The Top 5 Worst Motherboards of All Time

No mention of Abit"s early jumper-free boards?
Granted, they were innovating, but at the beginning, those pieces of shit kept losing their setup and resetting the CPU settings to default. Lots of returns on that one, when they even left the shop. I loved the idea of jumper-free setup, but being an early pusher of the tech cost me a lot of time.

Of course, as mentioned,
Pc chips, but also as mentioned, everyone knew they were pieces of shit. We used to get them by the box and a 10-20% DOA rate was guaranteed. I used to test the boards before popping them in, to save the work of popping them out when they were bad. Fake cache was also a funny one, until people went to pipeline burst sticks

We have to include anything with a VIA fucking chipset, back in the Windows 95 and 98 days. Holy shit. Paired with Cyrix chips, those cheap pieces of shit were the top of the bottom.

Things got better when Abit got their heads out of their asses and fixed their BIOS issue, VIA put out working drivers and Cyrix went under(or got bought out, don't remember).

Good times, back in my bench tech days, until I figured out there was much better money for an illegal spic handling business networks.

Ps: +1 on nforce 2/3 boards. I loved the hardware, but the drivers sucked, especially sound, then they went away.
 
Last edited:
I'm normally a fan of Asus motherboards, but the old Asus P4T533 (for rambus memory) had serious issues. Pretty sure there's an old forum post about it, but it was a long time ago. I swore off Asus motherboards for a few builds after that, but came back and pretty much only buy Asus now.

ETA: Hardocp review by Brent and Kyle with comments about the big forum thread of issues and my (bmg) comments in it: https://www.hardocp.com/article/2002/08/13/asus_p4t533/6

The old thread doesn't seem to be available any longer, but it was 15 years ago...somehow it doesn't seem all that long ago...time flies.
 
Last edited:
The owner's favorite phrase regarding that was "Do you think the customer did it?"

Blame the customer and never accepted the fact that he had no business owning a computer store.

This owner's name would not happen to be Anu would it? I worked for someone like that for about two weeks before I quit when I caught him in the act of conning an elderly couple. He told them that the computer they just bought from him less than three months ago was not fixable. I had repaired the computer and attempted to talk to the couple, but he kept going on about it not being fixable even after being caught in the lie.

Yes, the 680i SLi motherboard. I remember it well for the pain it caused my tail and wallet. The only good that came of it was pushing me into purchasing the X58 platform right away.
 
No mention of Abit"s early jumper-free boards?
Granted, they were innovating, but at the beginning, those pieces of shit kept losing their setup and resetting the CPU settings to default. Lots of returns on that one, when they even left the shop. I loved the idea of jumper-free setup, but being an early pusher of the tech cost me a lot of time.

Of course, as mentioned,
Pc chips, but also as mentioned, everyone knew they were pieces of shit. We used to get them by the box and a 10-20% DOA rate was guaranteed. I used to test the boards before popping them in, to save the work of popping them out when they were bad. Fake cache was also a funny one, until people went to pipeline burst sticks

We have to include anything with a VIA fucking chipset, back in the Windows 95 and 98 days. Holy shit. Paired with Cyrix chips, those cheap pieces of shit were the top of the bottom.

Things got better when Abit got their heads out of their asses and fixed their BIOS issue, VIA put out working drivers and Cyrix went under(or got bought out, don't remember).

Good times, back in my bench tech days, until I figured out there was much better money for an illegal spic handling business networks.

Ps: +1 on nforce 2/3 boards. I loved the hardware, but the drivers sucked, especially sound, then they went away.

LOL, many Ryzen owners have experienced the settings going back to default if they had them customized.

Review was a fun read which brought back some fond memories or the opposite in this case. I hope Dan does a top 5 motherboard of all time as well. For some positive recollections.
 
I will never own another EVGA mainboard, or anything else they make for that matter.

680 SLi was a turd.
780 likewise.
Had 2 X58 boards...one a Classified that was > 200 dollars which at the time was crazy money.
I got rails of shit trying to RMA that bastard. Got a warped replacement.....

Those guys shouldn't brand mainboards.

The mention of EVGA gives me rage even today.

Similar experience with my 680i RMAs with EVGAs. My first replacement's voltages were off the charts, you could literally, no joke smell the ozone from the Northbridge. My second came with a broken DIMM slot so obvious my grandmother would have noticed. Third blew out a tried-and-true PSU first time I powered up likely due to some short on the board. They were obviously just recycling returns for RMAs with little-no QA. That, or Foxconn was sending them lemons and they weren't testing them at all.

