Originally posted by: Zap
Originally posted by: myocardia
Am I the only one who remebers how horrible past Via chipsets were?
You aren't, but really, you shouldn't live in the past.
#1 reason VIA chipsets were terrible in the past... drivers. That was fixed with the "4 in 1" (later Hyperion) drivers many, many years ago.
#2 reason VIA chipsets were terrible in the past... used on cheap and crappy motherboards. I've seen the same VIA chipset on $35 motherboards as on $70 motherboards, and can tell you that even though the chipsets were exactly the same and used the same drivers, the cheaper boards sometimes... performed crappy, had stability issues or just plain had a high failure rate. It is an unfortunate byproduct of VIA having to price their chipsets lower than Intel chipsets.
#3 reason VIA chipsets were terrible in the past... not very overclockable. Reason was that it took them a while to implement locked busses. The first one I can remember was the PT880 chipset for Pentium 4, and that one even had problems with the SATA ports failing even with a 1-2MHz FSB overclock (but AGP/PCI locked fine).
So, #1 was just a driver issue that was resolved way back in the heyday of Super 7 boards and AMD K6-2 chips. #2 is a problem of cheap/crappy motherboards, not chipset. #3 is a problem for overclocker/enthusiasts, and isn't even a problem, just a missing feature (good overclockability).
Not to say that there weren't some really crappy VIA chipsets. Their first Super 7 chipset was the MVP3 (maybe?) and I recall a lot of problems... but again, that was before the 4in1 drivers. Their VIA Apollo 693(A) chipset had definate performance issues. However, their 694 chipset was an excellent alternative to the Intel BX and 815 series chipsets. The first SDRAM chipsets for Athlon chips were nice, the KT133(A). However the first DDR chipset, KT266, didn't fare as well. That was partly due to motherboard manufacturers cutting corners and slapping the chipset on existing KT133A PCB designs, IIRC. However, VIA redeemed themselves with the KT266A chipset, and later the KT333 and KT400(A). The problem was that not only was their Intel variants not as exciting (PX266/333/400), but around that time AMD was playing around with locking the multiplier. At first, people found ways around it (pencil trick anyone?). However, later Barton cores were hard-locked, so unless you were able to jump a full FSB ahead, such as from 266-333, or 333-400, you weren't going to get much of a stable overclock. This is where Nvidia started to become known for overclocking, with their Nforce2 series of chipsets and locked AGP/PCI busses. Was the Nvidia chipset "better" in that it performed better or was more stable? Probably not. However, it grabbed the enthusiast/overclocker hearts like the VIA chipsets never did.
Everyone hating VIA chipsets and loving Intel chipsets probably fail to remember the stinkers that Intel put out. How about the 810 series chipset that didn't even have an AGP slot capability? How about the 815 series chipset that was touted as the BX replacement, but which had an artificial 512MB RAM limitation (so it wouldn't compete with higher end chipset)? How about the 820 chipset which used RAMBUS and was even recalled once? How about the original 845 chipset for Pentium 4 that crippled the P4 with only SDRAM support (so it wouldn't compete with higher end chipset)? How about the 915/925 series chipsets which nobody seems to reminisce fondly over (
Intel's overclock lock anyone)?
Facebook
Twitter