New Affordable AMD B650 Motherboard Roundup

[Steve]

Motherboard makers release revised models of existing boards all the time. But what is unusual here is that all these new additions are entry-level, and released rather quietly. Let's find out why.

Read the full article here.

New Affordable AMD B650 Motherboard Roundup

 

[WhiteLeaff]

There should be at least one test in games. Most people don't use PCs for rendering.

 

[Disiw]

700/800 chipset boards are incoming. Most of these boards are 2 years old at this point.

Ryzen 9000 reveal in 49 days. I bet they will show new boards shortly after.

Would only buy a 600 series AM5 board at this point if 25-50% off and it supports USB firmware flashing.

 

[Theinsanegamer]

It's a real shame that OEMs continue to produce low quality micro ATX boards. For whatever reason this has been a problem with AMD for a LONG tome, back in the phenom II days there were only a handful and AM3+ never got one at all.

 

[Thatsdisgusting]

only buy a 600 series AM5 board at this point if 25-50% off

THE only reason I'd wait for AM5 X770

 

[Disiw]

THE only reason I'd wait for AM5 X770

I will wait because of much better designs. The current AM5 boards are mostly terrible.

I will be rocking 700/800 series chipset + 9800X3D ASAP unless Arrow Lake is any good.

My 7800X3D and board is already sold and will be picked up when I upgrade.

Could not care less about 2 year old boards with terrible designs, that needs firmware updating before booting with a Ryzen 9000 chip inserted.

New and shiny only for me. Not buying a brand new top tier CPU without a brand new board. New features will come for sure.

 

[bviktor]

There should be at least one test in games. Most people don't use PCs for rendering.

Those are less demanding on the CPU than rendering, so the test should be just fine.

 

[bviktor]

Not a single mini ITX board

 

[jpvalverde85]

I had a hard time with the position of the images in respect to the review of each motherboard, but certainly the results of this roundup are terrible, certainly the ones to avoid.

 

[Atmajaya2020]

When I read "affordable", I thought there was one itx b650 that was affordable..

 

[DukeJukem]

Those are less demanding on the CPU than rendering, so the test should be just fine.

I gotta shed some light on this. you're both right but the second guy is being dismissive of the first, which is fine and technically correct so I'll explain why...but most people won't know why you're correct and may assume you're...just not a nice guy. anyways here's the point.

guy number 1 demanding a gaming only test. watch your cpu usage while downloading a steam game, let alone installing something massive. Your cpu usage, if the rest of your hardware and ethernet connection are capable, will shoot up to 100% usage, for a very long time and will mimic a cinebench run, if not more depending how long it takes the game to download/install.

so if you buy a piece of trash board thats failing those cinebench tests and you're just using it for gaming, chances are it'll hit those temps during a massive download or install that properly utilizes your hardware. Even if it's getting too hot for 5 seconds, too hot is too hot.

 

[hwertz]

Holy crap! I thought when you said you were going to show which boards to avoid, you were going to quibble over like a 1-5% difference. That's insane that you can lose like 50% under load! (And 20-30% on other tests!)

And to those who say, "Why test with rendering instead of games?", it's to get it at full load for sustained lengths of time. You'll thank them for this test if you keep your computers for very long; even if a B650 build has a fast enough CPU that current games are not hitting full CPU load on it, in a few years some of them could. In addition, my non-gaming workloads keep getting more CPU intensive year-over-year; browser scripting keeps getting a bit more bloated, more photos, animations, and videos each year. Larger resolutions, larger data sets, more virtualization, more applications running at a time. Of course if you're stuck with Windows they're now going to start throwing AI junk onto your system (probably running on the CPU since they seem to be favoring using a "NPU" or the CPU rather than using GPU offload.)

In years past, certain vendors *cough*HP/Compaq*cough* just LOVED to undercool their systems, like "This P4 is so fast, it won't be maxed out for long anyway. Let's just put like half the cooling in there that it needs, why not!". This was distressingly common up to the Core 2 Duo/Core 2 Quad era, they viewed the throttling that was (in my opinion) put in as a failsafe as a license to just undercool the system and let the throttling take care of it (to be clear, back then they didn't have boost clocks, so this wasn't "the CPU not reaching full boost"... it's the CPU dropping below it's rated speed to keep from melting). Well, even 4, 5 years later, that was a joke, just regular tasks were CPU-intensive enough to just keep the thing pegged out indefinitely and it'd start throttling 10, 20, 30%.

 

[BobHome]

Sounds like some bean counters said, 'We can get these older off-brand parts cheap, what can we do with them?'

