Motherboard ASRock X370 Gaming-ITX/ac and AB350 Gaming-ITX/ac

[m_42]

Mainboard: Asrock Fatal1ty AB350 ITX/ac (BIOS 4.51)
CPU: Ryzen 5 2400G
CPU Cooler: Noctua NH-L9a M4
Memory: G.Skill Flare X DIMM Kit 16GB, DDR4-3200
SSD Drive: Samsung M.2 960 Evo
Power: HDPlex 160 W / 300 W Combo Power Supply

Running Windows 10 Pro with RAM @ XMP Profile 2

Those two high temps (1st picture) are bothering me a little. The system is running flawlessly though .....

 

[RedArt]

With these VRM temps dropping under load, I'm thinking whether it's somehow reporting Tdelta-to-Tmax... It would make sense. Suppose Tmax is ~160C. 115C reported would mean the actual temp is 45C.

That would be awesome lol

I didn't hide anything.
Why do you think that?

Because the user TheHig said he disabled that reading and in his screenshot there was no temp on vrm so I thought u did the same.

 

[parlinone]

So many people around here with a 2200G/2400G and nobody can check if they can disable a core or SMT? Just make sure you save your BIOS settings if you'd be willing to check.

I'm getting that bizarre temperature reading as well with Speedfan but in the BIOS all temps seem normal so it seems a bug to me. Not sure why there's such a big fuss about it...?

 

[m_42]

... Not sure why there's such a big fuss about it...?

Because it's always better to be safe than sorry.
Yes, it must be a bug considering all the reports I've come across so far.

 

[ChinStrap]

People have reported issues with the 960 Evo NVMe on bios 4.40. Just a heads up, I'm using beta 4.51 and my 960 Evo just disappeared last night. I had just finished moving a large number of pictures off my phone to my NAS and the computer restarted. It just didn't load back into windows, it loaded into the bios. I thought that was odd, since I didn't strike delete. Powered down and restarted, only to be dumped back into bios.

No NVMe listed in Bios. I plugged the 960 Evo into my backup rig, An Asrock z270m-itx and it showed up in bios on that board. Plugged it back into the x370, not listed.

I ended up grabbing a 256g 830 out of the stash to get me going again. I'm going to load the 960 Evo on the z270 board tonight and make sure it's working okay.

Just a heads-up to people using an 960 Evo on a newer Bios, make sure your stuff is backed up.

x370 Gaming-ITX (Beta 4.51)
2400G
G.SKILL F4-3200C14D-16GVR

 

[ChinStrap]

anandtech just posted a review: https://www.anandtech.com/show/12569/the-asrock-x370-gaming-itx-ac-motherboard-review

 

[theGryphon]

anandtech just posted a review: https://www.anandtech.com/show/12569/the-asrock-x370-gaming-itx-ac-motherboard-review

That's a little weird. A new chipset is here and they're reviewing the previous gen. Cuckoo...

LOL, they have a whole section on it:

Wait a Second, Why Test X370 if X470 is Just Around The Corner?
Section by Ian Cutress
A lot of discussion is about AMD's upcoming launch, which is set to include a new motherboard chipset known as X470. We have had emails about our last B350 review, stating what was the point given that this new AM4-based chipset was around the corner, and I fully expect more comments on this article about it as well. The reasons are multiple.

One) X470 isn't replacing any of the current motherboard stack. X370, B350 and A320 are still going to be produced and sold alongside X470. There might be fewer new models compared to the latest and greatest, but motherboard vendors are still set to sell all of them side by side.

Two) All of these motherboards will have forwards/backwards compatibilty with the AM4 socket. Dropping in a Bristol Ridge or the latest mainstream Ryzen, with a sufficiently updated BIOS, will be possible. AMD has stated that the AM4 socket, and my extention the chipsets, will be on a long-term support structure. So the motherboard vendors are still set to run updates for the chipsets that exist.

Three) When we started our AM4 reviews, we tested a number of motherboards all at once, and have gone through the process of writing up our analysis one-by-one. The analysis, for current owners or to-be owners (of new or used), is a very important part of that process to help making purchasing decisions. Not everyone is buying the latest, and not everyone is buying new. We believe having that data on hand, especially when enabled through a Google search six months later down the line, is important.

We still have motherboard vendors asking if we can review their current-generation AM4 motherboards, regardless of what is coming around the corner. You might be amazed and how soon after recieving a sample we get emails asking 'when will it be published?'! But the bottom line is that sales do not stop just because something newer is about to come out, and we think that having our analysis online can act as a good reference for those that need it.

For those waiting on X470 reviews, we have one or two waiting in the wings. Stay tuned!

Final edit: The reason Two there is BS. Not only a lot of good features of new gen CPUs are not supported by the 3XX chipset, the overall firmware update experience is not very good (from complaints I've read).