From my experience and what I recall in the EVGA forums (before the mods started redacting posts with even a hint of problems), I recall there was a point where EVGA seemed to toss in the towel and started hardlining customers.

I do recall moving to my next mobowas a glorious experience (Gigabyte P35, I believe, which was by all rights a cheaper, less feature-rich board) only due to the hell I suffered at the hands of the 680is.
 
Similar experience with my 680i RMAs with EVGAs. My first replacement's voltages were off the charts, you could literally, no joke smell the ozone from the Northbridge. My second came with a broken DIMM slot so obvious my grandmother would have noticed. Third blew out a tried-and-true PSU first time I powered up likely due to some short on the board. They were obviously just recycling returns for RMAs with little-no QA. That, or Foxconn was sending them lemons and they weren't testing them at all.

From my experience and what I recall in the EVGA forums (before the mods started redacting posts with even a hint of problems), I recall there was a point where EVGA seemed to toss in the towel and started hardlining customers.

I do recall moving to my next mobowas a glorious experience (Gigabyte P35, I believe, which was by all rights a cheaper, less feature-rich board) only due to the hell I suffered at the hands of the 680is.

Foxconn has taken a lot of flak for those boards, but it's not their fault. Or at least, not entirely. Foxconn was just the company contracted to build NVIDIA's reference design. NVIDIA's design sucked ass. If you hand Foxconn a good design, they'll build you a good board. You hand them a shit design, they'll built you a shit motherboard.
 
Honestly, the worst board I ever had was an 939 Era Nforce4 board by Asus. I believe it was the Asus A8N Deluxe.

That thing had the worst NB fan ever to exist on a high-end board. I believe i killed THREE of them too. Once, the Winbond BIOS chip separated from the board. One it just died, and another time the Chipset died when I switched out the 40mm HSF.

One time I had a board sent back from RMA (for the Winbond de-bonding) with just SMT components, it was pretty rad, but the 3 week turnaround to Canada really pissed me off. No caps or larger components. Just BGA and SMT.
 
I went through quite a few DFI boards back in the day as well, just brought a chill down my spine......oh crap now my eyes are twitching..
 
This is the first HardOCP article I've ever LOL'd before! I can totally feel the disgust in the article! Short and to the point. Thanks for this!

I've always wanted a DFI board back in the day. Glad I never got one.
 
I had this piece of sh*t KT266 ECS Socket A board for my 1900+.

I got it because Fry's had a killer combo for the proc & board for like $200 back in summer 2002.

For an 18 year old kid, this was an epic deal!

Had to return 3 of them before I got a working board.

It was actually an ok mobo for a while, UNTIL I decided to transfer all the guts into a newer case.

Bricked the board somehow.... Maybe static electricity, but I sincerely doubt it.

Whoops.

Replaced it with a 3500+/Gigabyte combo in '04.
 
I had a EVGA 750i SLI and it was also a giant piece of shit.

We need a top 5 motherboards now, #1 Abit BH6 obviously.

Asus P2B (440BX!). I was rocking a 133Mhz bus from day one on that board before 133Mhz was a thing. P-II 350 running at 133x3=400. Back when you could manipulate the multiplier on Intel but only downward. On 24x7 rock solid until the day it was replaced.

I think I replaced it with a Asus CUSL2 and ran that at 150Mhz bus with Kingmax BGA ram. First time I had seen BGA Ram.

Man. Good ole overclockin days. :)
 
Honestly, the worst board I ever had was an 939 Era Nforce4 board by Asus. I believe it was the Asus A8N Deluxe.

That thing had the worst NB fan ever to exist on a high-end board. I believe i killed THREE of them too. Once, the Winbond BIOS chip separated from the board. One it just died, and another time the Chipset died when I switched out the 40mm HSF.

One time I had a board sent back from RMA (for the Winbond de-bonding) with just SMT components, it was pretty rad, but the 3 week turnaround to Canada really pissed me off. No caps or larger components. Just BGA and SMT.

I liked the A8N-SLI Deluxe. I had one and it was great. I will say that the north bridge fan did suck ass. That's a fact.
 
Ouch, I think this thread explains a lot....

My one friend had nothing but issues with computers being finicky for the longest time. He had a socket A soyo board (Not the exact same one at least), and the Evga 680i after that. The last system we setup has actually been decent, but it took over 10 years before he had a stable one... (He didn't replace systems every other day like [H] does) I will say I can completely back up what Dan said though, the 680i was fine with the C2D, when we dropped a C2Q is when it started having issues. I have 3 780i boards and other than the 25 Watt north / southbridge combo, they work fairly well.

