMs check out the Youtube channel "Actually Hardcore Overclocking". A professional overclocker named Buildzoid talks A LOT about VRMs when he goes over motherboards. He likes to ramble a bit so be ready for that. Gamers Nexus also hires him to do motherboard analysis for them. Those videos are slightly more length controlled with slightly less rambling.


Too bad he doesn't do written articles. I can always sneak an article on the side in a spare moment. It's rare I can watch more than a few minutes video.

I feel like I can usually extract the same amou t of useful information from a well written article in about 10 minutes, as I wouöd get from an hour plus Long video.

Ain't no ody got time for that.
 
Overclocking is dead, long live overclocking! I enjoy tweaking but if the CPU can auto give me all its performance I guess that's okay too. The CPU cant build itself into a water loop just yet anyhow.

Exactly. There's a lot more to thinker with in today's computing world. Trying to extract a few more MHz is stupid, unless it's job :)
 
Yeah. I don't know a good written source for a lot of this stuff.
It's the same reason you wouldn't consider a job working for free. No money in long form written articles anymore. Bygone era.
 
It's the same reason you wouldn't consider a job working for free. No money in long form written articles anymore. Bygone era.

Very sad. It is by far the most effective way to convey technical details.

There is a reason the FAA doesn't do youtube video accident reports.

What I kind of decry is the financial motivation. I feel like with the exception of a handful of very successful sites, this wasn't done as a job in the past. It was a hobby.

Now everyone wants to be the next YouTube celeb.

I blame Google's decision to start sharing ad revenue with users. That was the beginning of the downturn.
 
Very sad. It is by far the most effective way to convey technical details.

There is a reason the FAA doesn't do youtube video accident reports.

What I kind of decry is the financial motivation. I feel like with the exception of a handful of very successful sites, this wasn't done as a job in the past. It was a hobby.

Now everyone wants to be the next YouTube celeb.

I blame Google's decision to start sharing ad revenue with users. That was the beginning of the downturn.
Yes, those videos are not produced for those of us working at a high level. I would really like to see that done by AMD in the future.
 
Wait! Did he write 4.75 Jigohertz!? Marty, fire up the Delorean, we're going shave off some render time!
 
Yeah, as I said, details are light.

Correct me if I am wrong though, but wasn't this previously only a Threadripper feature? Specifically a second gen Threadripper feature. Now it's coming to Ryzen 3000 series chips.

I knew that PBO used various aspects of chip health to determine overclocking capacity and dynamically and did so pretty well, but I did not know it had a data link to the motherboard to also probe and monitor the VRM's. This was new to me.

officially PBO was only on threadripper, unofficially most of the motherboard manufactures never removed the option from their bios when AMD decided to drop PBO support on am4 right before launch.
 
Thanks for clarifying that.

Well in that case I agree with previous posters. This isn't really any new info.

In the past I never paid much attention to the VRM's when motherboard shopping. I just assumed that if I was buying a quality high end motherboard, it would come with good power regulation. Maybe I should pay more attention going forward.

This blog entry suggests motherboard manufacturers are trying to trick users by putting fake chokes and other components on the boards to make them look like they have more phases than they actually do.

This suggests that access to quality hardware reviews are even more important than ever.

My favorite x570 motherboard this far is the Asus Pro WS x570-ACE. This is in part because of its more sedate old school aesthetics, but also because of its claimed official ECC support 12+2 VRM phases and the fact that I've been using an Asus WS board with a rather extreme high clock and voltage overclock for 8 years and it had been Rock solid all that time.

I'm hoping there is no trickery with the VRM's on this board, and it winds up being as solid as my old P9x79 WS. It, according to reviews from when it launched had an 8 phase design.


It looks like a decent setup, the IR 3555 isn't the best but it's pretty decent. One thing you might not consider (and a lot of custom loopers do.not) is active airflow over the VRM area.

Some people will rig w 120/140mm fan depending on their case etc but I actually found that using an extreme positive pressure setup works wonders. I have always had my 360mm rad up top with 3x Gentle Typhoon AP-15s as an intake. The extremely low FPI of my rad allows a ton of air and to pass through it and my VRM sinks are barely warm running 100% load 24/7 most of the time.

Just a friendly heads up, especially since you are planning on the 16 core for sure from what you have mentioned. That thing is going to draw a fair amount of power, especially since I know you are going to push the clocks.

