View Full Version : The state of Motherboards today.

7th February 2000, 12:52
Is it just me, or is there a funk in motherboard land?

Just a few months ago, things were crystal clear in mobo world. If you went intel, you bought a BX board. There were a ton of them, and a number of mans. like asus, soyo, gigabyte, were making extremely stable and feature complete motherboards. There were others, like Abit, who were also pushing outside the lines and designing unique things into their boards (softmenu, dual ppga). The BX proved to be a very solid performer, technically and comercially.

Now if you wanted a SS7 board, things were a little trickier, but none-the-less worked out. VIA was most prevelant, but did have a few quirks (in my experience). Mostly with compatibility (i.e. my g200 on FIC 2013). But a few boards emerged that were stable (ASUS P5-A, for instance), and were also feature complete.

But then things changed. Technologies were announced that were going to change computing. AGP 4X, ATA/66, PC133, RAMBUS, Athlon, coppermine.

But, behold, there were no chipsets. i820 was a disaster, and, IMHO won't make it. I for one will avoid it like the plague, if on prinicple alone. i840 sounds like a neat archetecture, but intel won't push it anywhere but into the high end, which eliminates home users like me. i810 is a micky-mouse chipset, made for people like my mom (will run the same computer for 5+ years, and wonder if it is safe to take the cover off the computer). So intel's offerings have botched the viability of rambus, and most people are running their shiny new coppermines on their old trusty BX boards.

Now we are back to VIA. She of the 59 and counting AGP Gart drivers. They've had a chipset supporting PC133, AGP4X, ATA/66 around for a spell now, and some mobo mans have incorporated them.

Now, they are going to come to the rescue of the athlon with another 133 spec chipset. Like intel's new chips, these have been delayed for quite awhile. Since the Athlon roll out, as a matter of fact.

My fear is this: in a market dominated by the Wintel two headed monster (or is it the beast with two backs? http://forums.murc.ws/ubb/wink.gif ) can we rely on a via driven product?

The worst part is, I'm not sure where the problem lies. For instance, when win98 came out, it claimed full support for SS7 chipsets like the MVP3. But, I still had to install all these patches so the OS would recognize everything and operate properly. Why? MS always ballyhoos about their complete hardware support. SO why is it that (pre-Athlon) the second most important chip-line in the world (K6-2,3) wasn't getting full platform support in their best selling OS? Already on these forums I'm seeing: help! my g400 wont work on my Athlon(or)133PRO system. Solution:install latest Gart driver. Damn, here we go again.

So, is it MS/Intel's fault? I don't know. Maybe VIA's engineering sucks, thought I doubt it sucks any more than intel's. Even so, there will be no SMP for athlon as yet, which is a shame.

Maybe in time, there will be a clearer picture. In my opinion Intel had better push for a renamed chipset. One that supports ram that people other that bill gates can afford. AMD had better keep a close eye on the chipset development for their athlon platform. If it takes this long for new chipsets which don't have a complete set of features, they are in trouble. Also, they had better convince the mobo people that athlon will make them money.

man, now I'm tired...

7th February 2000, 19:59
I aggree but you are quick to dismiss the horde of chipsets that Intel pushed in late 97-late 99. The FX VX LX and finally the BX were all on the market at one point and the user had to wonder what to buy. Waiting about 4-5 months gave us the BX chipset as a clear winner (with VIA and ALi trailing closely). The dust settled for about a year and now the same situation is happening all over again, compounded by the intoduction of a new CPU and chpset. I would just sit my ass down on the trusty BX till the end of the summer, waiting till the dust settles

7th February 2000, 20:02
I think we'll see a change as the newer boards based on the VIA KX133 chipset come into play. I have read that AMD has been working very close with VIA on this project. Of course new technology tends to cause everyone problems. Just look at the problems that Intel is having with Coppermine and the i820 chipset. Personally I have three AMD CPU based systems, two with the MVP3 chipset, and they all run pretty stable. I mean call it luck, call fate, call it general know how, call it whatever you want to but all I know is that I have not had most of the problems with my G400 as other SS7 owners. And I probally have one of the earliest G400 boards produced, overclocked to MAX speeds at that from almost day one.

Joel http://forums.murc.ws/ubb/smile.gif

7th February 2000, 21:22
This is a bad time to upgrade things like motherboards and cpus IMO. Things will get interesting with socket A and the Athlons with on die L2. From benchmarks I've seen the KX133 or whatever it's called is still slower than AMD's irongate with super bypass enabled, but it does offer "features" like AGP 4X and support for 133MHz ram. All the Intel offerings lately have been rather underwhelming, then again, their MO is add 5% improvment, make it incompatible and try to convince you that you need it. http://forums.murc.ws/ubb/smile.gif

8th February 2000, 00:15
Himself: Have you taken a look at tomshardware updated benchmarks? It cleary shows the KX133 a winner.

edit. damn english language.

[This message has been edited by mj12 (edited 08 February 2000).]

8th February 2000, 05:23
No chipset that requires extensive patching to work with prevalent operating systems is a "winner", by any definition of the word.

- Gurm

Listen up, you primitive screwheads! See this? This is my BOOMSTICK! Etc. etc.

