VIA Releases Free HyperTransport™ Analyzer Utility for AMD64 Based Motherboards

[Belwarrior]

People buy the products, get pissed off, forget, and keep buying the products.

unfortunately, i dont think this is going to change anytime soon

My feelngs are that it is a pretty baisc benchmark ment to show the faults of ONE of the competng chipsets, (ALi seems to be the guess at the moment) then spin this through advertising to give the reader the impression hat VIA is the only one who dosen't suffer from the problem.

This is why i feel what VIA's utility is irrelavant, and agree with Admiral that an AMD made utility would be the better choice.
I also consider VIA as being unethical IF indeed this utility will be used in the manner funky thinks as this may cause the some other manufacturers to come up with campaigns deriding their opponents' (through such utilities) rather than publishing their own products' merits. VIA's well-known pci bandwidth issues could be targeted easily enough so i am a bit surprised why VIA would bother to expose another chipsets' faults.

@funky - Every new or even the top performing chipsets have errors and issues including current amd holygrail Nforce2 , it is just that nvidia admitted its problems and released patches to resolve them. VIA will getting the stick until they change. Not to say that they make very bad chipsets, just that they are not addressing the issues feedbacked to them.

[Nowhere]

Originally posted by funky-d-munky
This is becoming a bunch a biased crap. With nothing being said based on fact.
What if VIA gets it right with this design, and NForce3 is the KT133 of the new era?
...

You want unbiased (from your point of view ) facts? How about something directly from oh-holy-via? Check what is used as southbridge. It's not connected using hypertransport at all. And check that this is THE SAME chip that is used in KT600 and KT400A, which is just a slight evolution of the ones from KT400, 333, 266A & 266. Oh, and they're all crap when it comes to pci implementation. This thread isn't a place for question "will via in this new generation finally repair their pci?", because it's NOT new generation (regarding the part of the chipset that is causing majority of problems...).
BTW, Nforce3 is for some time on the market (it's not chipset for Clawhammer only...) and I didn't hear that it's crap.

Originally posted by funky-d-munky
Doc, I expect so much more from you. Were talking about a brand new architecture here, and you bring up crap about KT266a? Not biased my blank.

Wombat again all your fact points have NOTHING to do with the thread. Again how many reponces to this thread yet no one has brought any fact of the new K8 sytems to the table. What point are you trying to make.? Ya you hate VIA, so what, big deal. Get over it. The topic of this thread is not "VIA is the best, please debate here"

GT98 "Well 99% past performance is normally indicative of future performance"

Well SiS used to make some crappy MB's. Intel made 256kb L2 400bus P4's, Matrox made the G400 ect.. ect..
When companies design a completely new architecture it is guesswork who will be the best.

As I pointed above, Doc had every right to bring KT266A in this discussion. I'll say it again, this is NOT completely new architecture. And Wombat showed you what this architecture was capable of in previous incarnations

And you're wrong about SiS. They've made a good P3 chipsets, the same is true about P4. BTW, I don't think P4's with 256kb L2 were a bad product...there's a difference between product that's helded back in performance because of design decision and the one that doesn't work properly. And I have no idea what were you thinking when you gave as an example G400 - for several months those were the best cards on the market

Originally posted by funky-d-munky
"But with Via you can see they didnt do much with fixing the PCI problem from the KT133 to the KT266"

Thats the same architecture, just updated. K8 is a completly new chip.

You know what I'd say here already, don't you?

Originally posted by funky-d-munky
"Well since Via tend to add to old designs rather than starting anew yes you can expect old design flaws and limitations to be passed down."
Ya good unbiased argument against a completely new architecture.
...
When has VIA produced a stellar performer, ever? Much less on a brand new architecture..."
Well name any company other than Intel who is more then 1 generation deep into a good chipset?
funky

And again...
Oh, and SiS of course.

Originally posted by funky-d-munky
Anyone have anything to say on the topic of this thread.? Getting into the pros and cons of some of the new K8 chipsets?
funky

Sure.
Nforce3 - it's on the market, and I didn't hear criticism on it so far.
SiS - given their previous record I'd say that this will be sub performing, but reliable and cheap chipset. (question is, will somebody who want to buy new AMD cpu will be concerned about price?)
Via - they even didn't bother to design new southbridge, just slammed in the old one, crappy design. Nothing wrong with that in itself, assuming that the damn thing is working as it should (actually, I wouldn't be surprised if SiS did that too, perhaps not only now but also between P3, K7 and P4, but southbridge from SiS is working as it should...when manufacturer does that with crappy product, we have a problem)

btw, question to everybody, what's with AMD chipset boards? Will they show up? It's awfully quiet about them...and I'd trust them most at the beginning.