Any idea on the MSRP of that board? I do not want to go back to paying north of $300 for a mobo, especially with DDR5 coming with the next socket. Zen2+ is going to be the last we get on AM4 from IIRC.
 
It looks like a decent setup, the IR 3555 isn't the best but it's pretty decent. One thing you might not consider (and a lot of custom loopers do.not) is active airflow over the VRM area.

Some people will rig w 120/140mm fan depending on their case etc but I actually found that using an extreme positive pressure setup works wonders. I have always had my 360mm rad up top with 3x Gentle Typhoon AP-15s as an intake. The extremely low FPI of my rad allows a ton of air and to pass through it and my VRM sinks are barely warm running 100% load 24/7 most of the time.

Just a friendly heads up, especially since you are planning on the 16 core for sure from what you have mentioned. That thing is going to draw a fair amount of power, especially since I know you are going to push the clocks.

Any idea on the MSRP of that board? I do not want to go back to paying north of $300 for a mobo, especially with DDR5 coming with the next socket. Zen2+ is going to be the last we get on AM4 from IIRC.

Link says it is a work station motherboard, what I have gathered from the whole X570 not being cheap and that there motherboards in the line up that goes near $1000 I would not get my hopes up on it being under $500.
 
It looks like a decent setup, the IR 3555 isn't the best but it's pretty decent. One thing you might not consider (and a lot of custom loopers do.not) is active airflow over the VRM area.

Some people will rig w 120/140mm fan depending on their case etc but I actually found that using an extreme positive pressure setup works wonders. I have always had my 360mm rad up top with 3x Gentle Typhoon AP-15s as an intake. The extremely low FPI of my rad allows a ton of air and to pass through it and my VRM sinks are barely warm running 100% load 24/7 most of the time.

Just a friendly heads up, especially since you are planning on the 16 core for sure from what you have mentioned. That thing is going to draw a fair amount of power, especially since I know you are going to push the clocks.

Any idea on the MSRP of that board? I do not want to go back to paying north of $300 for a mobo, especially with DDR5 coming with the next socket. Zen2+ is going to be the last we get on AM4 from IIRC.


I have seen nothing yet as far as pricing goes.

Well, there was one leak from some Danish or German etailer with motherboard pricing a while back, but they had the exact same price for every Asus board, so I am presuming they were just placeholders.

I'm guessing it will probably be at least a $400 board. That's what my last Asus WS board (an x79 model) cost in 2011, and I can't imagine things having gotten cheaper since then. Everything just seems be moving in the opposite direction instead for no apparent reason.

I appreciate the added VRM information. I know nothing about the different offerings here. Do you know where I might find more information about the IR3555 and how it compares to other offerings on the market?

Appreciate the heads up on the air flow. My system actually runs with a slight negative pressure. I know this isn't the favorite configuration, but I wanted to have a radiator up top, and one up front, and I neither wanted to force hot air to go downwards (why fight physics if you don't have to?) nor push hot air into the case, so I set them both blowing out.

To help make up for this a bit, I reversed the direction of the rear fan, and made it an intake, and have a fan in the bottom pulling air in as well.

I was concerned this would result in uncontrolled dust problems, but thus far (~3 years) it hasn't. Thank god I don't have pets :p

I wonder if it makes sense trying to put some sort of low restrictive water block on the VRM's. A WS board like this likely won't get a monoblock, and even if it did, I'm not sure I'd want it (the difficulties in tolerance stackups seem to usually result in worse CPU mating, and thus higher temps)
 
Thanks for clarifying that.

Well in that case I agree with previous posters. This isn't really any new info.

In the past I never paid much attention to the VRM's when motherboard shopping. I just assumed that if I was buying a quality high end motherboard, it would come with good power regulation. Maybe I should pay more attention going forward.

This blog entry suggests motherboard manufacturers are trying to trick users by putting fake chokes and other components on the boards to make them look like they have more phases than they actually do.

This suggests that access to quality hardware reviews are even more important than ever.

My favorite x570 motherboard this far is the Asus Pro WS x570-ACE. This is in part because of its more sedate old school aesthetics, but also because of its claimed official ECC support 12+2 VRM phases and the fact that I've been using an Asus WS board with a rather extreme high clock and voltage overclock for 8 years and it had been Rock solid all that time.

I'm hoping there is no trickery with the VRM's on this board, and it winds up being as solid as my old P9x79 WS. It, according to reviews from when it launched had an 8 phase design.


