Comments: "Then and Now: Athlon Platforms Compared

[Spitfire_x86]

This is another example THG's recent poor articles. Though it sucks less than "Benchmark Marathon: 65 CPUs from 100 MHz to 3066 MHz", it sill sucks. Both ideas were good, but the articles sucked finally.

1)Why GeForce2 MX400 is used to benchmark these mobos? What kept him away from using a faster video card like GeForce4 Ti4600 or Radeon 9700 Pro?

2) Why they selected AXP 2000+ for bechmarking? Was it impossible to use a faster 266 FSB AXP for this benchmark?

3) Why ABIT NF7-S as reference nForce2 board, since ASUS A7N8X is the best nForce2 mobo.

<b> "You can put lipstick on a pig, but hey, it's still a pig!" - RobD </b>

 

[Quetzacoatl]

1-No idea, i'm up for any guesses

2-I don't think the Kt133a would support anything beyond the 2100+ Palomino, if that even

3-The A7N8X is one of the worst boards I have seen, has had nearly as many reported problems as the K7S5A. The Abit is fast, but doesn't overclock as well compared to the Epox or MSI even

Instead of Rdram, why not just merge 4 Sdram channels...

 

[NIKOLAS]

Which platforms were running at 133/266 FSB and which if any, at 166/333 FSB?

I didn't feel the article made it clear what the FSB situation was.

 

[tombance]

I agree. Why does everyone say the A7N8X is so good even after hearing all of these reports of problems? And no, you cant come up with the excuse that it is this way because the A7N8X is the most popular board, as I know of many people who bought the Abit equivalent who have never had any problems. You never hear problems on THG about the Abit boards do you? I think the A7N8X is just a bad board.

<A HREF="http://service.futuremark.com/compare?2k1=5467618 " target="_new">Almost Breaking 12k!!</A>

 

[jdn999]

Yeah, like the "Which Operating System is Best For Games" article that wound up showing games tended to play the same under all operating systems - in terms of bench marks. And said nothing about which were more stable or which required more fiddling to get to work or which operating system played a wider variety of games. Wee, show me some bench marks for a few games and call that an article. In depth coverage - for a bird bath...

 

[longhorn]

Another thing about this review was the memory
configurations. It appears that every board was
tested with a single stick of 256MB PC2700 DDR.

-The KT133A won't even work with DDR.

-Several of these boards only support PC2100.
So was the PC2700 stick run underclocked?

-Were the boards that support DDR400 only run at
DDR333? I have heard that running the memory
and the FSBs synchronous may be faster anyway
(DDR333/166FSB may be faster that DDR400/166FSB)
but the article never explains.

-If only a single stick of RAM was used, the Nforce2
chipset cannot take advantage of Dual DDR.

A pretty shoddy piece of work, IMHO.

 

[phsstpok]

1) Doesn't matter which video card. All testing was done at 640x480 anyway to take video out of equation.

2) Don't understand what you are asking. XP 2000+ is 266 FSB. Did you mean why not 333 FSB?

3) Arbitrary decision.



<b>99% is great, unless you are talking about system stability</b>

 

[phsstpok]

2-I don't think the Kt133a would support anything beyond the 2100+ Palomino, if that even

I have a Tbred B XP 1700+ running at 2.1 ghz (15 x 140, ~XP2600+) on a KT133A mobo right now.

<b>99% is great, unless you are talking about system stability</b>

 

[Spitfire_x86]

2) Don't understand what you are asking. XP 2000+ is 266 FSB. Did you mean why not 333 FSB?

They could use AXP 2400+ or AXP 2600+ (266 MHz FSB)

Submit your opinion <A HREF="http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=28537#28537" target="_new"> Should Tom Fire Omid? </A>

 

[Spitfire_x86]

3-The A7N8X is one of the worst boards I have seen, has had nearly as many reported problems as the K7S5A. The Abit is fast, but doesn't overclock as well compared to the Epox or MSI even

It may suck in real world, but THG still thinks it's the best nForce2 board. So what's the point for using other nForce2 mobo as reference?

Submit your opinion <A HREF="http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=28537#28537" target="_new"> Should Tom Fire Omid? </A>

 

[LumberJack]

Bah.. you try running a P1 200 Mhx with a 9700 pro.. the whole idea is to provide as even a basis for benchmarking as possible. This means using a 1x2x4x graphics card which can be used in ANY board with an AGP slot! Problem solved.

To err is human... to really screw things up you need a computer!