 

[someusernam]

When AM5 came out most motherboards had overkill VRMs. So instead of designing new motherboards that have good enough current handling capability without overkill to reduce cost. All four vendors went and made matx mobos that are barely good enough for ryzen 8000; a series of CPUs that only makes sense in small form factor builds. Sigh.

 

[mbk34]

"How can you go wrong?" was exactly what I thought when I bought an MSI PRO A620M-E motherboard last year. I've built lots of PCs over the years and wasn't expecting any issues this time round. What I found was the motherboard placed the GPU slot at the bottom of the board directly over the PSU meaning the exhaust from the GPU blows directly into the PSU exhaust with a gap of a few millimetres (1/8 inch) between them. This was also the place on the motherboard where nearly all the connections are making it a pain to set up or add new components. All these wires obviously get in the way of what little airflow there was. The memory slots only had clips on one side and I found fitting RAM quite difficult to past builds but perhaps it saved them 5c. On the plus side it did have something called EZ debug which are debug lights telling you if anything is badly connected - unfortunately they all blinked an alarming red when I first built the PC and nothing posted so, after a few rebuilds, I just returned all the components for replacements. It was only later that I found out they stopped showing red after a minute or so. Hey ho.

My build works but I'm forced to run the GPU (6650X) in economy mode to stop it overheating. I don't play AAA games so it's not a big deal but I won't be buying another MSI board.

 

[Avro Arrow]

Steve, I would first like to thank you for the hours that you put into articles like these. You sir, have the work ethic of a bulldozer!

I do have a question to ask however. It's clear that none of these boards do well in CineBench with an R9-7950X but do you think that they were really meant to?

Under less strenuous circumstances, these boards just might shine. I wouldn't ask you to redo all of your testing with a different CPU (because that would be unreasonable to ask) but how about testing only the three worst of these boards with something like the R7-7800X3D with CineBench and a few CPU-intensive games like Starfield, Hogwarts: Legacy and CP2077? Hell, even try it out with a few rounds of Fortnite (which you'd probably do anyway).

Using a relatively low-power CPU that doesn't overclock like the R7-7800X3D might change your opinion of these boards from "Avoid at all costs" to "Great value for gaming but garbage for productivity".

These might turn out to be the better-priced budget AM5 motherboards that so many gamers have been waiting for. If the worst three actually do alright on the scenario I proposed, then it would be safe to say that the rest would be ok as well.

Even if you don't do the gaming tests, just the worst three boards doing CineBench or Prime95 with a 7800X3D might show that these boards are situationally decent. I could be wrong, I'm just theorising here but I think that little test might do a great service for budget gamers looking to jump onto AM5 but have found the entry price to be somewhat intimidating.

Thanks again for all you do.

 

[Athlonite]

Why would you pair an R9 7950X with a low end mobo that's just ridiculous the only CPU's that should be going into these are APU's or 7600/X 7700/X or a 7800X3D at most anything over that and you wasted your money on a top tare CPU

 

[Steve]

Steve, I would first like to thank you for the hours that you put into articles like these. You sir, have the work ethic of a bulldozer!

I do have a question to ask however. It's clear that none of these boards do well in CineBench with an R9-7950X but do you think that they were really meant to?

Under less strenuous circumstances, these boards just might shine. I wouldn't ask you to redo all of your testing with a different CPU (because that would be unreasonable to ask) but how about testing only the three worst of these boards with something like the R7-7800X3D with CineBench and a few CPU-intensive games like Starfield, Hogwarts: Legacy and CP2077? Hell, even try it out with a few rounds of Fortnite (which you'd probably do anyway).

Using a relatively low-power CPU that doesn't overclock like the R7-7800X3D might change your opinion of these boards from "Avoid at all costs" to "Great value for gaming but garbage for productivity".

These might turn out to be the better-priced budget AM5 motherboards that so many gamers have been waiting for. If the worst three actually do alright on the scenario I proposed, then it would be safe to say that the rest would be ok as well.

Even if you don't do the gaming tests, just the worst three boards doing CineBench or Prime95 with a 7800X3D might show that these boards are situationally decent. I could be wrong, I'm just theorising here but I think that little test might do a great service for budget gamers looking to jump onto AM5 but have found the entry price to be somewhat intimidating.

Thanks again for all you do.

Thanks mate. So with this testing you have the maximum sustained package power for each motherboard so you can easily work out what is the most powerful CPU each board can handle.

Naturally if you use a low-end CPU they will work, but that's not really the point for a few key reasons discussed in the article. For example...

