SiS would appear to have been on a massive improvement drive over the last year, to the point where I'm actually willing to go with an SiS 735 based board in my main machine.
And it HAS been the least problematic board I've ever dealt with.
Nforce is seeming like a nice solid platform these days, I'm using an SiS 735 board here and it's NO trouble whatsoever... the AMD 761/762 (uni/smp) northbridges are stable.
If I hear ONE more person say "I'm not buying AMD 'cause VIA chipsets suck", I'm going on a killing spree.. damnit.
Since I just had an SH6 (i815E board) that was a little over a year old suddenly get REALLY flaky for no apparent reason, I'm not about to trust Abit in a hurry.
I will give them credit for one thing though, they know how to lay out a board well.
the SH6 is/was about as close to perfect in physical layout as a slot 1 board can get. if only they were using more highly trained monkeys for QA.
It's _impossible_ to get ahold of Abit UK, so they are right off my list.
Twitchy boards + lousy support (in this area of the world) = me not buying Abit
btw, the KT7 has Crucial PC133 and an Antec 350w psu feeding it, it does have an SB Live! on it but the twitchiness it shows is NOT the usual live/686 problem, It'll lock randomly unless ALL the ram timings are dropped to their slowest and all the performance enhancing features (4-way interleave, enhance chip performance etc) also need to be turned down.. bah
some Epox boards are nice, others less so (the 8K7A is quite a nice board)
It saddens me to see Abit get such high placement though.
at a guess, they just guesstimated a running order based on speed > tweaking features > stability/reliability.
which is highly screwed up, reliability should be top of the list, it doesn't matter how high it can overclock a processor if it's not 100% stable at any clockspeed:/
I second this, I've got a K7S5A with a 1700+ / 512MB of Crucial PC133 (I shall be selling off an excess gigabyte of PC133 I got when the prices were rock bottom and putting the proceeds towards some DDR) / AOpen 300w PSU (rebadged Sparkle I think?) and after I replaced the absolutely useless Coolermaster HSF (60c load temps with random hangs.. GG Coolermaster) It's been absolutely solid.
More solid than the Abit SH6/P3-933 rig it replaced certainly:)
By the way, they seem to have stopped putting QA markings on boards.. I know of more than 25 Abit boards that have _no_ mark in the QA box on the sticker on the last expansion slot.
15 of those boards are either problematic or outright dead >:(
The article seems to put featureset ahead of stability, oh, and VIA chipsets (KT266A/KT333 _possibly aside_) aren't 100% stable, they can be very twitchy (the Asus P3V4 I've got here isn't exactly a cheap board, the Abit KT7 in my dad's machine is the most twitchy and unstable board I've had the misfortune of using, and yet it comes highly recommended from most hardware sites because they are blinded by featureset and "Hey, it managed to run quake 3 demos for 4 hours without crashing!"
On the other hand, I know of around 80 ECS K7S5A's being used out in the field that are working without any problems whatsoever (including the machine I'm sitting at right now.)
Now, in general, I wouldn't trust ECS all that much.. but I have yet to get burned by them, Abit have burnt me several times:( gah.
Actually, no, I'm using one without problems, and I know of one site that has 40+ machines based around them, and they've had _one_ board DOA and _one_ die after a couple of months.
Compare that to the joys of Abit quality control....
Slashdot Mirror
SiS would appear to have been on a massive improvement drive over the last year, to the point where I'm actually willing to go with an SiS 735 based board in my main machine.
And it HAS been the least problematic board I've ever dealt with.
"Although I guess now the new Macs are all based on some RISC processor... "
:)
Um, 6 years ago (1996), the Mac's of the time would have been PowerPC based already.
Oh, yeah, Windows 2000/XP IS a huge leap over Win9x/DOS/Win3.11 etc.
Microsoft finally released something that was actually stable.... only took them.. oooh.. 20 years?
Try not to forget that Rambus Inc. are EVIL INCARNATE.
naa, 3.9
(that said, OS 4 is due RSN)
Nforce is seeming like a nice solid platform these days, I'm using an SiS 735 board here and it's NO trouble whatsoever... the AMD 761/762 (uni/smp) northbridges are stable.
If I hear ONE more person say "I'm not buying AMD 'cause VIA chipsets suck", I'm going on a killing spree.. damnit.
Quite often flight sims are _VERY_ cpu intensive, and a 300Mhz speed bump could have made quite a noticable difference.
*cough* Christmas gets on my nerves.
I've got a 4 year old BH6 that's still going strong (aside from one of the fan headers dying)
I wouldn't touch any of Abit's recent boards with a VERY large bargepole however.
FYI, my board is only a month old or so, it came with the 01/11/09 bios installed.
"Don't believe me? Throw in a 1.4 Ghz processor in their K7S5A and tell me its quality."
I have a 1.47Ghz processor on my K7S5A...
Actually, I think you'll find that it's your personal opinions ON how things are.
