What B450 MB for ryzen 2200g + 2x4GB kingston hiperx predator?

[MaxTheFast]

Hi all,
I'm building a desktop pc with amd ryzen 2200g. I was going to buy a MSI B450-a pro MB so I had already bought a 2x4GB kit kingston hiperx predator @3200mhz cl16 ram (code HX432C16PB3K2/8) that would has been ok for that MSI; I know it's weird to get RAM for 1^ and then MB but that's my actual situation. Now I'm pointing to other MBs because that MSI has gone out of my budget, I could wait its price goes down but there're 30 days to give RAM back if it's DOA or not compatible and I can't test it anyhow right now.

I know 2200g works @2993MHz and my goal is to make that RAM work @3200mhz stably without any problem in the long period so I'll overclock it and I'm looking for a such good and capable MB. The following MBs are my best deals for price right now:
- ASUS PRIME B450-PLUS (90MB0YN0-M0EAY0)
- asrock b450 pro4 (90-MXB8B0-A0UAYZ)
- asrock Fatal1ty B450 Gaming K4 (90-MXB8A0-A0UAYZ)

My main concern is firstly about RAM compatibility because mine isn't listed anywhere in QVLs tough we can find something similar for each MB:
- MSI B450-a pro
https://www.msi.com/Motherboard/support/B450-A-PRO#support-mem-13 (then choose 2200g)
HX432C16PB3K2/32 -> kit 2x16GB @3200 cl16
- ASUS PRIME B450-PLUS
https://dlcdnets.asus.com/pub/ASUS/mb/SocketAM4/TUF_B450M-PLUS_GAMING/Memory_QVL_For_AMD_Ryzen_2000_Series_Processors.pdf
HX432C16PB3K2/16 -> kit 2x8GB @3200 cl16
- asrock b450 pro4
https://www.asrock.com/mb/AMD/B450%20Pro4/index.asp#Memory
HX430C15PB3K2/16 -> kit 2x8GB @3000 cl15
- asrock Fatal1ty B450 Gaming K4
https://www.asrock.com/mb/AMD/Fatal1ty%20B450%20Gaming%20K4/index.asp#Memory
as asrock b450 pro4
My second concern is if RAM will work @3200mhz with those MBs and if they'll do a good job, I mean keeping system stably.

What can you suggest? Are you experienced in those configurations?

[Paul]

I won't research this, but I'll offer these observations.

1) Select a company that will release a good quality BIOS.
Asus or MSI or Gigabyte etc. After seeing more than one
questionable call by Asrock, I can't promote them. They
make nice hardware, but their BIOS teams are absent minded.

2) If the RAM appears on *any ones* QVL, it should work on
another motherboard having the same computing device. So if
the MSI listed your RAM exact model, I wouldn't be super-concerned
it's an "Intel only" or something.

As for the RAM speed, really, who gives a rats ass :-) Sorry,
but the RAM I got here, doesn't run at the rated speed,
and I really don't notice. The only time it might have
made a difference, is running 7ZIP ultra compression. When
you run the RAM at a slightly lower clock, the CAS will
be bumped down proportionally.

And you want two sticks, rather than four. That reduces
bus loading, and leads to less disappointment concerning
the final settings. That's probably why my RAM had to be
turned down, is all slots are full on the machine, and
it simply isn't stable turned up. However, on my system,
the XMP worked first time and at the rated speed. But
adding a second set of RAM, XMP doesn't work for four sticks,
and the setup would not work at exactly the same speed.
A couple adjustments and I can run Prime for hours, so
no errors. And it computes just fine.

The measured (Streams) bandwidth is no-where near the
theoretical, and on my particular platform, bus efficiency
is around 25%.

It's really a whole lot of fuss about nothing. Plug it in.
Try the rated values. If it doesn't work, turn it down
and move on. Don't forget to run the various tests before
using it for stuff.

A good BIOS should be able to sort the RAM values. The
"rating" on your RAM stick, might not be encoded in the
sticks, and you could be expected to enter the values
manually. For enthusiast RAM, companies use slower settings
in the table, to "guarantee" the RAM runs without crashing
before you can access the BIOS. Then you can set it to
the rated value and voltage. For systems which support
XMP, you enter the BIOS and enable XMP, and the hard work
is done for you. But if you don't have an XMP platform,
it's not that hard to dial it in.