It looks like a decent setup, the IR 3555 isn't the best but it's pretty decent. One thing you might not consider (and a lot of custom loopers do.not) is active airflow over the VRM area.

Some people will rig w 120/140mm fan depending on their case etc but I actually found that using an extreme positive pressure setup works wonders. I have always had my 360mm rad up top with 3x Gentle Typhoon AP-15s as an intake. The extremely low FPI of my rad allows a ton of air and to pass through it and my VRM sinks are barely warm running 100% load 24/7 most of the time.

Just a friendly heads up, especially since you are planning on the 16 core for sure from what you have mentioned. That thing is going to draw a fair amount of power, especially since I know you are going to push the clocks.

Any idea on the MSRP of that board? I do not want to go back to paying north of $300 for a mobo, especially with DDR5 coming with the next socket. Zen2+ is going to be the last we get on AM4 from IIRC.
I have seen nothing yet as far as pricing goes.

Well, there was one leak from some Danish or German etailer with motherboard pricing a while back, but they had the exact same price for every Asus board, so I am presuming they were just placeholders.

I'm guessing it will probably be at least a $400 board. That's what my last Asus WS board (an x79 model) cost in 2011, and I can't imagine things having gotten cheaper since then. Everything just seems be moving in the opposite direction instead for no apparent reason.

I appreciate the added VRM information. I know nothing about the different offerings here. Do you know where I might find more information about the IR3555 and how it compares to other offerings on the market?

Appreciate the heads up on the air flow. My system actually runs with a slight negative pressure. I know this isn't the favorite configuration, but I wanted to have a radiator up top, and one up front, and I neither wanted to force hot air to go downwards (why fight physics if you don't have to?) nor push hot air into the case, so I set them both blowing out.

To help make up for this a bit, I reversed the direction of the rear fan, and made it an intake, and have a fan in the bottom pulling air in as well.

I was concerned this would result in uncontrolled dust problems, but thus far (~3 years) it hasn't. Thank god I don't have pets :p

I wonder if it makes sense trying to put some sort of low restrictive water block on the VRM's. A WS board like this likely won't get a monoblock, and even if it did, I'm not sure I'd want it (the difficulties in tolerance stackups seem to usually result in worse CPU mating, and thus higher temps)


I accidentally lied in my reply, but it is wasn't on purpose....My curreny case, a TT VIEW 71 Tempered Glass Edition had a 45mm thick low fpi rad in the front (3x140mm EK Vadars) and my 360 at the top the GT 120s....I actually had it setup with the front as intake, with the top rad as exhaust and my rear 120mm GT 120mm as an intake like you are running.


My issue is the fact that I no longer can use a manual fan controller (RIP 4chan Lamptron) to control fan speeds, and my Mobo doesn't have enough headers for 7 high amperage fans. So I have a 1 to 4 4x PWM powered splitter on each rad, so I only needed 3 fsn headers.

The second issue is the Asus fan software is garbage. If I were running 7 of the same fans it might not be that bad but the EK fans are LOUD at anything over 1krpm...slow them down, and they barely move any air. I kept my GT 120s speed up and my temps with 3 VEGAs and a 2700 @ 4.35Ghz all mining 24/7 never exceeded 63C on the CPU and mid 30s on the GPUS.

The thing with this setup is that I did not really have any active airflow over my VRMS, so while it was rock solid I was never happy with the lack of airflow. Now that my loop is torn down I am going to switch the 360mm to Intake.

My logic is the that 3/4" gap between the glass side panels and the case (something NO REVIEW mentioned) along with fact you can put two 140mm fans in the bottom and the mesh slot covers will allow me enough exhaust.

The issuse as you mentioned is that it physics. Given my great results using this kind of setup in. The past with a 5.2Ghz 3770K and 3x290s @ 1.3Ghz mining (never bothered to ubdervolt back then I just wanted as much Eth as possible so I was moving the waste heat of nearly 1200W downward) never gave me issues.

As far as the IR 3555, I would not worry with it being a 12+2 setup. If you went up to the higher end 3565 (IIRC the model number correctly) like on the 14 phase GB gaming board you would probably pay $100 more MSRP. Maybe more.
 
My issue is the fact that I no longer can use a manual fan controller (RIP 4chan Lamptron) to control fan speeds, and my Mobo doesn't have enough headers for 7 high amperage fans. So I have a 1 to 4 4x PWM powered splitter on each rad, so I only needed 3 fsn headers.