I haven't had _one_ crash from this ECS board.
I had plenty from my Abit's, I've had quite a few from the Asus board I've had sitting around.
Whilst I've had, ZERO, yes, ZERO hardware related crashes (I had a heat problem, but that was my crappy Coolermaster HSF) with this SiS 735 board.
I probably wouldn't use one for something mission critical (that's what Supermicro/Intel/Tyan boards are for) but for a workstation?, sure.
(btw, until I rebooted it to install an IE patch this morning, this machine had been up for 20 days)
Since I just had an SH6 (i815E board) that was a little over a year old suddenly get REALLY flaky for no apparent reason, I'm not about to trust Abit in a hurry.
I will give them credit for one thing though, they know how to lay out a board well.
the SH6 is/was about as close to perfect in physical layout as a slot 1 board can get. if only they were using more highly trained monkeys for QA.
It's _impossible_ to get ahold of Abit UK, so they are right off my list.
Twitchy boards + lousy support (in this area of the world) = me not buying Abit
btw, the KT7 has Crucial PC133 and an Antec 350w psu feeding it, it does have an SB Live! on it but the twitchiness it shows is NOT the usual live/686 problem, It'll lock randomly unless ALL the ram timings are dropped to their slowest and all the performance enhancing features (4-way interleave, enhance chip performance etc) also need to be turned down.. bah
"I know this article dismissed that idea, says it's only a problem with cheap motherboards."
:)
they are smoking crack
they still haven't managed to get a PCI bus that works properly on the KT266A >:(
I also find it strange that a lot of people have VIA/Abit problems, yet they've put Abit at the top of the list?!?!?!?!??!?! EXCUSE ME?
"Buy Abit, Abit are good!
BUT IT KEEPS CRASHING!
Then don't buy a crappy mainboard!"
*sigh*
some Epox boards are nice, others less so (the 8K7A is quite a nice board)
:/
It saddens me to see Abit get such high placement though.
at a guess, they just guesstimated a running order based on speed > tweaking features > stability/reliability.
which is highly screwed up, reliability should be top of the list, it doesn't matter how high it can overclock a processor if it's not 100% stable at any clockspeed
VIA have only been selling boards under their own name (rather than producing chipsets for other board integrators to utilise) for a year or so.
"requires weird registered memory" I think you'll find that's pretty much the case with all AMD 762 Northbridge based boards.
:)
a problem like you were having sounds like a bad power supply to be honest
Oh, btw, www.shacknews.com recently (Feb 20th) switched to AMD on Tyan servers, they don't appear to have crashed and burned yet.
I second this, I've got a K7S5A with a 1700+ / 512MB of Crucial PC133 (I shall be selling off an excess gigabyte of PC133 I got when the prices were rock bottom and putting the proceeds towards some DDR) / AOpen 300w PSU (rebadged Sparkle I think?) and after I replaced the absolutely useless Coolermaster HSF (60c load temps with random hangs.. GG Coolermaster) It's been absolutely solid.
:)
More solid than the Abit SH6/P3-933 rig it replaced certainly
Abit quality control is very poor.
:( gah.
By the way, they seem to have stopped putting QA markings on boards.. I know of more than 25 Abit boards that have _no_ mark in the QA box on the sticker on the last expansion slot.
15 of those boards are either problematic or outright dead >:(
The article seems to put featureset ahead of stability, oh, and VIA chipsets (KT266A/KT333 _possibly aside_) aren't 100% stable, they can be very twitchy (the Asus P3V4 I've got here isn't exactly a cheap board, the Abit KT7 in my dad's machine is the most twitchy and unstable board I've had the misfortune of using, and yet it comes highly recommended from most hardware sites because they are blinded by featureset and "Hey, it managed to run quake 3 demos for 4 hours without crashing!"
On the other hand, I know of around 80 ECS K7S5A's being used out in the field that are working without any problems whatsoever (including the machine I'm sitting at right now.)
Now, in general, I wouldn't trust ECS all that much.. but I have yet to get burned by them, Abit have burnt me several times
Considering that they had 12 out of 20 Abit's die on them, yes.
"Actually ESC K7S5A is a pretty lousy board."
Actually, no, I'm using one without problems, and I know of one site that has 40+ machines based around them, and they've had _one_ board DOA and _one_ die after a couple of months.
Compare that to the joys of Abit quality control....
Who the hell are Kbit?, you mean Abit?
?!!?
You'd never heard of Tyan?!!??!
(that truly astounds me)
As I just posted elsewhere
h tml">Tyan Tiger K7</a>
<a href="http://www.tyan.com/products/html/tigermpx.
I don't know why nobody goes on about stability problem with Intel boards, because I've seen enough of them with my own eyes.
There's a few websites being hosted of AMD hardware with no problems (yet) too, <a href="http://www.shacknews.com/docs/amd.x">like Shacknews</a>