Paul

[MaxTheFast]

Thanks for your explanation and experience. As you could guess I'm very newbie and it's my first PC build.

Now I see I've made a mistake when I got 3200mhz ram instead of the cheaper and TESTED 3000mhz hiperx, but I've red the followings:
-http://www.tomshardware.co.uk/answers/id-3742308/memory-choose-ryzen-2200g.html
"3000mhz (2966mhz) / 3200Mhz memory is best as it strikes a balance between costs and speed, remember Vega shares sytem memory."
-https://www.quora.com/Which-RAM-should-I-get-with-a-Ryzen-3-2200G-which-can-be-overclocked
"something with 2400mhz is as low as i would suggest, but it’s better to get at least 3000mhz, anything above that will just improve processor function even better,"
so i thought it'd has been better to get higher speed ram to improve graphics due to the fact it's integrated - I won't get a graph card. Anyway I've got this hiperx and I'll keep it if it works!

As I said my ram's *exact* code insn't listed in any QVLs of the above MBs, I just wrote the "similar" codes but as far as I know it's a lottery to guess if mine will be campatible or not on this base. Anyway I'll avoid Asrock due to its BIOS so thanks for your suggestion. I think I'll get the Asus if the Msi price won't go down in the next week.
Are you experienced in both asus and msi BIOSes? If yes, wich one and why do you prefer? Consider I've always used only a very old intel notebook with XP so I only know its dos bios!

[Paul]

BIOS development involved two companies.

A separate company writes the "bringup code". Award, AMI, Phoenix, Insyde...

The motherboard companies tune the parameters in the BIOS,
and the motherboard company doesn't have all the source code
with which to "compile" binaries with. Drivers for onboard
peripherals are modular (i.e. add a NEC/Renesas code module
to make a USB3 chip work). With the right tool, you can
see a list of the modules in the BIOS.

The motherboard companies must have some "extension" capability,
because they do adjust dividers that wouldn't normally be
visible. I've not seen any evidence of the really techie
controls being "in some GUI", so they must have some
other controls for playing with stuff. Sometimes one
company figures out some undocumented controls, better
than some other motherboard company. But once the second
motherboard company figures it out, they can usually
duplicate the effort in a couple months.

When Asrock screwed up my board, I thought it was "just one
of those things". When I see carelessness a second time,
then I've got to say something... For my Asrock board, I
used a hacked BIOS a guy in Germany did (and that is done
at the level of "enabling features" - again, no source code
is available and most of the typical end-user-hackers aren't
assembler/binary editor people).

Any of the other major players should be a safer bet.

Things you should do:

1) *Always* download the motherboard manual in advance.
Does it have a DRAM section for BIOS settings ?
Does it look like there are enthusiast voltage settings,
if something isn't stable. A shabby manual might have
only 40 PDF pages. A midrange manual might have 120 PDF pages,
plus the 120 page one has a BIOS section. This does not mean
the documentation is "excellent" - it merely means they
put more of their "token effort" into it :-) Like, sometimes,
we get enough info to figure out the PCI Express wiring and
which slots get disabled when you do this or that.

2) Check Newegg or Amazon comments on the product, in the
customer review section. I find for the electronics I buy,
several of the observations were spot on. When they said
a certain brand of hard drive "made a strange noise", when
I bought one, well damn, it made a strange noise and I knew
exactly what they were referring to.

Use the available info sources to reduce the risks in your build.

Paul

[MaxTheFast]

Thanks for the new infos.
Well if the MB's manual is the MB's fingerprint I think I'd buy the MSI:
- 100 total pages VS asus's 30
- 20 pages about BIOS VS asus's 4
- 4 pages about OC VS asus's 0
Therefore as far as I see MSI'd be a better bet while ASUS'd be an almost blind choice, right?

Maybe I missed something because it's my first time so here're the links so you can check it out:
- ASUS
https://www.asus.com/Motherboards/PRIME-B450-PLUS/HelpDesk_Manual/
https://dlcdnets.asus.com/pub/ASUS/mb/SocketAM4/PRIME_B450-PLUS/E14214_PRIME_B450-PLUS_UM_WEB_060418.pdf
- MSI
https://www.msi.com/Motherboard/support/B450-A-PRO#down-manual
http://download.msi.com/archive/mnu_exe/E7B86v2.0.zip

[Paul]

Of the two you've shown me, the MSI admits to "DRAM voltages"
in the BIOS section. While some of the features might be
interpreted as "enthusiast", they also serve to help people
running "stock". For example, on the machine I'm typing on,
one day out of the blue, I started getting RAM errors.
By bumping up Vnb (Northbridge) a couple notches, stability
returned. It seemed like perhaps the Northbridge was
suffering from electromigration or some similar kind
of parametric shift. That was about four years
ago, and the board is still running.

