Tom revisits KT266... VH's reaction....

[LXi]

http://www.tomshardware.com/mainboard/01q2/010416/index.html

http://www.viahardware.com

 

[LXi]

^^^^^^

 

[NFS4]

Well, Tom was a bit harsh with his comments...

 

[LXi]

Was Tom referring to VH or OCers?

 

[NFS4]

I'm quite sure he was referring to VH...Overclockers doesn't have a VIA slant as far as I can tell.

 

[LXi]

Thats what puzzles me, OCers publically scrutinized Anand and Tom, and yet Tom retaliates to VH which hasnt done that much talking IMO.

 

[NFS4]

<< Thats what puzzles me, OCers publically scrutinized Anand and Tom, and yet Tom retaliates to VH which hasnt done that much talking IMO. >>


Tom posted this article before the Overclockers article was published. I think that Tom was referring to comments made by VH last Thursday/Friday.

 

[LXi]

OCers posted two editorials, the first one is published on the 13th, before Tom's revisit.

 

[KevinH]

Hehe, gotta luv the drama. Good updates there LXi, I've been following this matter myself as you've seen me poking my nose in your other thread . Anyhow I'm not at all surprised that Tom would attack Via hardware. His delusionals of grandeur has definitely blinded him to a lot of things and random attacks seem somehow fitting.

 

[LXi]

I see what you mean by &quot;POS site&quot; now.

 

[KevinH]

Did you read the latest OCer's observation of Anand's review? He came up with the same things that was bouncing around in my head when I was reading Anand's review. He never ran benchmarks with ONLY the mod. It was mod + bios update. I sure would have liked to have known if the mod was the source of the increase or was it A BIOS UPDATE ON A FUNCTIONING CORRECTLY PLACED RESISTOR BOARD. Reread the AT artciel on the update again. The article just doesn't sit well with me.

 

[LXi]

Yea me too, I noticed it but didnt realize the importance until I read the OCers commentary. IMO Anand really should've done the unmodded + beta BIOS.

 

[KevinH]

Does anyone know what Anand uses in his torture tests?

 

[Anand Lal Shimpi]

LXi and KevinH

The only problem I had with running the unmodded + upgraded BIOS setup was that the system was incredibly unstable, and considering that every one of our benchmarks with the exception of cachemem and Quake III Arena were run under Windows 2000, I wouldn't have been able to paint a very accurate performance picture with the unmodded + upgraded BIOS setup.

Scientifically speaking, yes, it would make more sense to have done an unmodded + upgraded BIOS setup however without a stable platform to run the benchmarks on that wasn't a possibility unfortunately.

Another thing you'll want to take into consideration is that OCWorkBench has reported that the original person that submitted the R126/R127 modification did so after finding that the new BIOS did not improve performance.

Basically from the standpoint of the consumer, the conclusions should not change, the MSI K7T266 Pro and the KT266 in general are not mature enough platforms at this point in time to consider. When they do become more mature we (and I'm sure other sites will too) will provide recommendations accordingly, but until then we just can't.

Take care and thanks for the feedback guys, it really does help us make this an even better place for you all.

Anand

 

[ElFenix]

he speaks!

maybe the other sites should become more mature too?

 

[KevinH]

Whoa, I guess going to the front page pays off. .

I won't dispute the fact that this chipset has some stability issues as I'm seeing that from a lot of user testimonials. As an enthusiast tho I gotta try this one out. I guess I'll have to twist the arm of our IT guy to order a board for one of our game testing rigs here and maybe poke around for myself.
Keep up the good work and I'm glad u did do an update!

 

[formulav8]

You still could have done a much better job. I definitely see where OCers is coming from.

 

[Anand Lal Shimpi]

formulav8

That's what I'm here to do, listen to your comments and suggestions so that I can improve the articles here. I appreciate your comments and do take them to heart, hopefully I'll be able to change your opinion in the future with upcoming articles

Take care,
Anand

 

[formulav8]

Thankyou for listening. Thats the main reason I like coming here.

 

[KhakiBoy]

Anand

My hat definitely goes off to you as it's not easy to run a site like this and provide reviews to everyone's satisfaction. However, it definitely would be helpful to have a way to compare if it was the mod that gave the performance increase or the BIOS update. Like you pointed out though, the original mod came about because the BIOS update alone didn't provide enough of a performance increase. It'd just be nice to have some numbers to back this up. I understand that the unmodded/updated BIOS boards were unstable and thus unable to be benchmarked. Did you try a modded board (R126) with the original (1.00) BIOS? At least this way we can compare the 1.00 BIOS with the mod/unmodded board and then use the other benchmarks with the other BIOS updates to extrapolate the results, though not as reliably.

 

[Anand Lal Shimpi]

KhakiBoy

Let me see where I can grab a copy of the original 1.00 release bios and I'll try doing just that. I won't be able to tonight because I promised myself I'd work on this server article, but I'll definitely get to it tomorrow if I can get my hands on the BIOS.

I'll either provide numbers as an update to the article or post them here at least if they aren't significant enough to update the entire set of graphs with.

Take care,
Anand

 

[MrHelpful]

Have you figured out what the positioning of the resistor actually does? I'd love to know.

Keep up the good work.

 

[KhakiBoy]

Anand:

I sent you an email with the 1.00 version of the BIOS. I've used it on my own system, so I know it works. I'm curious to see how the right/fixed/modded version of the board benchmarks with the original 1.00 BIOS.

 

[Rand]

With results of the &quot;fixed&quot; board using the 1.00 BIOS should give us a better impression of howmuch the fix alone benefits performance. Assuming it is stable enough with the 1.00 BIOS to get some results that is.

 

[KhakiBoy]

Ok, I did my own informal review of the &quot;wrong/unfixed/unmodded&quot; board based on the different BIOSes available. Since I'm running WinMe I haven't run into the stability problems others have using Win2K. I flashed my BIOS and set the BIOS setup values to an optimal setting (CAS 2, 4-way interleaving, T1 enabled). Here are the results:

BIOS 1.00
ALU: 495
FPU: 648

BIOS 1.09
ALU: 541
FPU: 717

BIOS 1.10
ALU: 497
FPU: 645

My readings are still lower than what Anand got on his updated review. Maybe I'm missing a setting in the BIOS or something. I ran the benchmarks 3 times and took an average. The 1.09 BIOS definitely offers the best performance. I can't comment on stability since I did not run the system long enough. Right now I'm still using the 1.10 BIOS since it was the last one I flashed but will probably switch back to 1.09 and see how the stability is. In the meantime I'll just sit and wait to hear what MSI plans to do for us who have the &quot;wrong&quot; board.

Here are my system's specs:
MSI K7T266 Pro-RAID
1.33GHZ AXIA Thunderbird
256MB PC2100 Crucial RAM
3D Prophet II GTS Pro 64MB
SB Live 5.1 Platinum
27GB Maxtor 7200rpm ATA/66