 

[Spitfire_x86]

Another noticlable thing is terrible gaming performance of nForce (1). I was very surprised to see that

Here's link to original nForce benchmarks. Both articles used GeForce3 instead of GeForce2 MX400

<A HREF="http://www.anandtech.com/chipsets/showdoc.html?i=1535&p=11" target="_new"> Anandtech nForce benchmark </A>

<A HREF="http://www17.tomshardware.com/mainboard/20011126/kt266a_nforce-17.html" target="_new"> THG nForce and KT266A mobo comparison </A>

In these articles, nForce is only very little slower than KT266A and Epox 8KHA+. But is this Athlon platform comparison article, nForce is performs lower than SiS 735/AMD 760 and in Comanche4, even lower than KT133A. But in other benches, it performed up to mark. Is the graphics card guilty for poor gaming performance with nForce? Does nForce punish GeForce2 MX400 to show it's IGP as good as GeForce2 MX400? But it doesn't look in the Anandtech's nForce IGP benchmarks. The only reason of lower performance in nForce m/b may be the use of IGP instead of discrete GeForce2 MX. But THG said nothing about it

Submit your opinion <A HREF="http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=28537#28537" target="_new"> Should Tom Fire Omid? </A>

 

[jimbo99]

I've read just about every A7N8X mobo thread here and at nforcerhq.com and have concluded that the vast majority of the issue people have has nothing to do with the board itself. In the majority of the situations the issue are with the stuff put into the board and not the board itself along with a heavy dose of inexperience.

After reading the threads about "why the A7N8X" is a crappy board I have concluded the same as above.

Having worked with several of them myself, having flawless results, I have concluded that I have real practical experience (significantly more then the most common nay-sayer posting here) with the board.

There are even some that have complained loudly that have retracted their complaints and now applaud the board.

Junk in, junk out.

 

[Quetzacoatl]

That's what i'm thinking, that maybe if the indeed used the IGP version, the integrated graphics (even if turned off) were sucking some of the mem bandwidth. I don't think it would have that much of a performance hit though, especially at low resolutions. Consistency, doesn't seem to be THG's strongpoint

Instead of Rdram, why not just merge 4 Sdram channels...

 

[Quetzacoatl]

I try to think of Toms Hardware Guide as a <b>enthusiast</b> website, although lately, it's been more like a business/speculation website, which is very disappointing. Honestly though, if they wanted a comparison of the Nforce2 platform with the other chipsets, they might as well have used the best of the best, not the buggiest.

Instead of Rdram, why not just merge 4 Sdram channels...

 

[Spitfire_x86]

There's no reason for thinking that nForce IGP sucks memory bandwidth even IGP is not used. Check next two pages of the Anandtech benchmark.

I read Anandtech more than before. Previously it wasn't in the favourites menu of my browser, there was only THG. Now Anandtech has made it's place there. There's big difference between "Tom driven THG" articles and current "Omid driven THG" articles. One reason is Tom wrote most of the stuff, currently he doesn't. The another reason is Omid is only serious about his worthless articles/rants. He cares about the "truth" more than the article quality of the site. He doesn't take more than one day to respond to my poll by writing a column, but doesn't try to correct the article "Benchmark Marathon: 65 CPUs from 100 MHz to 3066 MHz" which is full of wrong info and pics.

Submit your opinion <A HREF="http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=28537#28537" target="_new"> Should Tom Fire Omid? </A>

 

[Spitfire_x86]

-Several of these boards only support PC2100.
So was the PC2700 stick run underclocked?

This is most possible

-If only a single stick of RAM was used, the Nforce2
chipset cannot take advantage of Dual DDR.

From the high performance of nForce2, it seems that they used Dual DDR at 266 MHz sync with FSB. With async operation or in single channel more, it isn't that faster than KT400/KT333.

Submit your opinion <A HREF="http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=28537#28537" target="_new"> Should Tom Fire Omid? </A>

 

[Quetzacoatl]

I don't trust Anandtech either, they're better, but not perfect. Looking at my friend's MSI Nforce2 IGP board, when I put in an external geforce3, the performance always lagged by 2-3 percent compared to my Epox Nforce2 SPP board. Same memory speed, sychronous at 333, 333, and yet there was a performance gap

Instead of Rdram, why not just merge 4 Sdram channels...

 

[tombance]

Yeah but still, if people put these 'things' into an abit or MSI board it would not cause a problem. So your basically suggesting that the Asus board has major incompatability problems? and that the only way to get it to work is to chuck away all of your hardware and buy stuff tyat is compatible with the A7N8X? I [ersoanlly would rather just keep my hardware and buy either the abit or MSI board which dont seem to have incompatibility problems. I believe that the few people who have managed to get their board going have only managed to do this by getting the lucky compatible configuration, and are just fortunate enough to own no pieces of hardware that 'trouble' the A7N8X.

<A HREF="http://service.futuremark.com/compare?2k1=5467618 " target="_new">Almost Breaking 12k!!</A>

 

[Spitfire_x86]

So which site is perfect (or almost perfect)?

Submit your opinion <A HREF="http://forumz.tomshardware.com/community/modules.php?name=Forums&file=viewtopic&p=28537#28537" target="_new"> Should Tom Fire Omid? </A>