And that's one of the things we look for, via the BIOS
screens, is "features not documented via other means".

If the manual doesn't have a BIOS section, it doesn't
necessarily mean the features are missing. Maybe the
Asus happens to have DRAM voltage adjustment too. But
if the manual doesn't confirm it is present, how
do we know for sure ?

I can give another illustration. The feature "ECC" for protection
of memory, is a particularly slippery feature. Industry
"rumors" will say "yadda yadda this board will have ECC".
*But*, if you don't see a picture in the manual of the
ECC screen, virtually 100% of the time you will be ripped
off, and ECC won't be there.

On my current motherboard, the Northbridge was supposed to
have ECC. But there was some waffling about which kind
of RAM (chipset supports DDR2 and DDR3), and ECC might
not work with one of them. On some BIOS, the ECC screen
will simply not even appear, unless ECC RAM is installed.
So Paul forks out for some ECC RAM and... no BIOS entry.
Nada. It's not there.

This is why the manual is *so* important. If you had
a must-have feature, and the motherboard "contract"
doesn't show it's there, you really have no reason
to complain. It doesn't matter how many false industry
rumors there are, it's the manual which is your final hope.

The MSI manual has "A-XMP" which stands for "AMD version
of Intel XMP". This automatically operates the RAM at
the spec speed. If your processor isn't rated for that
speed, it might not be stable. But as long as you stick
with two sticks, for reduced bus loading, that
gives you some small chance it'll work.

Now, another smart thing you did:

HX432C16PB3K2/8

https://www.kingston.com/dataSheets/HX432C16PB3K2_16.pdf

You've bought yourself two sticks of *single sided* RAM.
This gives absolutely the lowest bus loading, and makes
it more likely your RAM purchase will work with the A-XMP
switch. If TEAM corporation was making enthusiast RAM,
that's how they'd make it, with the single sided config.

If it doesn't work (like the RAM on my other machine
didn't, when the bus was heavily loaded), you can
still tune the numbers and "make" it work. No shame
in that. Plenty of options. We can't see the detailed
DRAM screen (which would have a ton of settings in it),
but since they claim to have a screen, there is probably
one present.

Some motherboards even have an "auto-overclocker", but
for the most part these are not all that good. In
some cases, they apply way too much voltage (the
equivalent of using a hammer). If they built a tuner
that really worked, it would need to run all night,
and it would need a non volatile storage to keep
the results between runs. Real hunting for stability
takes "time and small settings steps". In fact, you
purposely search for "instability" such as a RAM error
every ten minutes, as you "walk the edge of the settings".
The idea is, once you've constructed a curve for the
hardware, you then "back off a bunch", knowing that
now the error rate will be zero. Then you do a
Prime95 run (on Linux so your Window disk won't
get damaged) for 16 hours and prove it's good.

Paul

[MaxTheFast]

Thanks a lot Paul for your time, I've learned so much about that.

I'd like to ask you a last opinion about this Gigabyte MB which is within my budget as well:
Gigabyte B450 AORUS PRO (rev. 1.0)
https://www.gigabyte.com/Motherboard/B450-AORUS-PRO-rev-10#support-manual
http://download.gigabyte.eu/FileList/Manual/mb_manual_b450-aorus-pro-wifi_1002_e.pdf
48 total pages (+ another separate 27 pages PDF of useless "unique features"), 13 bios pages, a couple of pages about "MIT" (advanced settings such as XMP, etc), but there's only the following about DRAM:
"CPU Vcore/CPU VDDP/DRAM Channel A/B Voltage/+3.3V/+5V/+12V/VCORE SOC - Displays the current system voltages" so nothing explicit about DRAM manual setting though I know from forums this is a MB for "enthusiasts".
Moreover its QVL contains HX432C16PB3K2/16 (as the ASUS') so one of my HX432C16PB3K2/8's "twins".