The second issue is the Asus fan software is garbage. If I were running 7 of the same fans it might not be that bad but the EK fans are LOUD at anything over 1krpm...slow them down, and they barely move any air. I kept my GT 120s speed up and my temps with 3 VEGAs and a 2700 @ 4.35Ghz all mining 24/7 never exceeded 63C on the CPU and mid 30s on the GPUS.

Yeah, I have struggled with fan control over the years as well.

I find I don't need a TON of fan channels, but as you say, you do want to keep some fans separate as they have differing fan blade designs and max speeds so they make different amounts of noise at the same settings.

When I first built my water loop, I wanted automated fan control, but also the ability to switch to manual at a whim when I so desired. I also didnt want to voltage control the fans, but preferred PWM. I did a lot of searching but couldn't for the life of me find a manual PWM controller. Zalman had apparently made a FanMate PWM at one point but it was discontinued and nothing else was on the market.

After some googling - however - I found a writeup on how to build my own. I built the dual timer (556 chip) version on this page. I have no real experience in soldering circuits (which will be excruciatingly clear when you see my solder joints) but I followed the schematic and it works really well.

proxy.php?image=https%3A%2F%2Fc1.staticflickr.com%2F9%2F8487%2F29403639391_05a4e6ab88.jpg
proxy.php?image=https%3A%2F%2Fc1.staticflickr.com%2F9%2F8536%2F29194870320_8e1e1c1cf6.jpg



Here is the final shot, covered with some hotglue just to make sure no components move and short on eachother.


65439_IMG_20160908_152941.jpg


Then I got an attractive looking knob and a couple of switches, and mounted them to a spare drive bay cover for my case:


65440_IMG_20160908_214604.jpg


Rightmost switch switches all fans from off to on mode. I don't use this very often, but sometimes if there is a noise and I want to troubleshoot (is it caused by fans?), it has come in handy.

The middle switch switches between auto and manual fan control.

If set to manual fan control, the dial determines the PWM duty signal % to send to the fans.

Now for the auto fan control, I wanted to control it based on loop temp, but my motherboard (which I bought years before building the loop) did not have the capability to hook up additional temp sensors to use its fan control, so I bought a $7 Chinese fan controller controller on eBay.

64791_pwm.jpg


It had dual channels, so I could set two different models of fans separately. So, when in auto mode the two different fans would run at different speeds, but when I switched to manual, all would run at the same manual speed. It was amazing in what it could do given its price, but I was never happy with its linear fan curves, always giving me higher fan speeds at idle than I wanted or needed.

Because of this I kind of kept my fans in manual mode 99% of the time. Over time I learned what fan settings would be needed in what games at what room temperatures.

I had always wanted to try out one of Aquuacomputer's Aquaero fan controllers, but I had always balked at the kind of crazy price, until this past December, when I just decided to do it, and picked up an Aquaero 6 XT. I also bought 4 of their calitemp high accuracy temperature sensors and one of their flow meters.

I wish I had done this long ago. It has 4 PWM fan channels, (and I think 8 voltage control channels) and they can be programmed in amazing ways. I still ahve to use some PWM splitters, but it is all good.

The coolest feature is that they have a true PID controller in them, so you can have it set fan speeds based on target temp, NOT based on fan curves, which completely eliminates the problem with fan curves where your fans are running faster than they need to at lower loads, and are keeping your temps lower than you really need them.

upload_2019-7-4_20-37-30.png


I had originally planned on using their high accuracy temp probes, putting one before and after each of my two blocks, setting pump speed based on the delta T across the block, and setting the fan speed off of the max of the four sensors, but it turns out even high accuracy probes are not accurate enough to do delta T across a block, so instead I control pump speed based on the max of the two temp sensors on the "out" side of the blocks, and fan speed based on the max of the two temp sensors on the "in" side. Both running off of PID targets instead of fan curves.

208907_upload_2019-4-21_17-53-8.png



I love this fan controller. Wish I had just sucked up the price and bought one years ago. It's pretty expensive for a fan controller, but well worth it. I still like having my custom manual controls, but once I got the Aquaero set up the way I like it, I very rarely adjust anything manually anymore. I have a target loop temp set at 35C, which is enough that my GPU never hots 40C under full load. I just don't have to touch anything.

