"Flaws in Via chipsets hit ATA/133, SCSI performance"

[Duvie]

Yeah round up your buddies azgamer and jerboy and have them tell us where the chipset issues are...They talk about sis problems but everything they mention is related to motherboard specific issues...ie doa board, cmos batteries dieing, picky power supply and ddr requirements, and incorrect ohm resistors causing data corruption with 1.4ghz tbirds...

C'mon what is the sis chipset issues....Even if you want to state memory bandwidth and ide performance is better in kt266a, is it really that better in real-world performance and good enough to be worth 50 percent more cost???

 

[Pabster]

Candoman wrote:

"SiS and ALi chipsets, on the other hand, have to use the standard Windows drivers."

Which, not coincidentally, both perform better than 686B and 8233.

 

[ToBeMe]

<<

<< Heh. "Performance Type Drivers"? You mean 4-in-1 "patch" packs? SiS 735 has the speediest IDE performance out there today. >>



No, I don't mean the 4-in-1 drivers, sorry. You can find performance type driver for all the VIA southbridges at the bottom of the page here. It reads as 'IDE performance drivers. 580_3012.zip ', and it does improve IDE transfers over the standard 4-in-1 IDE driver. SiS and ALi chipsets, on the other hand, have to use the standard Windows drivers. In fact, there're some ALi IDE drivers, but the company recommends to use Windows default ones for better performance . >>


Gotta' ask........have you ever even owned an SiS 735 board and got it up and running?????? Believe me, i've got Via and SiS as we speak WITH the Via "Patch" you mentioned and there is no comparison..........the SiS is noticably faster.........


Also.......just an FYI.........those "Performance" drivers are NOT for XP before anyone goes and tries to install them in XP..........

 

[rahvin]

<<Intel chipsets are, by and large, rock solid. We can agree on that. But I'm a little tired of hearing people offer up the excuse that SiS 735 has "issues". Where are they? It's rock solid, and high performing. It's also one of the cheapest! >>

Take those X15's you own, put them on one of those K7S5A's and get it to boot. If you can get a bootable setup without setting the drives to narrow scsi I will be impressed. I spent 3 days trying to get the two to work (I formated so many times it was sick and tried 3 different OS's.) It could be a BIOS issue but I know there are other problems with that board and they can be found at the OCworkbench site devoted to it. It does have issues, they are bizzare and unseen most of the time. If the issues are actually with the chipset I would wager that it is a result of the incomplete product testing (whereas VIA just appears not to care).

SIS isn't perfect and neither are their products.

 

[Doomguy]

The ECS K7S5A is plagued with problems. Just check out the OcWorkBench forum. It's not even close to perfection.

 

[Budman]

<< Take those X15's you own, put them on one of those K7S5A's and get it to boot. >>



This is a Motherboard issue,NOT A SIS735 issue. all your lame excuses are tiresome, Show me REAL documented SIS735 issues.

 

[Budman]

<< The ECS K7S5A is plagued with problems. Just check out the OcWorkBench forum. It's not even close to perfection. >>



When you get a cheap board, cheapscates will buy it & skimp on other components, PSU's Video, Ram,etc..... that's why some have problems with the ECS k7S5a, use quality stuff with it & 98% of the problems just disapear.

 

[Pabster]

Rahvin wrote:

"Take those X15's you own, put them on one of those K7S5A's and get it to boot. If you can get a bootable setup without setting the drives to narrow scsi I will be impressed. I spent 3 days trying to get the two to work (I formated so many times it was sick and tried 3 different OS's.) It could be a BIOS issue but I know there are other problems with that board and they can be found at the OCworkbench site devoted to it. It does have issues, they are bizzare and unseen most of the time. If the issues are actually with the chipset I would wager that it is a result of the incomplete product testing (whereas VIA just appears not to care)."

Prepare to be impressed

I'm using a 19160 with a 36.7GB X15-36LP on a K7S5A right now. And it is the boot drive. No issues whatsoever. And, suffice to say, the drive performs considerably better with SiS 735 than it did with any VIA chipset.

"SIS isn't perfect and neither are their products."

And I don't think anyone ever insinuated that. I simply asked for a list of the "bugs" and "issues" with SiS 735. So far, I've received absolutely nothing. For the last time I ask: where are the bugs/issues with SiS 735?

 

[mithrandir2001]

VIA fanboys! Ha ha, that sounds funny.

 

[Sunner]

Damn these threads give me self confidence, since I have a bunch of rigs, running VIA and intel chipsets, and I dont have any problems with any of them.

Guess Im just so damn good

 

[Flatline]

Yeah, I've got SiS, VIA (the venerable ASUS P3V4X), and Intel chipsets...if you set them up properly and don't OC the hell out of them, they seem to work pretty well

 

[Damascus]

<< there's no Creative glitch >>



The 'glitch' is the Creative product itself...



<< no nVidia glitch >>



Just don't put 3 DIMM's on your nForce board!

 

[Texmaster]

<< Coming from The Register, eh. I guess a lot of people are ***confirming*** this one, since everyone is taking it as fact. I'm not effected, and my other home machine isn't effected. The fact of the matter is, even if there is a flaw, is this really going to effect enough people to even mention, or is it just one of those usual very concentrated flaws with very specific components. Lol, I remember reading an article that said that if you had ANY VIA chipset, a GeForce 3 series card, and Win95/98/ME, that you would suffer from a 30% performance decrease. Lol, that obviously was true [hint, sarcasm].