Anyway I've got the 2 asrock Mbs' manuals as well and they seem pretty well done:

- asrock Fatal1ty B450 Gaming K4

https://www.asrock.com/MB/AMD/Fatal1ty%20B450%20Gaming%20K4/index.asp#Manual
http://asrock.pc.cdn.bitgravity.com/Manual/Fatal1ty%20B450%20Gaming%20K4.pdf
89 total pages, 22 "UEFI SETUP UTILITY" pages, a couple pages of "OC Tweaker Screen", in particular: "DRAM Voltage - Use this to select DRAM Voltage."

-asrock b450 pro4
https://www.asrock.com/mb/AMD/B450%20Pro4/index.asp#Manual
http://asrock.pc.cdn.bitgravity.com/Manual/B450%20Pro4.pdf
82 total pages, the rest is as above

Of course the "quantity" of pages doesn't matter in these cases because fonts are huge! One thing about both asrocks impressed me: in both bios sections (eg: https://www.asrock.com/mb/AMD/B450%20Pro4/index.asp#BIOS) we can read "We don't recommend users to update the BIOS if their system is already running normally". As far as I know enthusiasts update bios to improve features too, especially for OC, so why this message? Just for newbies or it's a general matter of asrock bioses reliability, in the sense "our bioses are so bad that you'd update yours only if you've no other solutions to carry your system on"? That has came to my mind as you wrote about your experience of "hacked" asrock bios.

So what MB out of those 4 can you finally suggest me for the "AMD overclocker easiest life" after these last comparisons?

See you and thanks again for sharing.

[Paul]

They all support a 105W processor, so the VCore goes at least
that far. We have to check on AMD boards, because of this one
incident where AMD told the motherboard makers to use 65W VCore
circuits, making one generation of boards underpowered for the
high end CPUs. This doesn't appear to be a repeat. You can still
find high definition pictures of each of your quarry, to count
phases and *compare the size of the cooler*. The mistake I made
on my last purchase, is the cooler on VCore was too small, and
I had to fit a blower next to VCore to help out.

https://www.asus.com/us/Motherboards/PRIME-B450-PLUS/HelpDesk_CPU/

Ryzen 7 2700X(3.7GHz,8C,L3:16M,105W)

https://www.asrock.com/MB/AMD/B450%20Pro4/index.asp#CPU

Ryzen 7 2700X(YD270XBGM88AF) 105W

https://www.asrock.com/MB/AMD/Fatal1ty%20B450%20Gaming%20K4/index.asp#CPU

Ryzen 7 2700X(YD270XBGM88AF) 105W

https://www.gigabyte.com/Ajax/SupportFunction/Getcpulist?Type=Product&Value=6635

Ryzen 7 2700X 8C/16T 3.7GHz/4.35GHz 4MB 16MB N/A Pinnacle Ridge 12nm B2 105W

The two Asrock boards look nearly identical, and have a VGA connector
on the back. I can find at least one discussion thread, which mentions that
the VGA doesn't work out of the box with say a 2200G or a 2400G,
and you cable up with say HDMI first, go into the BIOS, *turn on*
the VGA, then move the connector over to the VGA :-/ Not
as much of a slam-dunk as you'd think... A trip to
pain in the ass ville. (You can get DP to VGA and
HDMI to VGA active adapters at the computer store, if
you're stuck. I'm using a DP to VGA on my new video card.
The video card companies are laughing themselves
silly this year, as we all go around buying adapters
adapters adapters to fix this.)

The PCI Express lane count is limited on the hardware, and
so far, I don't see any signs they wasted any additional silicon
on PCIe switches. I'm mostly seeing x16 on one
slot and x4 on a second slot, exactly as it comes out
of the CPU. Sometimes, a company might split to x8,x8,x4 say,
by using bifurcation logic outside the CPU (CPU supports the
x8/x8 interface option, but lacks sufficient pins to do all the lane
switching inside the CPU itself). This has largely
stopped, because the company that used to make cheap components
for that purpose, was swept up and crushed by a larger company,
meaning the "bottom layer" of components is kinda missing.
It means some unique features enabled by moving some
lanes around, is less possible than it used to be.

This also means, if you check the "SATA versus M.2 usage table",
the sharing of some chipset wires will work the same way on
all the products. You might plug in an M.2 and some SATA
get turned off.