Originally posted by funky-d-munky
...
If the benchmarking software is bogus, it will be well known very quickly. My theory is that if they are going to put up a benchmark like that, front and center, there must be some problems with their competitors chipsets, that they want to exploit.
...
You know real questions that have nothing to do with K7 boards.
So lets keep this thread stricktly K8 MB related. It is my thread after all.
funky

Or even they're "fair", but are measuring something in which perhaps Via chipset is superior, but it's totally insignificant in real life. Probably something that's usefull for games performance (since in Via's implementation Hypertransport is only between cpu and northbridge, which is basically just agp bridge now) - in this Via boards aren't that bad
And you should know by now funky very well that, in case of Via & K8, there's a reason to talk about their K7 boards.

Originally posted by funky-d-munky
"Sis 648 chipset wasn't a bad chipset at all and the 645dx wasn't bad either."
Thats the same generation, kind of like a KT-133a and a KT-333.
I am saying from PI to P2 to P4. You know, a hole new architecture. No one but Intel has had any consistancy. So I thought it fair to point out that because VIA had some problems with their K7 chipset, does not mean their K8 chipset will suck as well.
It very well might, but you understand if I don't take your word as fact The PIT.
But of course I would gladly read any facts you have on any K8 chipset (good bad or otherwise) as thats what this thread is about.
funky

Let's see...you should know about SiS by now and about "new" architecture.
Yes, it will suck, it's the same southbridge, only northbridge is new (?), so PCI issues MUST remain.


BTW all, I've read above that Via has implemented fixes etc. in their new drivers. Currently I'm using built in winXP drivers on my KT133A with via latency patch, but will 4in1 work better finally?


edit: oh, and aye!

[Wombat]

Funky? Where'd you go? Funny, you were posting in this thread an awful lot before.....

[The PIT]

Originally posted by Wombat
Funky? Where'd you go? Funny, you were posting in this thread an awful lot before.....

Probably his boss at Via is back in and he's trying to solve some bugs in the chipset. Plus he might be designing some more fud so Via can blame other companies as normal.

[funky-d-munky]

PIT, Wombat good to see your still padding your post stats with relevant technical info. So if someone mentions via, they work for via? Yawn' get a life' yawn.

Belwarrior- "and agree with Admiral that an AMD made utility would be the better choice."

AMD is not going to release any benchmark to show deficientcies of any chipset for K8

"VIA's well-known pci bandwidth issues could be targeted easily enough so i am a bit surprised why VIA would bother to expose another chipsets' faults."

Maybe they aren't scared? I would love to see a PCI benchmark or any other benchmark utillities from competing chipsets. Like I say, if they are faulty benchmarks that news will spread like wildfire. If not what a great tool for reviewers to use. I would like benchmarks for every aspect of these boards.



Nowhere- good to see someon is doing their research. You are making some good points on some of the same chips being used. But it is also guesswork on how it will actually perform untill we get some solid reviews. It might very well suck (and I have said so in this very thread) my only defense of the product is towards people who say it will suck because their KT133 sucked. A little premature narrow minded judgement IMHO.

"And you're wrong about SiS. They've made a good P3 chipsets" if you say so. I don't recall much fanfair.

"I don't think P4's with 256kb L2 were a bad product...there's a difference between product that's helded back in performance because of design decision and the one that doesn't work properly. And I have no idea what were you thinking when you gave as an example G400 - for several months those were the best cards on the market "

I think you misread the context of the the origional post. It was products from one generation to the next. So from good to bad or bad to good. Sorry I didn't make it clearer.

"btw, question to everybody, what's with AMD chipset boards? Will they show up? It's awfully quiet about them...and I'd trust them most at the beginning."

No AMD. They are not in the chipset biz, and only in the K7 scene out of a need to fill gaps.

"Yes, it will suck, it's the same southbridge, only northbridge is new (?), so PCI issues MUST remain."

Guesswork again. But you may be right.

"Currently I'm using built in winXP drivers on my KT133A with via latency patch, but will 4in1 work better finally?"

4-in-1's have been stable for quite some time. No need for a latency patch as it is part of. The trick is when to install them. XP-> SP1 -> IEsp1 -> windows updates -> directX -> 4-in1's -> drivers.
Although I have never had problems with whatever 4- in 1's XP installs automatically.
funky

[The PIT]

Originally posted by funky-d-munky
4-in-1's have been stable for quite some time. No need for a latency patch as it is part of.

Stable don't make me laugth.

[The PIT]

Originally posted by funky-d-munky
PIT, Wombat good to see your still padding your post stats with relevant technical info. So if someone mentions via, they work for via? Yawn' get a life' yawn.

What Facts have you produced Zilch.

I have posted one or two facts but you don't like the truth do you.

Zip drives and Tape drives didn't work with the vp3 and mvps. Zip drives still didn't work properly with the KT133/133a chipsets. Can't remember about the 266/266a.