In retrospect the XT model was a waste though. Only difference (I believe) is the bluetooth remote unit, which I never use.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
I have seen nothing yet as far as pricing goes.

Well, there was one leak from some Danish or German etailer with motherboard pricing a while back, but they had the exact same price for every Asus board, so I am presuming they were just placeholders.

I'm guessing it will probably be at least a $400 board. That's what my last Asus WS board (an x79 model) cost in 2011, and I can't imagine things having gotten cheaper since then. Everything just seems be moving in the opposite direction instead for no apparent reason.

I appreciate the added VRM information. I know nothing about the different offerings here. Do you know where I might find more information about the IR3555 and how it compares to other offerings on the market?

Appreciate the heads up on the air flow. My system actually runs with a slight negative pressure. I know this isn't the favorite configuration, but I wanted to have a radiator up top, and one up front, and I neither wanted to force hot air to go downwards (why fight physics if you don't have to?) nor push hot air into the case, so I set them both blowing out.

To help make up for this a bit, I reversed the direction of the rear fan, and made it an intake, and have a fan in the bottom pulling air in as well.

I was concerned this would result in uncontrolled dust problems, but thus far (~3 years) it hasn't. Thank god I don't have pets :p

I wonder if it makes sense trying to put some sort of low restrictive water block on the VRM's. A WS board like this likely won't get a monoblock, and even if it did, I'm not sure I'd want it (the difficulties in tolerance stackups seem to usually result in worse CPU mating, and thus higher temps)

A Danish site supposedly leaked motherboard a price list a bit ago. The price leak for Zen 2 was a bit high, at least when doing conversions and accounting for VAT, so the MB price probably is too. Either way, the conversion and accounting for VAT it comes to around $480 for that board. So maybe in the $450-ish range or a little less. The handful of prices we already have for X570 boards are pretty expensive, so it wouldn't surprise me.
 
Yeah, I have struggled with fan control over the years as well.

I find I don't need a TON of fan channels, but as you say, you do want to keep some fans separate as they have differing fan blade designs and max speeds so they make different amounts of noise at the same settings.

When I first built my water loop, I wanted automated fan control, but also the ability to switch to manual at a whim when I so desired. I also didnt want to voltage control the fans, but preferred PWM. I did a lot of searching but couldn't for the life of me find a manual PWM controller. Zalman had apparently made a FanMate PWM at one point but it was discontinued and nothing else was on the market.

After some googling - however - I found a writeup on how to build my own. I built the dual timer (556 chip) version on this page. I have no real experience in soldering circuits (which will be excruciatingly clear when you see my solder joints) but I followed the schematic and it works really well.

View attachment 171939 View attachment 171940


Here is the final shot, covered with some hotglue just to make sure no components move and short on eachother.


View attachment 171941

Then I got an attractive looking knob and a couple of switches, and mounted them to a spare drive bay cover for my case:


View attachment 171942

Rightmost switch switches all fans from off to on mode. I don't use this very often, but sometimes if there is a noise and I want to troubleshoot (is it caused by fans?), it has come in handy.

The middle switch switches between auto and manual fan control.

If set to manual fan control, the dial determines the PWM duty signal % to send to the fans.

Now for the auto fan control, I wanted to control it based on loop temp, but my motherboard (which I bought years before building the loop) did not have the capability to hook up additional temp sensors to use its fan control, so I bought a $7 Chinese fan controller controller on eBay.

View attachment 171943

It had dual channels, so I could set two different models of fans separately. So, when in auto mode the two different fans would run at different speeds, but when I switched to manual, all would run at the same manual speed. It was amazing in what it could do given its price, but I was never happy with its linear fan curves, always giving me higher fan speeds at idle than I wanted or needed.

Because of this I kind of kept my fans in manual mode 99% of the time. Over time I learned what fan settings would be needed in what games at what room temperatures.

I had always wanted to try out one of Aquuacomputer's Aquaero fan controllers, but I had always balked at the kind of crazy price, until this past December, when I just decided to do it, and picked up an Aquaero 6 XT. I also bought 4 of their calitemp high accuracy temperature sensors and one of their flow meters.

I wish I had done this long ago. It has 4 PWM fan channels, (and I think 8 voltage control channels) and they can be programmed in amazing ways. I still ahve to use some PWM splitters, but it is all good.