8th February 2000, 06:22

Have a look at Anand's KX133 review, it shows that AMD750 is faster than the KX133. http://forums.murc.ws/ubb/smile.gif I don't trust Pabst, he gets into "have to prove a point" mode too easily. Only real performance advantage KX133 has is being able to run faster ram, 133Mhz vs 100MHz, that's it. The other advantage being able to use a 4 layer board so it's cheap. Anand's numbers show that super bypass give the irongate an edge even with the KX133 having virtual channel sdram at 133MHz. That says pretty clearly to me that the irongate is the faster northbridge. Pabst, or whoever wrote the bit, on the other hand would rather ignore the existance of super bypass, harp on the glory of AGP 4X, and go out of his way to find benchmarks showing the effect of faster ram and AGP 4X, which is not testing chipsets, but ram bandwidth and video cards. If someone buys one of these cheap KX133 boards and uses 100MHz ram with it they are not going to get better performance than they would with the AMD 750.
The next VIA chipset with DDR support is the one to wait for, 133MHz is not worth your time, especially if you expect much from AGP 4X.

My $02

8th February 2000, 06:40
Himself: you may already have read this, but i'll post it here: http://www7.tomshardware.com/mainboard/00q1/000205/index.html
the implication, from the numbers, not the author, is that the drivers previously were dogging the hardware.

Let's not forget, AMD doesn't want to, or can not provide chipsets to the platform. The VIA offerings will be the only game in town, barring SIS or someone else getting in on the act.

8th February 2000, 06:47
BTW, in computerland there is plenty of blame to go around. Most times it is difficult to know where it should stick.

For instance, microsoft gets blamed for just about any software related issue known to man. And in a lot of instances, I would agree. But, don't forget that every piece of hardware and every piece of software you install puts files all over creation. there is stuff going in you windows, windows/system, windows/system32 directories. There is stuff going into your registry. Hardware is accessing chunks of (in win9x, unprotected) memory. dll's get overwriten with older or incompatible versions, in spite of the microsoft driver model.

And what does microsoft have to counteract this, besides underhanded marketing practices? They can tighten up the windows certification process, as they did with Win2K. But then software designers say, hey that is to tough, so we are just going to release our software anyway, with or without your certification.

Just playing devil's advocate, BTW. http://forums.murc.ws/ubb/wink.gif

8th February 2000, 08:17
While I agree that this is a bad time to upgrade motherboards, chip shortage aside, I think it's about to become a pretty good time for overclockers to upgrade processors.

The "low end" of the Coppermine E-series appears to be very overclocker-friendly. If Intel can get their supply act together, I think we'll be seeing a lot of activity with the 500E, 550E, and 600E in March and April.

I think the chipset situation couldn't be more dismal. Much of this, I suspect, is due to a considerable amount of heavy handedness on the part of Intel. They recently announced they intend to quicken the transition from Slot 1 to FC-PGA in the absence of a decent motherboard that supports the new socket Coppermines.

Intel knows OEM's actively dislike the i810 and i820 solutions, so they are now pressing OEM's to use slockets. I find this particularly ironic. One of the benefits of the socket format is that it is cheaper to produce, and the savings were to be passed on to OEM's and the enduser. So much for that.

The Solano II chipset is suppose to correct a number of the problems inherent to i810 and i820. I've read that it will be launched in April.


8th February 2000, 11:37

I agree that VIA will be the only game in town, but for the simple fact that AMD doesn't want to do chipsets and 4 layer boards are cheaper than 6 layer boards. That's why I mentioned the DDR version of the KX133 they say they will eventually get out. And of course AMD will probably have to do it's own SMP chip rather than waiting for VIA to come up with something. VIA tends to go for the bottom feeder market, so SMP is kinda out there for them.

As for MS, I tend to blame them for generally bad design under the hood rather than the execution of the API, but there are some large goofs there as well. (Today's example being the registry entry for opening folders that doesn't get used when running explorer but does when you are running an alternate shell.) Windows has a lot of nifty gui stuff, other than that, it's still basically a dos shell. There should be a warning label on Windows, "Not for use with non MS applications." http://forums.murc.ws/ubb/smile.gif

8th February 2000, 12:28
Doesn't the VIA Pro133A already have SMP for coppermine chips. I think ASUS already has a dual CuMine board...

8th February 2000, 16:00
I've read somewhere that AMD is finding the idea of producing its own chipsets less objectionable. If true, maybe it has something to do with the fact that VIA intends to produce lowend processors.


21st April 2000, 02:18
oh i hope so any one remember the last chip set amd produced ? it was rock stable, just like the irongate, heres to hopeing amd makes more chipsets.

21st April 2000, 06:03

And do you remember the AMD640? Hmm?

- Gurm

Listen up, you primitive screwheads! See this? This is my BOOMSTICK! Etc. etc.

21st April 2000, 08:19
mumble mumble , you would have to bring that up http://forums.murc.ws/ubb/smile.gif

[This message has been edited by merchant2112 (edited 21 April 2000).]

21st April 2000, 09:01
The first post in this thread was Feb 7. Its a sad state of affairs that in the following two and a half months nothing has changed. Neither Intel, AMD or VIA has come out with a new and improved version of their chipsets. Maybe next month - yeah right!


Oops, sorry, I turned cynical again.

21st April 2000, 20:37
i feel you guys' pain.. i am trying to buy a new motherboard too and i can't decide on which chipset to get.. sigh..