13 Boards with KT266A and nForce 420D @ Tom's Hardware

[Guilty]

This article really made me lose a lot of respect for THG and whoever the people are that reviewd it.

First the most glaring problems:
http://www.tomshardware.com/mainboard/01q4/011126/kt266a_nforce-19.html

"In the "office performance" category, the Asus A7V266-E is the cream of the crop - followed closely by the Soyo and Soltek."
- The A7V266-E is absolutely tied with the Soyo at 321, ?


<P align=justify>"The MSI K7N420 is the fastest board at compiling the latest Linux kernel. It only takes 230 seconds - while the QDI requires a whole 270 seconds. "
-The MSI K7N420 is not even close to the leader in this test, it's in the very middle with 215 seconds, the QDI requires 232 as well, ?

"Archiving is a very practical application. WinACE 2.04 was used under Windows 2000 to archive a 178 MB WAV file while the clock was running. The Asus A7V266 is the top dog in this discipline."
-The A7V266 is at the bery bottom, the A7V266-E comes in at #3 with 231, the A7N266 and Epox 8KHA+ are 230, ?


Next:
http://www.tomshardware.com/mainboard/01q4/011126/kt266a_nforce-18.html
"The Lame MP3 Encoder under Windows XP is used to convert a 178 MB sound file from a WAV format to a "MPEG-1 Layer 3" format. In this test, both boards based on the nForce chipset come out on top."
-The nForce boards are tied 178 with 3 KT266A boards, ?

"Again, both nForce boards have pummeled their KT266A challengers."
Pummeled? He really likes throwing this word around, the A7N266 wins by 0.26fps, I cant tell you how many times I've died in Quake and said to myself if only I had an extra 0.26fps.


Theres plenty more, but it seems like they did a few rounds of testing and wrote of the comments based on initial rounds and failed to modify them for the final numbers. Besides that though they make a mountain out of a molehill. This is Athlon XP vs. P4 stuff, 0.26fps OMG THAT BLOWS THE OTHER OUT OF THE WATER, get a clue Tom.

 

[Leokor]

Why do Tom's Hardware folks claim that the memory access on nForce boards reverts to 64-bit mode when all three slots are taken? That is not what everyone else is saying. The access will still be 128-bit wide up to the smaller amount of memory per controller (and 64-bit for the rest of the memory). But that is also true if you plug in two sticks of unequal size.

Leo

 

[Guilty]

True Leo, everywhere else I've read that also. The DIMM's dont have to be equal in size; and occupying the 3rd slot doesnt negate the dual channel's.

 

[Zebo]

Did anyone notice where the article said;

The Shuttle comes wrapped in a generic package that doesn't stand out from the competition. Among its positive features are its four DIMM slots and its six PCI interfaces. The manufacturer scrimped a little on the sound chip, preferring an inexpensive AC'97-Codec. Hardcore AMD fans will be happy to hear about how far the CPU core voltage can be adjusted - all the way up to 2.025 volts. However, the FSB clock speed maxes out at 133 MHz. Compared to its more-established competitors, the Shuttle can't quite keep up, and brings up the rear in the benchmarks.

133 FSB Max? I am at 150x10 as I'm typing this.

 

[Leokor]

<< True, to some extent, but considering all those features on the Nforce are useless to most of us who already have GF3/Radeon and Audigys and GTXPs, there is no reason for any of us to get a Nforce. >>


I wouldn't say that nForce audio is inferior to Audigy. Rather, all the way around. But for those folks that already have Audigy, yes it can be a problem.

Besides, who said that nForce was not meant to be used with GF3? There is that AGP slot o'er there.

Leo

 

[Paul]

I wish he stuck the ABIT KR7A RAID in there.

 

[AGodspeed]

This article really made me lose a lot of respect for THG and whoever the people are that reviewd it.

The use of loaded terminology is what epitomizes Tom Pabst IMO.

 

[flexy]

<< Tom must be on crack or something. The nForce boards he tested were within 1-2% of all KT266A boards except one test, and he concludes that:

KT266A Trounces nForce 420D.

I swear to god, Tom must be smoking something. Someone please explain exactly how the KT266A "trounces" the nForce 420-D boards. I'd really like to hear a logical argument. >>



funny....i was thinking the same

you should also note that the nforce boards probaly all were tested running old beta bios....

However...according to this review i begin to re-think my decision to go for nforce again...but because of many other reasons, not because his review.

() nforce keeps delaying and delaying...since the mobo is the ONLY part i need yet i get somewhat nervous

() MSI...well.....ehrm....it must not be an MSI

() i really dont like the useless, dumb features of the Asus boards...realtek NIC....etc..."even" tom said that he was dissapointed by the newer (and especially nforce) asus boards

() it is not even clear yet in what configuration the asus will be delivered (strange, eh ?)

() the asus is MUCH MUCH MUCH MUCH too expensive....and i'd almost spend $220 dollar...and NOT get the nforce NIC ? No way !!!! As i see it..if i'd go with nforce i would HAVE to go with MSI.....i dont know...

() epox 8kha+ suddenly looks attractive again (i'd have to order a soundcard, too then....)

 

[flexy]

<<

<< Tom must be on crack or something. The nForce boards he tested were within 1-2% of all KT266A boards except one test, and he concludes that:

KT266A Trounces nForce 420D.

I swear to god, Tom must be smoking something. Someone please explain exactly how the KT266A "trounces" the nForce 420-D boards. I'd really like to hear a logical argument. >>



Price vs. performance

KT266A
Shuttle AK31a = 80
8KHA+ = 105

nForce
MSI420 = 187 >>



john,

you're totally off here.

epox = $105 PLUS audigy soundcard ~~ $165, AND you get all that cr*p (eg. gf2mx) with it for free...so priceing is NOT the factor here.