Fact 686 and the 686b suffered from faults that caused data corruption.

Fact theres a bug in the memory controller on the north bridge and this has been present from at least the KT133 days if not earlier. This maybe fixed on the KT400 onwards but I wouldn't know.

Fact George Beese has done a lot of work to make via based boards do the basic like boot up and run without corruption, cracking on audio and getting the disk transfer speeds up to spec.

So come prove that the new chipset is a totally new design and not a modified one.

Any football started so I'm off again now.

[Dr Mordrid]

I think the old saying goes; "there are none so blind as the munkys who will not see...."



Dr. Mordrid

[ALBPM]

Seems it only takes a little constructive criticism to throw a Munky into a Funk.

And this was from most of the AMD users on the Forum. I thought it was pretty mild.

[funky-d-munky]

Pit- The 4 in 1's do include the latency patch now, and have for a while. The stability issues regarding the latency patch are gone. Whats so wrong with saying that?

"Zip drives and Tape drives didn't work with the vp3 and mvps. Zip drives still didn't work properly with the KT133/133a chipsets. Can't remember about the 266/266a.
Fact 686 and the 686b suffered from faults that caused data corruption.
Fact theres a bug in the memory controller on the north bridge and this has been present from at least the KT133 days if not earlier. This maybe fixed on the KT400 onwards but I wouldn't know. "

Yes Pit, there are problems. I never said otherwise. Why are you so angry? I am the one who came here asking questions and never once denied via didn't have their share of problems. I don't know why I have been chosen the punching bag for all anger directed at via. I never once said, buy via, or via is the best or any such crap. If I did anything wrong it was posting two related topics at the same time. This seems to have unfairly branded me as a via fanatic. Instead I just ask questions as they come to me.

It is aparent that the MURC forums have become a backwoods boys club dominated by a inner circle of mega posters. I have to tell you I feel like I stumbled into the movie Deliverance. I think I will leave now before Pit and Wombat try to make we squeal like a pig.

This the only only decent response I have recieved in regards to my original post:
"Nforce3 - it's on the market, and I didn't hear criticism on it so far.
SiS - given their previous record I'd say that this will be sub performing, but reliable and cheap chipset. (question is, will somebody who want to buy new AMD cpu will be concerned about price?)
Via - they even didn't bother to design new southbridge, just slammed in the old one, crappy design. Nothing wrong with that in itself, assuming that the damn thing is working as it should (actually, I wouldn't be surprised if SiS did that too, perhaps not only now but also between P3, K7 and P4, but southbridge from SiS is working as it should...when manufacturer does that with crappy product, we have a problem)"

So after 4+ pages thats it? Ya great discussion on K8 chipsets. Sad. Good luck to all.
funkly

[magician]

To funky-d-munky:

I really don't get it... why are you so desperately trying to make a news out of clasic "man bit dog" canard.

You can't expect from people to be hilarious about that kind of statement from VIA, especially considering her (it's) long built reputation.

You are like: [abstract]You are all bunch of ignorant VIA haters! Why are you not glorifying next VIA's chipset when they clearly stated that it will be awesome? Don't give me opinions based on shity products that they have been making for years, rather discuss about how great they will be in the future.[/abstract]

Is there any moderators in this discussion bord anyway?

[Wombat]

There sure are. But there's nothing to be edited yet. It's not quite flames. He's being childish, but not offensive.

[magician]

P.S. My personal experience tells me to omit VIA chipsets whenever intense PCI load is required, and it is proved several times even with latest chipsets (KT400).

VIA has made it's way to the top of chipset makers only due to lack of real support for AMD processors. We have to admit that it meliorate over the years but it certainly can't be considered as synonym for quality and performanse, at least not when it comes to intensive I/O tasks.

Now, when AMD customers have nForce and even SiS as an option, VIA doesn't looks so good any more. And when it comes to Intel platform, VIA never couldn't do anything in that segment, simply because of upright competition.

I'm not saying that VIA's chipsets are crap in general, they are perfectly suited for vast majority of PC user: gamers, Internet surfers, maybe even office emploeeys... but when it comes to south-bridge dependant tasks VIA sucks big time. Just ask anyone involved in audio/video production or networking and you'll see what I'm talking about.

I would really like to see some truly good chipset from VIA, and when I say that I mean "good north-bridge coupled with truly good south-bridge, that have decent PCI bus implementation", maybe even with additional PCI-X, yet it's a science fiction as it looks for now.

[magician]

Originally posted by Wombat
There sure are. But there's nothing to be edited yet. It's not quite flames. He's being childish, but not offensive.

Yes, but they could simply close the thread. It doesn't leads us anyware but to meaningless flame.

[Wombat]

Oh, I don't know. It's not any worse than it was on pages 2 & 3, but if we'd closed it then I wouldn't have gotten to see Nowhere's very informative post.