The coolest feature is that they have a true PID controller in them, so you can have it set fan speeds based on target temp, NOT based on fan curves, which completely eliminates the problem with fan curves where your fans are running faster than they need to at lower loads, and are keeping your temps lower than you really need them.

View attachment 171938

I had originally planned on using their high accuracy temp probes, putting one before and after each of my two blocks, setting pump speed based on the delta T across the block, and setting the fan speed off of the max of the four sensors, but it turns out even high accuracy probes are not accurate enough to do delta T across a block, so instead I control pump speed based on the max of the two temp sensors on the "out" side of the blocks, and fan speed based on the max of the two temp sensors on the "in" side. Both running off of PID targets instead of fan curves.

View attachment 171944


I love this fan controller. Wish I had just sucked up the price and bought one years ago. It's pretty expensive for a fan controller, but well worth it. I still like having my custom manual controls, but once I got the Aquaero set up the way I like it, I very rarely adjust anything manually anymore. I have a target loop temp set at 35C, which is enough that my GPU never hots 40C under full load. I just don't have to touch anything.

In retrospect the XT model was a waste though. Only difference (I believe) is the bluetooth remote unit, which I never use.


Wow, nice write up. I do have a lot of room behind my Mobo and the glass side panel (it's designed to hold 4x2.5 or 3.x3.5" drives and I only have a single HD with a 2.5"SSD going into a StoreMi config for those big 4k Linux ISOs ;))..I could tuck some boards back there.

I liked the AC controllers as well, but the Lamptron I had could do like 5A a channel and had 4x chnnsels. A buddy of mine won this YT giveaway, it was in there ($89 at the time) this insane mouse pad that came with a big case and some other stuff. So I bought it from him.

My biggest issue is that while my case claims to have a 5.25" bay, it really doesn't in the sense that even with o ky a 360 up front on nothing will fit and the glass panel on the front would prohibit the knobs from being used.

I really have a love hate relationship with this case. I was scrambling to find something to replace my highly modiifed HAF932 that was silenced with sound deadener when I accidentally broke my front 230mm rad. I could not find a replacement for under $200, and it wwss going to ake weeks to get here so I switched. It was the cheapest option that could hold a 360+420.

Anyway, I need to stop detailing the thread.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
Wow, nice write up. I do have a lot of room behind my Mobo and the glass side panel (it's designed to hold 4x2.5 or 3.x3.5" drives and I only have a single HD with a 2.5"SSD going into a StoreMi config for those big 4k Linux ISOs ;))..I could tuck some boards back there.

I liked the AC controllers as well, but the Lamptron I had could do like 5A a channel and had 4x chnnsels. A buddy of mine won this YT giveaway, it was in there ($89 at the time) this insane mouse pad that came with a big case and some other stuff. So I bought it from him.

My biggest issue is that while my case claims to have a 5.25" bay, it really doesn't in the sense that even with o ky a 360 up front on nothing will fit and the glass panel on the front would prohibit the knobs from being used.

I really have a love hate relationship with this case. I was scrambling to find something to replace my highly modiifed HAF932 that was silenced with sound deadener when I accidentally broke my front 230mm rad. I could not find a replacement for under $200, and it wwss going to ake weeks to get here so I switched. It was the cheapest option that could hold a 360+420.

Anyway, I need to stop detailing the thread.

Just so you are aware, the Aquacomputer units come in many varieties. Some - like this one - are just boards not designed for bay mounting, with no screen you can just mount anywhere it fits using double sided tape or random screw holes. In this case you just control the unit using the software. The screen does look nice, but I have found in daily use I almost never look at it, so I don't think you'd lose much by going with a screenless version. At only about $100, it saves some money too. The Aquaero 5 is even cheaper, but it has many more limitations, so I wouldn't bother with it.

Stock, the Aquaero 6 has 4 fan channels, all of which can be either voltage or PWM controlled, and support up to 2.5A per channel. According to the spec sheet there some way to expand this up to 12 channels, but I am not entirely sure how it works. It probably requires some sort of accessory board. VanGoghComplex is pretty knowledgeable about these units and may know. I'd imagine with many channels or voltage control of fans, one of the optional heatsinks for the unit would probably be wise.

In my setup the amperage is pretty much irrelevant. I use a couple of high quality PWM splitters, which provide all the power, so I am not drawing any power from the unit at all, just tapping it for the PWM signals on two of the fan channels. Only thing I lose doing this is that I only get RPM readouts from one fan per channel, but I can live with that.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
It looks like a decent setup, the IR 3555 isn't the best but it's pretty decent. One thing you might not consider (and a lot of custom loopers do.not) is active airflow over the VRM area.

