![]() news news archive SEND NEWS!
articles
PC Database
SETI@HOME Team
|
31-May-2002 - Review by James "Agg" Rolfe Review unit provided by Altech Computers, a distributor with offices in Brisbane and Sydney. It was a long time coming, but the lack of an official DDR333 standard didn't stop many manufacturers of memory and chipsets from producing products. VIA's KT333 is substantially similar to their previous KT266A chipset, but adds support for 333MHz DDR memory, hence the name. The front-side bus of AMD's latest CPU's still runs at 266MHz and apart from the ATA133-supporting southbridge (which many KT266A boards were already using) there's little new in KT333. Not that that's a bad thing - KT266A is a proven, very fast and stable chipset and already the chipset of choice for DDR SocketA. We first saw Soltek about a month ago, with their SL-85DRS2 P4 DDR motherboard which impressed us overall. With this KT333 board they continue their approach of minimal extra features for low cost. No onboard video here, no RAID or USB2 controllers and no other extra features clutter the board. Onboard sound is present but it's the AC97 codec, which is unexciting for all but the least discerning of audio users. In fact, probably the most striking thing about this board at first glance is the colour: ![]() click to enlarge It's an unusual colour but in reality quite a bit darker than shown on Soltek's website. It's actually quite similar to the PCB colour of Asus's GF4 Ti4600 card (reviewed here) and they would make a good pair in a windowed case. Soltek seem to be following the current trend of feature-identifying stickers (saying "ATA133", "DDR333" etc) on their motherboards when shipping - this one having no less than four stickers scattered around the board, including one over the CPU socket warning about using a decent heatsink and fan. In terms of layout there's not too much to report - 3 DIMMS for DDR SDRAM and 5 PCI slots are pretty common these days. I'm not a big fan of having the "long" lug-side of the socket along the top edge of the motherboard as it can make attaching a cooler tricky in a minitower case. Given that this motherboard only has 2 IDE headers, it's fair to assume a lot of them are going to spend their lives in mini-towers. The power connector is in a spot that some people don't like, but I don't mind as I generally cable-tie the ATX power lead to the back of the case which keeps it out of the way of the CPU cooler. However, owners of long video cards such as GeForce4 Ti4600 and Ti4400 units will have to be careful of the IDE header location as they are very close to the end of the AGP slot. As seen in our Asus V8460 review, the tall capacitors on those GeForce4 cards can interfere with IDE cables on this motherboard: ![]() click to enlarge This isn't a major drama, but it does make routing cables to your IDE devices tricky if they are high up in a case, or if you have short IDE cables. Soltek's SL-75DRV4, their KT266A offering that is otherwise very similar in layout to this board, had the floppy connector directly in line with the end of the AGP socket, between the IDE headers and the RAM slots. This has been corrected for the SL-75DRV5 seen in this review. There's plenty of room on either side of the socket - some tall capacitors not far from the bottom lug-side, but they shouldn't cause a problem with most heatsinks. The hefty MC462A fits on this motherboard so you'll probably be right with everything else - bearing in mind again that the socket is close to the top edge of the board. No less than 5 fan headers adorn the board and there is an under-lug protection sticker under the centre lug only on the "long" socket side. A personal gripe of mine is with motherboards that require you to remove the AGP card before you can add or remove RAM modules - the Soltek has plenty of room there, so that problem is not present. Soltek have used what's popularly referred to as a "3-phase power solution" on this motherboard. This topic always spurs some debate, but the general concensus is that, by spreading the load of power regulation across 3 channels of components, component heat is reduced which is good for the quality of power and hence stability. "3-phase" is a misleading term as a friend (Boky) pointed out. 3-phase power implies some kind of AC weirdness going on when, of course, this is all DC power. 3-phase in this instance refers to their being 3 channels, or "phases", through which the regulated current flows. You can see the 6 MOSFETs and other groups of 3 components in the image below. You might also want to replace the stock thermal paste with Arctic Silver or at least apply it so it covers the whole North Bridge chip properly: ![]() click to enlarge Under the North Bridge cooling fan, heatsink and thermal paste is the KT333 North Bridge itself. Note that Soltek are using the CE revision, which is apparently more stable than the CD revision and earlier. Some people are going so far as to refer to this as a KT333CE board, but I don't know if the differences of this stepping warrant that kind of attention. Still, it's usually good to have the latest stepping, in theory with a few wrinkles ironed out:
The South Bridge chip is VIA's VT8233A. The main thing that this brings over the VT8233 that formed part of KT266A is native ATA-133 support. This means that the normal IDE headers on the motherboard will support the higher transfer rate of ATA133. In fact, Soltek used this A-model South Bridge, as did a few other manufacturers, to provide native ATA133 support on their KT266A board. We have some investigation into the differences that ATA133 brings here. The other major differences that KT333 brings over KT266A are of course DDR333 memory support and a 1/5 FSB/PCI divisor. This lets the PCI run at the standard 33MHz at a FSB of 166MHz - promising for overclocking, but also an indication perhaps of AMD's intentions in the future. Another interesting feature of this board is ABS II. This stops you locking your brakes up when.. hangon a second. :) No, it's Anti-Burn Shield II, also referred to as SmartDoc in the BIOS. You might have seen the videos floating around the net of someone pulling the cooler off a CPU on various platforms and seeing what happens (usually, lots of smoke). AMD's response was to say they were making a system to read the thermal diode inside the CPU core (Palomino onwards) and shut the system down when it reached a certain temperature. Soltek's ABS II is basically that. You enable the system via a jumper on the motherboard and by default when the diode reports 85C the system turns itself off. I'm not brave enough to test Soltek's engineers in this regard by yanking the heatsink off - but [H]ard|OCP did, with mixed results. This feature is certainly better than nothing and seems likely to protect your system in most instances of fan death or even spontaneous heatsink popoff. The BIOS menu for it allows you to configure the system to turn off when CPU fan rpm's drop below a certsain threshhold, or you can configure a different shutdown temperature. Other Recent Content:
|
Major Sponsors:
|