It works, but it clearly is not ideal and you're not getting the "by design" performance from the new CPUs, which pretty much defeats the whole CPU-update-only idea (disregarding the APUs). Sure, they will clock a bit higher still, but if you have money to throw, why not update the motherboard as well?

If anyone is buying a new Ryzen 2XXX CPU, there is totally no reason whatsoever to go and buy the 3XX series board. Lol, you wouldn't be able to use it right away either, because it needs a firmware update.

 

[greyhamism]

That's a little weird. A new chipset is here and they're reviewing the previous gen. Cuckoo...

Final edit: The reason Two there is BS. Not only a lot of good features of new gen CPUs are not supported by the 3XX chipset, the overall firmware update experience is not very good (from complaints I've read).

It works, but it clearly is not ideal and you're not getting the "by design" performance from the new CPUs, which pretty much defeats the whole CPU-update-only idea (disregarding the APUs). Sure, they will clock a bit higher still, but if you have money to throw, why not update the motherboard as well?

If anyone is buying a new Ryzen 2XXX CPU, there is totally no reason whatsoever to go and buy the 3XX series board. Lol, you wouldn't be able to use it right away either, because it needs a firmware update.

I wasn't aware that there where any major features supported by X470 that weren't supported by X370?

A reason to buy a 3XX series board to pair with your new Ryzen+ CPU might be form factor? As far as im aware there aren't any ITX X470 boards currently available, and given how long it took for decent X370/B350 ITX boards to be available after Ryzen's initial launch, a 2600/2600X/2700/2700X with an existing X370/B350 offering from ASRock of ASUS might be the only way to get an ITX/SFF system up and running.

 

[Midiamp]

I wasn't aware that there where any major features supported by X470 that weren't supported by X370?

XFR and precision boost 2.0. It's there on the press release that to utilize a better automatic clock boost, you will be needing the new X/B 4XX motherboards as it involves power delivery. If that's the case then, a BIOS upgrade won't matter since the current components aren't design to deliver said power.

Anandtech review of the Ryzen 2XXX draws a very clear picture why a motherboard upgrade is needed (but not necessary).

Here's a quote

The results are not quite as clear: it would seem that the 1800X draws, as a percentage of peak power, more at low thread count, but the 2700X draws more at a middling thread count.

It is worth noting that the end result of Precision Boost 2 is two-fold: more performance, but also more power consumption. Users looking to place one of the lower powered processors into a small form factor system might look at disabling this feature and returning to a standard step-function response in order to keep the thermal capabilities in check.

Putting that Ryzen 2700X in current X/B300 series will force the motherboard to deliver more power than it was originally designed and that troubles me. Again here's a quote from Anandtech.

Clearly 3% IPC is not enough, so AMD is combining the performance gain with the +250 MHz increase, which is about another 6% peak frequency, with better turbo performance with Precision Boost 2 / XFR 2. This is about 10%, on paper at least

You basically need Precision Boost and XFR 2.0 to get the advertised increased in performance. If not, you'll only be getting the IPC increase and the clock speed. However, no reviewer yet tested how the 2700/X fares on existing 300 series motherboard, so I can't draw any conclusion other than to wait for an ITX 400 series if I want to upgrade. Perhaps it won't matter much if you're overclocking the system, but I use mine stock.

 

[Phuncz]

To play devil's advocate, Intel's more recent flexible use of TDP also enables performance that only unlocks in certain scenarios. If you want the performance of the i7-8700K that the reviewers are showing, you need a above average Z370 board and above average CPU cooler, along with a limit on duration unless properly cooled on top of that.

I dislike these practices as they portray performance that's not available in most situations because most people still buy their computers from the shelves and even if they don't, these performance "disclaimers" are not well-documented.

Performance is basically optimized for review scoring, which was a popular practice with GPUs for quite some time, but now both AMD and Intel are guilty of doing this as well.

 

[a13antichrist]

So many people around here with a 2200G/2400G and nobody can check if they can disable a core or SMT? Just make sure you save your BIOS settings if you'd be willing to check.

I'm getting that bizarre temperature reading as well with Speedfan but in the BIOS all temps seem normal so it seems a bug to me. Not sure why there's such a big fuss about it...?

Are you asking if it's possible, or are you asking someone to do it in order to test something? There is an option to disable SMT, I could do it but I can't test anything as I'm running on a makeshift PSU right now and don't want to push any load at present..

 

[ChinStrap]

So many people around here with a 2200G/2400G and nobody can check if they can disable a core or SMT? Just make sure you save your BIOS settings if you'd be willing to check.

I'm getting that bizarre temperature reading as well with Speedfan but in the BIOS all temps seem normal so it seems a bug to me. Not sure why there's such a big fuss about it...?

SMT is right on the front OC Screen, Disabling Cores is buried in the AMD 'Zen' options. I'll grab a path for you tonight.

 

[Midiamp]

So... I just bought the 2700X... That was one stupid impulse purchase.

I use Asrock AB350 ITX Gaming, and here's my impression.

