Old, but interesting read I found on 486 motherboards w/ fake and missing parts

[DaveLembke]

I was looking up a PC Chips board that I had a ways back and was curious what ever happened with PC Chips manufacturer. Looking into where they are now, bought out or belly up and found this article rather interesting. The PC Chips board I had was a 386 SX 16Mhz. Found this article interesting in how low end boards could be ordered with false identification labels making a no name chipset look like INTEL because it says Intel and who would question a chip label. 

Also interesting that the cache on some boards back then were not populated or populated with fake chips. I remember having a 486 board with sockets for cache that wasnt populated which i got cheap ($40 486DX33Mhz with 16MB RAM ) from a guy who upgraded to a Pentium and at the time I didnt know that it was a severe crippler of the 486 board because I didnt have anything faster to compare it to and it did run faster than my 386. All I know is that it ran my mostly DOS games better or same as the 386 I had. Also way back in the day I ran SETI on the 486 and it took 59 hours to complete 1 work unit after upgrading to a 66Mhz DX2. Today it would likely take a few months to complete a work unit with SETI with a 486 if SETI through Boinc would even run on a 486. 

Cringed at the suggestion of flexing a PCB to test quality, but he did say its not good for boards. 

http://redhill.net.au/b/b-bad.html

[BC_Programmer]

I don't think the thickness of a motherboard is a direct indicator of quality. Especially comparing to a Gigabyte board- they've been intentionally thickening their boards for that reason for quite a long time, calling it their "Ultra Durable" technology. There are certain brands you just avoid altogether, I think. Chips, BioStar, and for a  time MSI, for example. (Somehow, MSI went from low-end scrap dealers to making high-end Gaming stuff though...)

I watched a youtube video from somebody who mostly makes content about older motherboards CPUs, benchmarks them, etc. and I seem to recall coverage on several PC Chips boards, including finding that numerous chips didn't serve any function or had traces that went literally nowhere, including the alleged cache chips.

PC Chips and Technologies also made graphics chipsets, for that they were a little bit better, but still far from perfect. My Toshiba Satellite 440CDX uses a Chips & Technologies 65554 Display Adapter and it has a few quirks.

The first quirk is that it claims to support things like OpenGL, but it doesn't; Or rather, it supports openGL, but doesn't support basic features, most notably texture mapping, nor does it run quickly. Even back then Quake at 8FPS without any textures was something I found unplayable altogether. It's possible this was specific to the implementation on my laptop since from what i can find the same Chips 65554 was used in the Powerbook G3 which didn't advertise or provide OpenGL support at all in that model.

[patio]

(Somehow, MSI went from low-end scrap dealers to making high-end Gaming stuff though...)

This caught my eye...i am really suprised at some of the glowing reviews of their high end products lately...
I guess things can and do change over time.

[DaveLembke]

I agree that MSI has shown improvement. The feedback ratings at sites that sell their products have actually averaged more than 4 of 5 stars for many products whereas in years past it would be 2.5 or 3 stars. I see their products when browsing through newegg and tigerdirect and other sites, and feedback seems more positive over the last about 2 years. It will still be a while before I would buy a MSI brand product though. There are others out there that are not so much a gamble still and around teh same price.

 The last MSI brand product I bought was a socket 478 motherboard for my last employer to repair a dead system for a user where the main board was toast, but CPU and RAM was still good so moved that into the replacement board and all was good for about 2 months and then the system would randomly reset itself as if someone pushed the reset button, but the reset wasnt wired up. I figured it must be the power supply so I swapped that and nope still did it. Checked temperatures to see if it was running too hot and resetting but temp of Pentium 4 2.8Ghz HT CPU was 49C which is normal. Thought maybe the soft power button was to blame on case since this is same case that the other motherboard died in, so I moved everything over to a new case and it still did it. Swapped out this MSI board with a used ASUS board from a friend and problem went away and the same CPU and RAM was still healthy. Tried to get a RMA replacement board and shipped out the board for replacement or repair and never got a replacement. Called support and got someone in India who wasnt any help but informed me that they could send me a replacement board if I would give them my credit card number to pay for replacement out of pocket. That was it with MSI in addition to friends who had all sorts of problems with their products lasting sometimes just long enough to die after warranty period was up. But to send a board back RMA and never see it again was just nuts. At least it wasnt my money lost it was a business that was out $80 which wont feel the pain of just a $80 loss when they did around 20 million in sales.

I think the biggest surprise reading the info at the website was the fact that it wasnt just 1 board that was made without cache or with fake cache and hacked BIOS to report that 256k cache was present when it wasnt, and this illegal activity didnt catch up with them. They were able to do this for quite a while with multiple boards and someone made a killing ripping off people by selling a lower cost product and so many people shop by price vs product name. If something like this were to happen today I would think that with all the people who benchmark boards it would be found out pretty quickly that a board is lacking functionality.

I think the only thing that can compare to this to me in last few years was buying a motherboard and installing my AMD FX-8350 to it and then confused by that the clock would never reach 4.2Ghz when running a heavy benchmark to pegging all 8 cores to 100% the CPU has 4.0Ghz as normal clock and wouldnt reach the 4.2Ghz Turbo clock mode that this CPU supported. I then found out that the motherboard that I bought did not support this 4.2Ghz Turbo clock mode. But the board manufacturer never stated that it would run it at 4.2Ghz or not, it just has a CPU support list and the AMD FX-8350 4.0Ghz was the heaviest CPU that the Biostar motherboard supported and the 4.0Ghz as listed was it not 4.2Ghz with Turbo. The ok thing is that with 8 cores at 4.0Ghz, the lack of the extra 200 Mhz doesnt really make a big difference and so I am not going to toss out this motherboard to spend $65-$75 to get a board that supports Turbo 4.2Ghz for the extra 200Mhz CPU speed. I could if I wanted to just run an overclock manually and have it run faster that way such as increasing the FSB clock. But I'm ok with it at 4.0 Ghz its still plenty of processing power.

Other board that kind of surprised me was the Gigabyte board with a Kabini 25Watt TDP APU build that I did and the memory controller doesnt support dual-channel as I wasnt aware until after buying 2 x 4GB DDR3 1333Mhz sticks thinking I was going to have 8GB of Dual-Channel memory, that its Single-Channel only with the socket AM1 builds so I could have gotten a single 8GB stick and saved a few dollars.  The memory controller is on the APU itself and it supports single-channel only. Corners are also cut in the design to how the BIOS is on the AM1 boards but for a $40 motherboard and $34 APU I suppose it should have been expected that its likely lacking something. 

But nothing illegal in these 2 somewhat recent experiences of missing or unsupported features. 

[patio]

I'm glad they have seemed to turn things around...however i'll never buy 1 of their products again.