This article has a table, showing what interfaces are included
on various chipset solutions on AMD.

https://wccftech.com/amd-b450-motherboards-official-launch-msi-asus-asrock-gigabyte-roundup/

https://www.anandtech.com/show/13091/analyzing-b450-for-amd-ryzen-a-quick-look-at-all-the-motherboards

Some of the boards in your list, have had reviews on Tomshardware,
so you can get info on nuances or quirks that way.

I have a feeling though, of the four PDF files I have sitting
in front of me, there aren't really a lot of "hidden features".
Just keep your eyes open, and see if you agree with
which USB3 interface was placed on the USB C connector
and so on. (I have no use for USB C, but you might.)

The two Asrock boards have practically identical component
layout (even if no board manual shows a realistic picture
of VCore and the VCore heatsink). As a result, you track
down photos at a decent resolution and "count phases"
or make guesses on whether the heatsink is big enough.

The computer I'm typing on, there's like 20-30SqIn of
heatsinks on the thing, for the chipset and VCore. This
is the good old days, where "the thing ran luke warm".
The sinks are a lot smaller now, but you can still compare
the sinks, see if the sinks are tied together with heatpipes
and so on.

My newer motherboard, with the inadequate VCore, cost $250,
but because that was the cheapest board "in the range", they
still saw fit to cut corners. And I missed that while shopping
for the thing, and didn't notice until bench testing
in the kitchen, that VCore was way too hot (65C while running
Prime95). That's one of the reasons you bench test before
putting the motherboard in the computer case, so you can
put your fingers on stuff... and burn them :-/ But, I could
fit a fan, make some BIOS adjustments and that's under
control now.

You still have some research ahead of you, because some
of those connectors on the back could be irrelevant to
your situation.

Paul

[MaxTheFast]

I've just bought the MSI B450-A PRO because its price went down again within my budget, I just hope this will be a good purchase for my configuration and will keep good for the longest period, I mean reliability.
This is a hi-res pic of the MB's upper side (about 10MB):
https://images.anandtech.com/doci/13091/msi-b450-a_pro-2d.png

I've found the following pages on this msi's phases management:
https://www.hardwareluxx.de/community/f12/pga-am4-mainboard-vrm-liste-1155146.html
https://nl.hardware.info/reviews/8541/7/amd-b450-review-9-betaalbare-am4-moederborden-vergeleken-stroomvoorzieningen-fase-choke-en-mosfet-hoe-zit-het
where we can read this:
PWM-controller: Richtek RT8894A (4+2)
echte Phasen (real phases): 4
highside MOSFET: 2x 4C029N
lowside MOSFET: 2x 4C024N
The 2nd linked page shows a comparative table between this MSI, ASRock Fatal1ty B450 Gaming K4 and Gigabyte B450 Aorus Pro but it's arab for me so I just hope I've got a "peer-to-other" MBs if it's not the best one among the ones I've pointed to.
What can you say about those data about phases?

Yeah I knew and "studied" the "VGA matter" about some cfg. AMD cpu + MBs. Here're some useful informations:
https://forum-en.msi.com/index.php?topic=312357.0
especially:
"My understanding is that on the newer 400 series chipsets, you should be able to use either the HDMI, VGA, or DVI-D port(s).
There's an issue on the older 300 series chipsets when using a Athlon 200GE whereby you may not be able to use the VGA due to how the pinout's have changed on the CPU (AMD's fault there....) and how it's supported.
So....with that being said, the 400 series chipsets shouldn't be a problem support wise. As long as the BIOS is up to date, it should work."
This is a real problem for me because I'd like to build this new PC using an old VGA 1440x900 @60Hz Acer monitor and I've NO monitor carrying on HDMI port! As far as I could understand I'll manage the VGA connection between MSI's MB and Acer by the standard VGA cable only if the MSI will came with a newer bios out of the box. If not I'll have to use a HDMI->VGA active converter as you suggested. I think a HDMI->VGA converter would be the best choice for this situation because I couldn't find any problem about HDMI output around the web while lots of troubles about VGA, DP, DVI-D outputs on those configurations. Due to that I've already bought this converter:
https://www.amazon.com/dp/B01GJO6HPW
So my strategy is: if VGA doesn't work out of the box I'll use HDMI port + converter.
There's only a last matter I couldn't find any answer about: what kind of HDMI signal output will came out from the MSI out of the box? My concern is the msi will "shot" HDMI output at its max values:
"HDMI 1.4 port, supports a maximum resolution of 4096x2160 @30Hz, 2560x1600 @60Hz"
In the case there's the converter I'm afraid it will convert the HDMI max output to VGA values that are too high to be displayed on the Acer monitor and I'll get black screen. In this case I'll run for mayor of "ass ville" (as you said) and I'll have to sell the Acer and buy a used HDMI monitor because my budget is null for now.
Is what I said correct or did I make mistakes in my "studies" and solution strategies? Can you possibly suggest other solutions about that?