I believe you can find more coherent and complete test on professional websites, as I will focus on power and thermal performance of the 2700X.

TL;DR: buy the 2700 and not the 2700X if your only choice of cooler is a low profile HSF.

To say that AMD 105W TDP is just a label is an understatement, this thing runs hotter than the 1700X I previously used. I bragged about my new build with Fractal Design Define Nano S + Coolermaster 120 Lite thermal performance, and this processor politely disagrees. Before, the 1700X peaks at 60 degrees C, with average load at around 56 degrees C. With the 2700X it peaks at 73 degrees C with average load at around 60 degrees C. It's not bad, but definitely it's not ideal on low profile HSF. The 2700 on the other hand only has 65W, reflecting on my data, it will be far more manageable.

Power wise, I'm a little bit confused. Maybe because I use the B350, power numbers are everywhere. On single core apps, I see a peak of 1.481V. My friend told me that's quite high, even an overclocked 2700 runs all cores at 4 GHz at just 1.25V. Probably that's the XFR/PB doing its thing. on single core apps, the processor runs at 4.1 GHz on 2 cores. I tested with World Of Tanks for that, and temp jumps from 53 to 67 degrees C with CPU voltages ranges from 1.2-1.48 V. I tested with Farcry 5 and Forza Horizon 3 that utilize all 16 threads and numbers are a bit more stable. Voltages stable at 1.3V and clocks stable at 3.9 GHz (slight tens of MHz variations). Temps are also more stable at 55 degrees C.

My friend recommended me to just overclocks the processor and pegged it at 4GHz. Well, I'm not much of an overclocker and I actually don't like the concept of it. I thoroughly advised to avoid 2700X until HSF maker look into the thermal envelope of the processor for SFF builders. The 2700 OTOH, is a great pick.

I'm still mulling whether to buy X470/B450 for the processor. Reading the reviews, the 2700X should be good for 4 - 4.1 GHz running on all cores. But I guess the XFR/PB 2.0 is only available on the new motherboard since it's built from the ground up to deliver stable power the processor. A 200 MHz increase, again I don't know how much performance increase.. But I'm wondering about the power efficiency if any.

 

[TheHig]

Totally agree. Get the 2700 or 2600 for sff and smaller cooling set ups.

I’m running a 2600 stock on the AsRock B350 itx and the voltage peaks out at less than 1.3 at full load. This is at 3800 all core gaming. 3.9 and even 4 show up on single core. It’s really impressive.

However the same board and bios 4.5 does push the 1600 to 1.39-4v to get 3.4 all core. The refresh has some really nice improvements.

The 2700x is a great cpu but to get the performance AMD had it operating out of its sweet spot. I found even the 1700 was too hot on this board with the L9a for me. The two extra cores add some heat especially when loaded IMO.

Edit. Turns out I’m not on the latest bios after all!

 

[Midiamp]

I'm on the latest 4.6 bios. I guess the platform is already matured as it is, the refresh motherboards should have minimal issues. I still remember some of my friend who sells computer parts got got yelled at because of a particular Asrock AM4 motherboard has memory compatibility issues.

 

[stree]

Beta BIOS up. 4.51D with the new AGESA for those on ASR AB350 ITX. probably X370 soon as well

 

[a13antichrist]

Beta BIOS up. 4.51D with the new AGESA for those on ASR AB350 ITX. probably X370 soon as well

That's weird that they would release a .51D Beta when the .6 Full is already released a week ago..

 

[Phuncz]

The 2nd generation X-models seem to be amazing auto-overclockers, allowing much higher performance with the right hardware (X470 board, good cooling). Ofcourse this requires more power and will produce more heat so consider the 105W TDP before buying ! The non-X models are indeed a lot more suited for proper SFF builds.

 

[ChinStrap]

That's weird that they would release a .51D Beta when the .6 Full is already released a week ago..

agree. for the x370 we have a couple new versions posted now:
4.60-04/24/18- "Enhance compatibility for Raven CPU with Intel WiFi."
Beta 4.53 - 04/30/18 - "Update Agesa Version to PinnaclePI-AM4_1.0.0.2a"

same situation, newer dated bios with beta4.53 has a lower number. hmmmm.

 

[ChinStrap]

Also found this: Post by User 'pg123' on a different board, posting something he found on Anandtech forums, posted by The Stilt:

Don't get confused by the AGESA version numbers, because they're different for each branch.

- SummitPI 1.0.0.6x (Bristol and Summit Ridge)
- RavenPI 1.1.0.1 (Bristol, Summit and Raven Ridge)
- PinnaclePI 1.0.0.0x (Bristol, Summit, Raven and Pinnacle Ridge)
- ThreadRipperPI 1.0.0.5 (Zeppelin SP3r2 MCM)

PinnaclePI is the most current development branch for AM4 platform.
Should be used for all AM4 parts, as soon as it reaches full maturity.