VIA chipsets

OK last one..........
What's the good (or bad) word on VIA Technologies. Anyone keeping up with the fast paced subject of chipsets, processors, and MB mixing and matching? I've reached pages so old out on the Web that told me VIA and Intel were still stubbornly resisting each other. Any inside stuff (or anything) on how well the bond was formed. What's with the "Bridge" naming system?" North & South Bridge chips, man whotdadeal?

 

I've used VIA boards since '98 and have not had any problems with them. The "party line" is that VIA's implementations of USB, as well as the AGP and PCI buses are not as good as say those of Intel. I believe this to be true. VIA had to reverse-engineer AGP in order to even implement it. It's a fact that their PCI bus implementation has "latency-related" problems that reduce total throughput and can lead to "stuttering behavior" under high-load conditions and some compatibility problems (soundblaster).

But have I suffered from using VIA all these years? - absolutely not! But, I will admit, being a hardcore gamer, alot of bells went off for me when I started hearing all the problems people were having with the VIA-WinXP-GeForce 4 combination. And after my excellent experience with the SIS 735 chipset on an ECS K7S5A, I decided that SIS was an excellent alternative to VIA and might help me avoid any future headaches. I just bought an MSI 745Ultra based on the SIS 745 chipset. I'm very impressed with SIS' stability and compatibility.

VIA chipsets are implemented with two chips: one called a northbridge (near the "north" part of the board) and one called a southbridge (near the "south" side of the board). They split up the duties of "running the motherboard ". The SIS chipsets (735 and 745) are implemented in a single chip thereby reducing the cost of the board.

 

Well, I have a VIA kt133a chipset on my Gigabyte motherboard. I've had no issues whatsoever since I purchased it about a year ago. Also have two friends with the same motherboard and no complaints from them either. Moreso, a buddy of mine has the kt133 chipset from VIA and has never had a single complaint. Makes me kinda wonder why I heard so many bad things in the past.....hmmm. Standard rumour kinda thing???
Not sure about this regarding the Northbridge and Southbridge naming system but, my guess is that the Northbridge IS the upper (higher on the board) and the southbridge IS the lower (lower on the board) and named after the standard compass settings of a map? North is up, south is down.

Dale

 

Count on this board to come through with the goods. Thanks fellow enthusiasts. Your words are reassuring. So this ECS_P4VMM2 MB won't be to much trouble. I haven't had to much to do with VIA chipsets or even AGP, and you have. I've just upgraded from P1 166MMX machine to a P4 1.8G system and I'm getting the lowdown on all the stuff in it.
Just what I was looking for. Thanks again

sidermike w/meanP4pc

 

Dale, as I said, I've personally never had any problems with VIA boards but others have - probably most problems related to the PCI latency issue.

Sidermike - I didn't realize you had gone over to the dark side! j/k

I really know nothing about VIA chipsets for Intel processors - my comments were related to AMD platforms. I suppose some of the same drawbacks could be true for VIA's Intel boards as well.

I've heard that the best non-Intel chipset for Intel boards is currently produced by SIS. Don't know the chipset name but I hear it rocks!

 

>>>>Dale, as I said, I've personally never had any problems with VIA boards but others have - probably most problems related to the PCI latency issue>>>>

Very curious as to this latency issue. Is this because Sisoft Sandra says so or, is this an actual fact that affects the perfomance of said machine? Sandra says so, on my machine,
but I do not find or feel any issues. Hmmm....again.

Dale

Gigabyte 7ZXR-C, XP 1600.....ect....ect......

 

Dale, I've got the link at work on a article all about "the VIA PCI Bus problem ". Tomorrow, I'll edit this post with the link. Basically it talks about how VIA's PCI implementation can cause problems because it interrupts data transfers on the PCI bus much more frequently (than say Intel, SIS, or ALI). But I'm paraphrasing - the quite extensive article will be much clearer.

One consequence of this is that VIA's performance with PCI Raid cards (something HEAVILY dependent on PCI throughput) is significantly less than other chipset makers.

Here's the link to the rather lengthy article:

http://www.tecchannel.de/hardware/817/index.html