I also have the DFI Expert in this thread, and I'm still not sure what I did to it, but I managed to break it. You can get it to light up if you short out the power on pin, but it won't post. I thought this board was a bit better than others, but I guess I'm not the only one who's had issues with theirs.
 
I still have that DFI board. I got one because it was supposed to be the best board for overclocking the Opty I paired with it but I've got to say, building a useful PC out of that board was soul crushing. Every time you tweak something for an overclock, you would inevitably have to return the CPU to stock clocks to see if the tweak you made is the cause of instability. It took months to get a stable system, overclocking or no.
 
Foxconn has taken a lot of flak for those boards, but it's not their fault. Or at least, not entirely. Foxconn was just the company contracted to build NVIDIA's reference design. NVIDIA's design sucked ass. If you hand Foxconn a good design, they'll build you a good board. You hand them a shit design, they'll built you a shit motherboard.

Back when I did helpdesk there was a way higher than usual number of mobo and psu failures with Dell Optiplex 270s. Foxconn was the board manufacturer (and had been for Optiplexes for a while).

*opens computer case in front of user, points to burn mark on northbridge*
"well, there's your problem"
 
Asus P2B (440BX!). I was rocking a 133Mhz bus from day one on that board before 133Mhz was a thing. P-II 350 running at 133x3=400. Back when you could manipulate the multiplier on Intel but only downward. On 24x7 rock solid until the day it was replaced.

I think I replaced it with a Asus CUSL2 and ran that at 150Mhz bus with Kingmax BGA ram. First time I had seen BGA Ram.

Man. Good ole overclockin days. :)

Oh yeah, that was a great board and a great chipset. Just saying "440BX" gives me a warm feeling.
 
I would say the ASUS A8N-SLI Deluxe & Premium could give the DFI LanParty a run for it's money. At work we used these in most of our 939 builds and I had a Stack taller than me of bad boards in my office by the time Conroe came around.

https://www.newegg.com/Product/Product.aspx?Item=N82E16813131517
https://www.newegg.com/Product/Product.aspx?Item=N82E16813131540

Ah the N-Force solutions. Hmm did any maker have a "Solid" N-Force board that didn't have problems?
 
When I started, with the company I worked for had many of the SOYO SY-6KBE boards installed when I started, nothing but headaches, some worked rock solid no issues (even years after the fans stopped)
The owner thought it wise to buy the next cheap solution up, the FIC 503 boards, come to find out some of the chipsets had the IRQ lines routed incorrectly internally, and the constant updating driver from VIA was trying to fix that. (granted the tech who built the systems overclocked everything and built as cheaply as possible)
The problem was he was using these boards to build client systems, then pay tech's to fix them. customers were not happy.

I got the authority to spend $50 more per system and build Intel board systems, the last of which were just retired over 10 years later. Were they $50 savings really worth the investment? data says yes, granted our tech service revenue went down, but our sales went up.
 
Back in the early 2000's I was working for what was at the time the largest PC manufacturer down here in New Zealand, a company rather imaginatively called "The PC Company". The company was ridden into the ground by stupid cost cutting purchasing decisions, none of which stands out to me more than the decision to use ECS K7VTA3 motherboards in a production run. I worked in the R&D department at the time, our job was to test parts for quality and reliability, find any issues and work with the manufacturers to resolve them. We dutifully tested the K7VTA3, told our boss it was a horrible cheap nasty board we should avoid at all costs, and were told "Well just make it work, we've ordered 3000 and production starts with them next week".. I'd say at least 3/4 of those PC's came back for warranty repair because of those boards, and it must have cost the company a fortune.
 
LMFAO

I just knew the 680i and a dfi nf4 dr expert board would be involved, I saw the thread title and the first thing that came to my head before I could even clicked the link was, evga 680i mobo and a dfi nf4 dr x.

Dfi board was a retarded piece of shit that made you wait days to reset the bios, god forbid it you needed to and you did on multiple occassions just to boot the fucking thing up and you would turn the pc off one day only to turn it on the next fully expecting it to work and it just would not start.

Evga 680i, lol you would be happy if it booted up and never fried itself before completing the boot up process.

As for the others, apart from the ud7, fuck knows ?
 
I owned a EVGA 680i SLI. I used it with my 8800 GTX SLI setup until it set its self on fire by the CPU socket and EVGA replaced it with a 780i. I still have the 780i in my closet. I didn't really have problems with both of them. I was able to run my Core 2 Q6600 at 3.5ghz with both of them.

That's a problem.
 
My God...I remember owning that FIC board. Trying to remember if I had issues with it or not...
 