The question is eg. if i'd be better off getting epox & audigy....because the audigy has EAX and i dont know if the builtin sound of the nforce has any use in real life anyway....eg . no eax...

 

[sandorski]

Perhaps he wrote the review in German then used one of them net language convertors?

Yea, over use of superlatives in such a close bench is kinda rediculous(sp).

 

[flexy]

<< i'm eyeballin the chaintek board.....

i've already decided not to get a nForce board(speed oc,ing and price)

at this point it's a tie between the the chaintek, soyo, and soltek, at least for me it is.


also any board i get has to be compatible with the Alpha 8045... >>



i got mine today (alpha 8045), neat

btw...where in the staes would you get a soltek board ? I didnt see it on pricewatch...neither do they have a section in the states. They're in europe and brazil...i think i have to disappoint soltek fans here with availability in the us of a . :/

 

[Bozo Galora]

link

 

[flexy]

Thanx for the link ! great !

hm..anyone ordered from them already ?

What would they want for shipping ?

 

[human2k]

<< Did anyone notice where the article said;

The Shuttle comes wrapped in a generic package that doesn't stand out from the competition. Among its positive features are its four DIMM slots and its six PCI interfaces. The manufacturer scrimped a little on the sound chip, preferring an inexpensive AC'97-Codec. Hardcore AMD fans will be happy to hear about how far the CPU core voltage can be adjusted - all the way up to 2.025 volts. However, the FSB clock speed maxes out at 133 MHz. Compared to its more-established competitors, the Shuttle can't quite keep up, and brings up the rear in the benchmarks.

133 FSB Max? I am at 150x10 as I'm typing this. >>




Yeah, what does he mean by "the shuttle can't quite keep up". According to anand's scores of the shuttle ak31, epox 8kha+, and msi kyt-pro 2, it does keep up. And when DID YOU START BENCHMARKING MAINBOARD BOXES? Does a better mobo box make it go faster (if it does then i want one)???? It doesn't seem like he take costs in to consideration, the AK31A is the cheapest of the crop to sport 6pci, and 4 dimms. PLUS its stable.. Oh wells, I LOVE my shuttle ak31a.

 

[GreenLantern]

That is the least technical and subjective roundup of that size that I've ever seen. Guilty says it all, glad someone had the patience to point out the problems. Someone was up too late writing that one - and I'm not going to touch the use of pummel and trounce when speaking of the most microscopic benchmark variants - whether it was in favour of the nForce or the KT266a.

I sure would like more of a scoop on what is taking manufacturing so long though.

 

[Executor]

Just ordered a soltek but from a different place. Cant wait to get the board

 

[ST4RCUTTER]

However, the FSB clock speed maxes out at 133 MHz. Compared to its more-established competitors, the Shuttle can't quite keep up, and brings up the rear in the benchmarks.



Tom must have had a pre-production board or was tokin' on the crack pipe again cause somebody needs to tell my mobo to stop running at 140Mhz...and I've just begun to overclock.

 

[GreenLantern]

<< because the audigy has EAX and i dont know if the builtin sound of the nforce has any use in real life anyway >>

nForce sound is designed to support EAX, is less CPU intensive than the SB, and will hold it's own for sound quality easily.

 

[JavaMomma]

Soltec in Canada...

NCIX

 

[Mats]

Considering the possibilities nVidia may bring with a driver or BIOS update, I would go with the nForce.

Besides, nForce is a more advanced architecture than the KT266A. I would like to support nVidia's m/b chipset initiative, it could mean greater things in the future.

 

[Mats]

"Tom must be tokin' on the crack pipe again..." - ST4RCUTTER

 

[Macro2]

TOMS HARDWARE GUIDE has gone to h$ll in a hand basket. Whomever wrote the article throws around words like "trounce" and "destroyed" like they pulled them out of a dictionary with no knoledge of the english language. On top of that, they couldn't even read a bar chart...and when they could, if one board beat another by some insignificant amout, it somehow "trounced" it?
The nForce boards seemed to be middle of the pack...but the key was that there wasn't much difference between any of the boards save the KT133 one. SO it comes down to what features you want and if you want to overclock.

Mac

 

[Macro2]

RE:"Perhaps he wrote the review in German then used one of them net language convertors?"

I think is more like those Taiwan compamies that name everthing "JUMBO" or...

Mac

 

[chasm22]

I can't see the advantage of the nforce cards. Their biggest drawback is what should be their biggest draw, i.e the graphics card. Instead of getting the "good" stuff, you get the "clunker" stuff.

OTOH, the asus A7V266-E looks solid to me. $161 for a solid overclocker with very good onboard sound from a company with a history of building rock stable cards. And, of course, the promise raid card. I don't think that price is out of line for the features.

BTW, did anybody else notice that the new abit kr7a wasn't included. This card is supposed to be a good performer from a company that has always been a major player in the overclocking field. I assume THG couldn't get his hands on one.

 

[LXi]

Guity,

Whoever did these comments are pretty stupid, however, one of the graphs/comments was interesting, if you look at the graph below this comment...

<<"The MSI K7N420 is the fastest board at compiling the latest Linux kernel. It only takes 230 seconds - while the QDI requires a whole 270 seconds. "
-The MSI K7N420 is not even close to the leader in this test, it's in the very middle with 215 seconds, the QDI requires 232 as well, ?>>

In WinACE 2.04, the K7N420 Pro got 230, and QDI got 270.