[Paul]

The RT8894A is more or less "designed for AMD". Richtek
is traditionally a "low frequency SMPS" maker, not that this
matters. My AthlonXP board had one. Richtek can do around
30-35W per phase (again, not that modern and older are comparable
but at least they've demonstrated the capability in the past).
When RichTek did a two phase, others were doing three or four
phase around the same point in time.

https://www.richtek.com/Products/Vcore/Vcore%20Controller/RT8894A?sc_lang=en&specid=RT8894A

Their diagram is a bit weird. The chip is supposed to be a 4+2,
yet the right hand side four phases, three phases are direct drive
and the fourth phase uses an RT9624A. The purpose of those eight
pin "pre-drive" chips are as a buffer, separating the 3000pF MOSFET
gate capacitance from the main regulator chip. When you use pre-drive
chips like that, it makes the voltage regulator run cooler. The pre-drive
chip gets warm, but it's separate from the main chip. You can cool
regulator chips by using a thermal slug on the bottom and
soldering the chip bottom to the motherboard. It's better if all
the phases are buffered, but you can see in the design of the
chip, the chip architecture is designed to hit a price point.
so it's a compromise between lower running temperature, and the
cost of buffering up all the phases. You can probably drive three
phases of large MOSFETs without boiling the main controller.

https://www.richtek.com/~/media/Richtek/Products/ProductSpecs/RT8894A/en/Version4/57744ctqaa.GIF

It's better if the MOSFETs have a good sized heatsink. The chip has
sensing, and can probably sense phasing properly. If I could find
the PDF datasheet, I could check if the stuff on the left is
intended for thermistor input (as then the circuit can monitor
operating temperatures). On a P4 board I own here, they had
a regulator with temperature compensation capability and they
"skipped" using the thermistor, and on that design, that adds
around a 50mV error during CPU load step changes (might affect
attempts to overclock and on an enthusiast class board too). It's
really better if the "optional gubbins" are installed on stuff like
that.

*******

As far as resolution setting, even if an adapter is in the path,
the EDID serial clock and data are passed through to the motherboard.
The motherboard can "read" the EDID. If the EDID says "I'm a
1440x900 monitor", whether it's a CRT or an LCD, the board will
set a resolution according to the "monitor declaration". The
max res a monitor should show, is its "native resolution".
Even if the HDMI puts out 4000x3000, once the EDID is read,
the HDMI mode line will be set to 1440x900 as you would expect.

Where you end up in trouble, is with "projectors". We had one
at the office, a projector that takes a laptop VGA and projects
a picture onto a meeting room screen. On a lot of those, there
is *no* EDID. When no EDID is detected, it's an industry tradition
to not blow up any fixed sync monitors, and they choose values
like 1024x768 or 1152x870 or so. They specifically don't allow
an EDID-free setup to run 4000x3000 by accident. This feature
is both a blessing and a curse of course... a blessing when
some poorly designed display device is not ruined, but a
curse the rest of the time.

It's possible to buy an "EDID faker" box for $50, which you
place between the laptop and the projector. And there are
various programming options via the box. Some boxes, you
can connect them to a regular monitor, and have the
"EDID faker" copy the resolution table in write mode.
Then for the rest of its life, the EDID faker box
runs in read-only mode, telling the world the "projector
runs at 1440x900". I haven't seen those boxes lately and
assume you can still buy them, but the market today
would be smaller. And with the deprecation of VGA as
a standard, a lot of the faker boxes would be from
the VGA era.

The only unanswered question, is why were the damn projectors
made that way in the first place ? Why avoid putting a $2 chip
in a $1000 projector, forcing a customer to spend $50 for a box
and power supply with a $2 chip inside ? :-/ The engineers
who did that should be taken off to the crazy house.

