Ya'gotcher AMD 760, VIA KT266A and 'yer VIA KT333? Huh? Is this here now?

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
This is not a "Should I wait or buy now?" thread. This is a WTH is this and when's it coming out. Rumor had it that EPOX was to release their KT333-based mobo this month (8KVA?) but now that's scrubbed.

Do they even sell PC2700 memory? Prolly like $200 for 128mb at today's memory prices.

Also, we all know that Via comes out with a chipset, and it has a regular numerical designation; ALA KT133, KT266. Then, they make all improvements to the chipset that they should've done in the first place and suddenly the chipset works really, really well, ALA KT133A, KT266A. me\<--is a KT133 owner...damn 100FSB...:(

Is it the KT333 or KT333A that is being released? Just curious.
 

Jeff H

Golden Member
Oct 11, 1999
1,611
4
81
MichaelD, you're correct in that unless you absolutely have to buy a board, you want to avoid the first iteration of a Via chip, and instead wait for the "A" version. However, if I recall correctly what I read last week, you won't have to wait for the "A" to arrive, as none of the board makers are going to release the original version. Rather they're going to move right to the A version. This may put board availability back a few weeks, but it should be worth the small wait. The bigger question is if the A version brings anything significant to the table, compared to KT266A.
 

johndoe52

Senior member
Aug 12, 2001
773
0
0
Pretty much what I've heard is that the kt333a chipset is just the kt266a only with offical support for a fsb of 166.
 

Athlon4all

Diamond Member
Jun 18, 2001
5,416
0
76


<< Do they even sell PC2700 memory? >>

There is OCZ and Kingston I believe PC2700 out and really, it's no more expensive than Crucial (maybe $10).

<< Is it the KT333 or KT333A that is being released? Just curious. >>

I honestly dunno, but if KT333 is what it was originally (266A+DDR333 support), then forget it.

Really, here's my take on this whole KT333(A) fiasco. KT333(A) from a performance standpoint is nodda, nothing unless mobo manufacture's provide 1/6 PCI dividers (I'll get to that later). The Athlon cannot benefit at all from DDR333 on the 266fsb platfrom, and I guarantee you that VIA isn't gonna pull another 266A out of their bag for KT333(A) . Really, as far as I'm concerned, KT333(A) will be made only by it's features and that means throw KT333 out because it won't have USB 2, ATA/133, AGP 8X, or V-Link 512. That's all for KT333A. And even then, there's a good argument for those feature's not even mattering because lets look at these features one by one:

USB2: On some current boards including K7T266 Pro2-RU, Gigabyte's 266a solution and a couple others I believe.
ATA/133: On many boards in the form of ATA/133 RAID.
AGP 8X: Look at how little AGP 4X brought to the table. AGP 8X will offer nothing in performance, just in future proof
V-Link 512: Not even worth taking into account.

Really, KT266A has been and will be "the" Athlon chipset prolly until it's departure. Sure KT333A will eventually replace it, but it was prolly the most significant chipset released for the Athlon. Really, the only thing that KT333A will offer is features on-chipset, which will do nothing for us, and only make it easier for mobo makers to put USB2 and ATA/133 on their boards. And, also. If board manufacturer's implement 1/6 PCI dividers, then KT333A may become something worth getting for ocing because you could hit 166fsb just like IWill XP333 but it'll be faster 'cause KT266A's memory controller is faster than MAGiK-C's. We'll see, but for me, KT333A is nothing to get excited about.

<< Pretty much what I've heard is that the kt333a chipset is just the kt266a only with offical support for a fsb of 166. >>

Nope. There is no confirmed 166fsb support. Only DDR333.
 

goog

Golden Member
Sep 8, 2000
1,076
0
0
As was posted here looks like the Inquirer got it wrong, VIA KT333 is still on the slate and the KT333A (now KT400?) will come later.

ocworkbench.com
"Unconfirmed news from our source reveals that there will be no more KT333A. The version which with improves the performance from KT266A will be named KT333CE. The version with AGP8X and Vlink 533 will be named KT400."

x-bitlabs.com
"The exact launching date for KT333 hasn?t been announced officially yet, however our source claim that it should take place on February 20"

EPoX 8K3A+ KT333 mainboard
EPoX 8KHA3 KT266A RAID, ATA133 = mid Feb.?
EPoX 8KHA2 KT266A RAID, ATA 100 = cancelled.
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
Thanks guys. Just as I thought; the KT333 is just a bunch of smoke and mirrors. And Athlon4All, you're right; the KT266A is the most significant chipset for the Athlon. Why, when I compare it to my KT133 *snicker* I_just_wanna_cry. ;) me\<--still stuck in 100MHz FSB-land....and the buses have stopped running :( LOL!

Everyone knows the KT133 chipset overclocks about as well as a fishing lure....right! It doesn't! And I've got one of the ORIGINAL, original 1GHz Athlon B's...it doesn't like a 10.5 multiplier, even at max voltage. 1GHz is it...unless I want to be kicked out of UT/Q3/MaxPayne in the middle of a game. NO thanks.

Anyway, I'll be upgrading this week..still undecided B/T an Abit KR7AR and a Shuttle AK35GTR...still researching. Will be pairing it with a 1800XP and 512mb of Crucial PC2100. Should be a significant step up for me.

Thanks to all for the info. :)
 

Athlon4all

Diamond Member
Jun 18, 2001
5,416
0
76
I hear ya MichaelD about the 100fsb and KT133. I'm on KM133 now, and I can't oc my Duron for nothing by fsb (and I don't even have multiplier or vcore on my board!!:disgust::|) but oh well. Soon, I'll be upgrading to nForce and plan to oc my Duron 800 to 133fsb for starters, and then to 1GHz or so:D. So, I hear ya. Oh and you got a PM comin' in a few mins.
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
Why NForce? Is it that Twinbank thing that's gotcha goin'? :D Or the built-in kick-butt sound? The KT266A platforms are just as fast if not faster than the NForce...but that's cool that you're taking the Path Less Traveled...something different.