Move along, nothing to see here... >>




I knew if anyone would downplay this most recent FLAW in Via chipsets it would be AGodspeed! LOL

Just another predictible blow to your warped and inaccurate theory about VIA chipsets! HAHAHAHA

 

[NicColt]

>"SiS and ALi chipsets, on the other hand, have to use the standard Windows drivers."
>Which, not coincidentally, both perform better than 686B and 8233.

I'd like to see proff of that if you don't mind.

 

[Zebo]

<< Damn these threads give me self confidence, since I have a bunch of rigs, running VIA and intel chipsets, and I dont have any problems with any of them.

Guess Im just so damn good >>



Add me to that list. Or maybe we're just to stupid to realise we have problems and can learn something from Pabster

 

[AGodspeed]

I knew if anyone would downplay this most recent FLAW in Via chipsets it would be AGodspeed! LOL

And only a good old flame shooter like Texmaster would make a statement like that without backing it up with facts. Just like how you blocked my pm's, you just can't seem to accept defeat. Were you getting tired of losing your arguments in pm lol.

 

[WarCon]

Are there any issues with that IDE performance driver? Will it improve performance for ATA100 or just allow ATA133 to function correctly? I am lost. My ATA100 benches as an ATA100 drive every time I test it and I have a VIA chipset. I personally don't know enough of the actual issues to contribute much to this little toasty place, but am curious as I am looking to upgrade in the next several months and am looking to have the same or better level of stability as I have now. I have an ABIT KT7A, what is the best (not best for the price, I have gone that route too many times and been unhappy) mobo currently.

I have seen lots of people praising the Dragon + and the Epox DDR boards, but from some of you diamond and higher members who currently run these boards, honestly what is really good overclocker and stable as lets say a BX board?

 

[gogeeta13]

<< >"SiS and ALi chipsets, on the other hand, have to use the standard Windows drivers."
>Which, not coincidentally, both perform better than 686B and 8233.

I'd like to see proff of that if you don't mind. >>



SiS735 primary IDE channel. Western Digital 30ddb(20GB platter density, 2 platters, both sides of one, 1 side of the other): 82. 8 mbps from HD Tach 2.61

Via kt133 with 686b primary HDD(wester digital, same drive as above) 40.9 mbps from hd tach

Via kt266a with 82xxx whatever the hell it is Maxtor +60(20 gig platter, same seek times and all as the western digital) 59.9mbps

These are all real scores, and were benched by me in the last few minutes.

A quick note, i have a promise ata controller in my kt266a system. it also has a maxtor +60 on it. It scored 79.2mbps

EDIT: ANother note, this is with default drivers from WinXP pro with ALL the boards, I never installed any new 4 in 1s in the via system, and I have never installed a SiS IDE update of any sort in the k7s5a(sis735)

 

[Pabster]

GASP!

Those numbers can't be right. I mean, they show dismal IDE / PCI performance from various VIA south bridges.

That's the beauty of SiS 735. The IDE performance is top notch -- right out of the box. You don't need to hunt down a new driver or go looking for some third-party "patch" to improve performance. In fact, it appears VIA south bridges (8233 included) are still sub-par, even with Breeses' Latency Patch. Better, yes, but not up to SiS 735.

 

[WarCon]

gogeeta, are those measurements the normal or burst mode readings?

 

[MadRat]

I hate to interject here, especially since this is a pissing match, but the VIA chipsets have been designed with an eye towards progression. The VIA chipsets are limited by the inherent limitations of "budget" designs, not because they cannot design performance. They build their performance with each design and regularly introduce new technology with every revision. Its a farce to argue that one design is better than the other when the designs were built for specific niche pricemarkets, not for maximum performance.

 

[gogeeta13]

<< gogeeta, are those measurements the normal or burst mode readings? >>



Burst readings from HD Tach. The seek times were pretty much the same across the board. I didn't let the test run long enough to get the average sustained transfer, but if you guys really want, I will.

 

[WarCon]

I just ran my KT7A (via chipset) and it tached at 79.8 before I installed the bus mastering drivers (which when I figure out the best way to remove them, I will do so) and now with the bus mastering drivers it only bursts at 68.3.

Wierdness.

Oh it was a Maxtor 20gig ATA100 7200rpm.

 

[Pabster]

MadRat wrote:

"I hate to interject here, especially since this is a pissing match, but the VIA chipsets have been designed with an eye towards progression."

LOL That's certainly a new one

Progression to what? The 5.xx series of 4-in-1 "patches"?

"The VIA chipsets are limited by the inherent limitations of "budget" designs, not because they cannot design performance."

Oh, baloney. VIA-chipset based mainboards command the highest prices in the Socket A arena. There's no "budget" issue there.

"They build their performance with each design and regularly introduce new technology with every revision. Its a farce to argue that one design is better than the other when the designs were built for specific niche pricemarkets, not for maximum performance."

Now you're argueing that VIA doesn't attempt "maximum performance" with each chipset revision? That just doesn't make sense. The fact that they rushed the "A" revision of KT266 out the door to quell the uprise from SiS 735 is a shining example. And, again, VIA-chipset based mainboards are commanding the highest prices in the Socket A arena.

 

[AGodspeed]

And, again, VIA-chipset based mainboards are commanding the highest prices in the Socket A arena.

Compared to what? Current ALi, Nvidia, and AMD chipsets (and their respective motherboards) are all more expensive than current VIA chipsets (and their respective motherboards). The only other chipset/motherboard that is cheaper is the SiS 735 (ECS K7S5A).

Granted, the current chipsets/motherboards Nvidia (nForce) and AMD (760) make are of higher quality and stability IMHO than current VIA chipsets (KT266A) (I just got the K7N420 nForce today! ), the fact of the matter is that current VIA chipset/motherboards are by no means "the most expensive Socket A boards."