I think you're in good shape on the VGA. As long as the
VGA isn't "outright dead", you now have a flow chart
of what to do about it. And that was my main concern,
that VGA was being put on the motherboard, and in
2019, that connector really shouldn't be on the
back of the system (forcing you to buy the HDMI-VGA
active adapter, right away, instead of it being
a conditional purchase).

Paul

[MaxTheFast]

Hi Paul, I've just built the PC and all seems to be alright so I just want to thank you for you help for RAM and video matters. RAM is compatible but it works @2400mhz by bios default instead of 3200mhz and now I'd like to OC it so I'd want to ask you a little help because this is a "new world" for me: consider I've worked with win xp and dos bios for the last 15 years.

This is my configuration:
- amd ryzen 3 2200g (details: https://www.amd.com/en/products/apu/amd-ryzen-3-2200g)
- mobo msi b450-a pro (manual: http://download.msi.com/archive/mnu_exe/E7B86v2.0.zip)
- ram 2x4gb kingston hyperx predator 3200mhz CL16 (datasheet: https://www.kingston.com/dataSheets/HX432C16PB3K2_8.pdf)

I've just run memtest for the ram @2400mhz and it seems ok: no errors in 43 tests after about 1h time.
What to do now? Do you recommend to upgrade stock bios to the latest release for 1st and then OC or try with this bios for 1st? Consider stock bios is from 2018.6.11 and the latest one is 7B86vA4 (V10.4) released on 2018.12.26 (https://www.msi.com/Motherboard/support/B450-A-PRO#down-bios).

[Paul]

I had started writing an answer, when it occurred to
me we'd already discussed flipping the XMP switch
in the BIOS.

Have you tried that ?

That's the easiest way I know of, of getting there.

From the manual:

"4x DDR4 memory slots, support up to 64GB*

ƒ Supports 1866/ 2133/ 2400/ 2667Mhz (by JEDEC)

Supports 2667/ 2800/ 2933/ 3000/ 3066/ 3200/ 3466 MHz (by A-XMP OC MODE)
"

You have the ideal situation. Installing two sticks
of RAM, should allow just about any cheesy XMP to work.

And since the RAM you bought is likely single-sided
(eight chips on one side), that's even better from
a stability point of view. A single rank of chips like
that on each channel, gives slightly better stability
at high clocks.

Once you try the XMP, and run your memtest86 or whatever,
that should be sufficient to meet a goal of running
the purchased memory at its "stock". Because
the MSI manual claims to support 3200 for the attempt.

It would be an entirely different matter, if MSI had
no intention of ever going over 2400. Then we'd be messing
with BIOS, Overclock Tuners, and other assorted stuff
I'd have to look up :-) I haven't had an AMD board
since my AthlonXP. That was my last one.

*******

If you were going to upgrade the BIOS.

1) Do it within the retailer return period for the motherboard.
Do it immediately. Or don't do it.
2) Find a forum discussing your board, and find out from
other users, how many bricked or didn't brick systems
while upgrading.

Generally, if the BIOS release notes don't indicate any
major improvements, it wouldn't be worth the bother.
However, if you see mention of a new AGESA version,
that might be worth it.

https://www.anandtech.com/show/11447/amd-announces-ryzen-agesa-1006-update

Maybe with AGESA and UEFI, and an OS-based Overclock Tuner,
you could "auto-tune" the board. If so, I'd want to use
a copy of the OS install sitting on an old hard drive,
in case the auto-tune corrupts it. (Some motherboards will
reset the BIOS after three power cycles, saving on
using the CMOS jumper.)

The biggest danger with that sort of tuning, is the
motherboard company getting lazy and using too much voltage
to make it "easy" to prove an overclock will work. When a human
overclocks, they "follow the edge of the curve", then
adjust final settings to ensure some margin and stability.
Whereas some of the poorer overclock softwares, turn up
the voltage into the red zone, just so the overclock only
takes one attempt. Which is stupid and might affect
long term reliability (electromigration). I've seen
signs of electromigration and parametric shift here,
even on equipment that shouldn't have done so. And
at least a "tool or two" in an overclocker forum,
who would say "my CPU won't even run stock any
more, time to Ebay it". So if you're wondering why
people sell used CPUs on Ebay...

Paul