1: They're advertised to work with Ryzen 9 processors (they don't).
2: They're advertised to offer exceptional power deliver and in some instances superior overclocking (neither is true).
3: They don't allow you to upgrade to a more powerful AM5 processor in the future, which is kind of the point of AM5.
4: There are boards such as the Asrock HDV which cost the same amount or less and allow you to use any Ryzen processor, Ryzen 9 included.

So really no matter which was you slice it these boards are trash, they do nothing to improve the AM5 platform, in fact they do the opposite, they're bad for consumers.

Why would you pair an R9 7950X with a low end mobo that's just ridiculous the only CPU's that should be going into these are APU's or 7600/X 7700/X or a 7800X3D at most anything over that and you wasted your money on a top tare CPU

The reasons really should be obvious, and I'm being blunt here because the reasons were explained in the content you're commenting on. We're doing you a favor here by testing correctly.

There should be at least one test in games. Most people don't use PCs for rendering.

Yeah which game? One that places heavy load on the CPU or one that uses a single core? Also would you like me to wait 4 years for when games are much more CPU demanding and you might want to install a more powerful CPU on your AM5 motherboard... because you'll almost certainly be able to. I bet then you'll wish you spent $120 on a good AM5 board and not a trash AM5 board.

 

[NumberNine]

I bought the ASRock 650m board not too long ago for a build, it had coil whine. Sent it back and went with ASUS ProArt, problem solved.

There's a small bit of real info for you.

 

[yesnt]

I gotta shed some light on this. you're both right but the second guy is being dismissive of the first, which is fine and technically correct so I'll explain why...but most people won't know why you're correct and may assume you're...just not a nice guy. anyways here's the point.

guy number 1 demanding a gaming only test. watch your cpu usage while downloading a steam game, let alone installing something massive. Your cpu usage, if the rest of your hardware and ethernet connection are capable, will shoot up to 100% usage, for a very long time and will mimic a cinebench run, if not more depending how long it takes the game to download/install.

so if you buy a piece of trash board thats failing those cinebench tests and you're just using it for gaming, chances are it'll hit those temps during a massive download or install that properly utilizes your hardware. Even if it's getting too hot for 5 seconds, too hot is too hot.

Steam download and install is nowhere as demanding as rendering. Steam will max maybe two cores, and rendering can max them all.

Your "light" is dim and points in the wrong spot.

 

[Resonator]

Thanks much for the informative article!

Just after the holiday season this year I helped my brother-in-law put together a basic gaming system. He wanted to keep it in the $1000 ballpark, so I ended up going with the ASRock B650M Pro RS as it was only $130 at the time, we didn't need wi-fi or the PCIe 4.0 x1 slot (or really the PCIe 3.0 x16) and it had four DIMM slots instead of two. I'm glad to see it performs very closely to the recommended ASrock B650M-HDV/M.2. All my previous builds have been based on Asus MBs and before that (dating myself here) A-Bit, but given a lot of recent negative press about Asus QA and a lot of positive reviews about ASRock, I decided give them a try.

I managed to put it together (sans monitor, keyboard & mouse) with a 7800X3D CPU, 32GB of G.Skill Flare X5 RAM, a 1TB M.2 SSD, NZXT H5 Flow Case & C750 PSU and an RTX 4060 for about $1200. As it's mainly going to be running older and retro games and maybe an occasional AAA title, it should be more than sufficient. I was even able to save a few extra bucks by using a new spare Wraith Spire heatsink I had kicking around. At first I was concerned it wouldn't be enough, but with the great airflow of the H5 case (I added a couple additional 120mm fans since I had them) I had no issues with keeping the temps reasonable when stress testing & benchmarking.

 

[takaozo]

1: They're advertised to work with Ryzen 9 processors (they don't).
2: They're advertised to offer exceptional power deliver and in some instances superior overclocking (neither is true).
3: They don't allow you to upgrade to a more powerful AM5 processor in the future, which is kind of the point of AM5.
4: There are boards such as the Asrock HDV which cost the same amount or less and allow you to use any Ryzen processor, Ryzen 9 included.

First let me thank you for all your hard work.

My question is related to new Zen 5 cpu's, are we expecting 125-350W parts only?

For me a "a more powerful AM5 processor in the future" can be a 9600x or 9800x 3D cache or not and if they keep it under 90W should be fine.

Cheers mate!

 

[Kam7r]

When I read "affordable", I thought there was one itx b650 that was affordable..

No love for us SFF enjoyers

 

[koblongata]

Wow, didn't know mainboards can make such differences.

 

[robert40]

Where is the coolest motherboard?

MSI B650M-PROJECT-ZERO

no cables on the front !!!

https://www.msi.com/Motherboard/B650M-PROJECT-ZERO