I don't think I have. I've worked with some ECS and other atrocious brands but none of them gave me the trouble that the above list has. Notice that I experienced issues with multiple copies of the above boards. The GIGABYTE 990FXA-UD7 being the one exception there.
Speaking of the Gigabyte board, you listed it as 1998, not 2008.
 
Lol, he didnt have a problem with it until it caught fire.

I know what he means, but its funny the way it translates in text.

I understood what he meant. The thing is, most motherboards don't catch fire. And again, the 680i SLI motherboards were great until they went tits up. I knew people that never had trouble with them. Years ago a forum member here called the 680i SLI the best motherboard they'd ever used.

The people that had good luck with them typically only used them with dual core processors, RAM with average clocks for the day, two DIMMs, one or two graphics cards and that's it. I ran an LSI MegaRAID 8308ELP with dual 8800GTX's, the fastest RAM available at the time along with a Core 2 Quad Q6600. Later I used 3x 8800GTX's with them when NVIDIA kept breaking storage controller support. They made changes at one point to the BIOS which allowed for better overclocking but fucked over anyone who used a PCI-Express controller.

I think it's clear that the VRM design was shit as were the automatic BIOS voltage settings. Everything needed more voltage than it should have. The board ran too hot and even water cooling couldn't stop that. Those boards fried RAM, fried their chipsets and ultimately themselves.
 
Yea, something funky with the power was happening as it damaged too many boards and hardware for it to fall within the law of averages.

Its probably for the best that nvidia left the mobo chipset scene, lets them focus on what they are really good at instead.
 
Yea, something funky with the power was happening as it damaged too many boards and hardware for it to fall within the law of averages.

Its probably for the best that nvidia left the mobo chipset scene, lets them focus on what they are really good at instead.

The funny thing, it was NVIDIA that made the AMD platform decent. Around the same time, AMD and Intel both locked NVIDIA out of the chipset market. I don't think AMD prohibited them from making chipsets as much as it didn't make sense to go with NVIDIA chipsets when AMD started making their own again. Well, once they bought ATI that is. Early AMD chipsets like Xpress 3200 were ATI chipsets. AMD made their own in the past, but stopped with the AMD 760 if memory serves.

NVIDIA wouldn't have done as well as it did in the Intel market were it not for the lockout on SLI.
 
Last edited:
I often get asked what the worst motherboards I've dealt with are, were, or who made them. I've decided to count them down in order of best to worst, and tell you why they suck.

#5 DFI LanParty nForce 4 SLI DR Expert -2005

I expect to catch some shit for this one. This was one of the boards to have in the Socket 939 era. It was feature rich, it had a good layout and it was a monster of an overclocker. At the time I was building systems for a local computer store and I built several machines using this motherboard. I can't tell you how many DOA or problematic copies of this I went through. I dreaded each and every one of them I ran across. I could assemble systems with the same RAM, CPU, PSU and this motherboard and get different results each time. Some were stable at stock settings, but many weren't. This is the crux of my hatred for this motherboard. It was inconsistent regarding its behavior, stability and compatibility with hardware.

#4 Soyo SY-6KBE - 1997?

This motherboard was awful. I built several systems based on this motherboard back in the day for a client. Within a year or so, all of them had developed some sort of fault. Two of them got replacements of the same model only to fail a short time later. The best of these only had two bad PCI slots but otherwise ran for years. The others all lost RAM slots, developed instability issues that couldn't be solved or plain died. Soyo was a brand loved by many, but my experiences with them were mostly negative.

#3 GIGABYTE 990FXA-UD7 - 2011

This is one I reviewed back in the day. The conclusion page summarized my issues nicely. Up until X370, this was the worst motherboard I had reviewed. In fact, reading over the notes in that article, I had a lot of memory compatibility problems echoing my experiences with X370. If you don't want to click the links, this line from the conclusion page should paint the picture for you: "This is probably the worst motherboard I’ve ever reviewed here at the [H]. I can’t recommend it to anyone for any reason as it stands today and feel like it should be avoided at all costs."

#2 FIC VA 503+ -1998

This was an easy call to make. The FIC VA 503+ was based on a VIA MVP3 chipset. This piece of shit had a number of shortcomings, one of which was that it didn't fully support all the K6 processors that were out at the time. Not only that, but the user manual was printed with inaccuracies for the jumper setup. The silk screening on the motherboard was full of errors too. Even the addendum in the user manual was wrong. You had to use a painful trial and error process to make it work. This motherboard also had a dismal DOA rate and a high percentage of failures. I worked for a computer retailer back in those days and we saw countless DIY systems come into the tech shop for service. Worse yet, AMD had their own clone of Intel's Retail Edge program and this motherboard was paired with a processor that the board didn't officially support. Virtually everyone who bought this deal at our store had me setup their systems because no one else could get them to work.