Some people will rig w 120/140mm fan depending on their case etc but I actually found that using an extreme positive pressure setup works wonders. I have always had my 360mm rad up top with 3x Gentle Typhoon AP-15s as an intake. The extremely low FPI of my rad allows a ton of air and to pass through it and my VRM sinks are barely warm running 100% load 24/7 most of the time.

Just a friendly heads up, especially since you are planning on the 16 core for sure from what you have mentioned. That thing is going to draw a fair amount of power, especially since I know you are going to push the clocks.

Any idea on the MSRP of that board? I do not want to go back to paying north of $300 for a mobo, especially with DDR5 coming with the next socket. Zen2+ is going to be the last we get on AM4 from IIRC.



I accidentally lied in my reply, but it is wasn't on purpose....My curreny case, a TT VIEW 71 Tempered Glass Edition had a 45mm thick low fpi rad in the front (3x140mm EK Vadars) and my 360 at the top the GT 120s....I actually had it setup with the front as intake, with the top rad as exhaust and my rear 120mm GT 120mm as an intake like you are running.


My issue is the fact that I no longer can use a manual fan controller (RIP 4chan Lamptron) to control fan speeds, and my Mobo doesn't have enough headers for 7 high amperage fans. So I have a 1 to 4 4x PWM powered splitter on each rad, so I only needed 3 fsn headers.

The second issue is the Asus fan software is garbage. If I were running 7 of the same fans it might not be that bad but the EK fans are LOUD at anything over 1krpm...slow them down, and they barely move any air. I kept my GT 120s speed up and my temps with 3 VEGAs and a 2700 @ 4.35Ghz all mining 24/7 never exceeded 63C on the CPU and mid 30s on the GPUS.

The thing with this setup is that I did not really have any active airflow over my VRMS, so while it was rock solid I was never happy with the lack of airflow. Now that my loop is torn down I am going to switch the 360mm to Intake.

My logic is the that 3/4" gap between the glass side panels and the case (something NO REVIEW mentioned) along with fact you can put two 140mm fans in the bottom and the mesh slot covers will allow me enough exhaust.

The issuse as you mentioned is that it physics. Given my great results using this kind of setup in. The past with a 5.2Ghz 3770K and 3x290s @ 1.3Ghz mining (never bothered to ubdervolt back then I just wanted as much Eth as possible so I was moving the waste heat of nearly 1200W downward) never gave me issues.

As far as the IR 3555, I would not worry with it being a 12+2 setup. If you went up to the higher end 3565 (IIRC the model number correctly) like on the 14 phase GB gaming board you would probably pay $100 more MSRP. Maybe more.


Newegg now has the Asus Pro WS x570-Ace for $379. Much better than I expected given the crazy price of that $700 "Godlike" board.
 
Newegg now has the Asus Pro WS x570-Ace for $379. Much better than I expected given the crazy price of that $700 "Godlike" board.

The only problem is that the CH8 Hero is the same price and has more "features." Go figure that you are paying more money to not have "bling."

ECC support with that WS board? I didn't look too closely at it.
 
The only problem is that the CH8 Hero is the same price and has more "features." Go figure that you are paying more money to not have "bling."

ECC support with that WS board? I didn't look too closely at it.

Yeah, the product page claims ECC support. And OOB management via a dedicated NIC port, if you are into that sort of thing.

I really don't care about "features". Probably going to wind up disabling onboard sound, NIC's and SATA anyway.
 
Yeah, the product page claims ECC support. And OOB management via a dedicated NIC port, if you are into that sort of thing.

I really don't care about "features". Probably going to wind up disabling onboard sound, NIC's and SATA anyway.


It sucks since you are paying for those features. It would be awesome if we could get stripped down OC'ing first boards (in the vein of the DFI LANPARTY SERIES) without all the insane stuff.

The issue is these days it cost peanuts to add a SB overlaid a Realtexc Audio codec, a few LEDs, and built in wifi , all in order to charge $1000 peanuts more.


Most of the YT generation just wants flashly lights that all match (they don't care if the fans themselves suck harder then a Hoover) so they can toss it together and run it at stock to play FN and PUBG/Apex Legends with their friends.
 
Back
Top