This piece of shit was so bad it had three hardware revisions and never was built correctly. I have no doubts that this motherboard's awful reputation is one of the many nails in the coffin for FIC's motherboard business. The company exists today, but exited the PC motherboard market years ago.

#1 EVGA 680i SLI -2006

This motherboard and the various rebrands of it were built by Foxconn. This was a great motherboard at the time on paper. Unfortunately, these pieces of shit were prone to failure. These motherboards had a myriad of problems. Some of these problems you could work around, but many heralded the motherboard's impending demise. These suffered from poor RAM clocking, overvoltage issues, overheating components, USB issues, and some problems I'm sure I've forgotten. Even watercooling these things didn't help with the heat issues, which I suspect was due to things like the chipset getting more voltage than they could stand. Often times, BIOS updates fixed one thing at the cost of something else. PCIe device compatibility was sacrificed to gain overclocking performance in one of the later iterations.

I personally owned 12 of these things. All 12 of them died within a year's time. Almost half of them were DOA replacements that were received from RMA or bought at the store. Even the hand picked EVGA Black Pearl I reviewed didn't last long. The only reason people tolerated these things, and sold so well is due to the fact that SLI functionality was locked to NVIDIA's motherboard chipsets back in the day. Eventually, I stepped up to Intel's D5400XS / Skulltrail platform to get away from this pile of trash. I had a stack of these I took out shooting one day. Putting a .500 Magnum round through the chipset's fancy heat sink was a very satisfying activity. I hope to never encounter a motherboard this bad again.

IDK man, DFI LAN Party was a great mobo and I didn't have all the issues with the nf4 boards. It was one of the best nf4 boards at the time. I elected to go with Asus and they were pretty big pieces of shit. Given how all the overclocking deal and gaming PCs were only starting to go sort of mainstream I think it was fair to cut them some slack tho. EVGA 880i SLI I also had excellent experience, my buddy is still rocking my old 680i in his old gaming shitbox, still all works. I also still have my old EVGA 780i SLI in my old gaming box which works as a test box. Reason I switched from 680i was lack of support for Q9450 processor. My brother had same exact board and it was rock solid until I built him an x79 rig back in 2012 or so. Never had FIC but used to be a Soyo fan back in P4 days. Still have my Asus P2B from the same circa as Soyo SY-6KBE. Still works with a 400MHz slot P2. I just can't let myself though it away as it was my ultimate Unreal Tournament computer back then with 3dfx Banshee 12MB.
 
I owned the 990FXA-UD7 board.

From day 1, I couldn't get my keyboard to work without being in Windows. In order to get into the BIOS or safe mode or anything that required me to hit a key before Windows loaded I had to plug in my ps/2 keyboard to do it.

I never overclocked it or messed with much but after about 2 months for whatever reason it just simply wouldn't boot up. For kicks and giggles I plugged my ps/2 keyboard up so I could get into the BIOS and do a reset/save and it worked. It left me scratching my head because I couldn't remember ever changing anything. Then every couple weeks it started happening and every single time I would just go in and do a reset/save and it would boot right back up. I could never explain it. I just dealt with it for about a year when I finally couldn't take it anymore and built a new system.
 
I owned the 990FXA-UD7 board.

From day 1, I couldn't get my keyboard to work without being in Windows. In order to get into the BIOS or safe mode or anything that required me to hit a key before Windows loaded I had to plug in my ps/2 keyboard to do it.

I never overclocked it or messed with much but after about 2 months for whatever reason it just simply wouldn't boot up. For kicks and giggles I plugged my ps/2 keyboard up so I could get into the BIOS and do a reset/save and it worked. It left me scratching my head because I couldn't remember ever changing anything. Then every couple weeks it started happening and every single time I would just go in and do a reset/save and it would boot right back up. I could never explain it. I just dealt with it for about a year when I finally couldn't take it anymore and built a new system.

I had to do the same thing with the keyboard.
 
I had a few ECS boards with via chipsets and I can hand on heart say that every single one of them died au naturel.

Well, they were natural but premature deaths.

Pieces of shit.

I was waiting for this response. Thank you. 10+ years ago, when I used to be able to sift through the computers at my local recycling center, I would find countless PI, PII and PIII machines with dead ecs boards. To be fair, sometimes dead caps were the culprit. But man, those boards were just pure crap.
 
My current FIC SD11 board. What a POS. AGP 1.0 supposed to do 1x/2x but only does 1x even with the most recent Bios update.
 
Back
Top