Whatever happened to AMD DTX?

All about them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
dougz
Posts: 317
Joined: Thu Nov 24, 2005 3:03 pm

Whatever happened to AMD DTX?

Post by dougz » Fri Aug 31, 2007 1:37 pm

Announcements earlier this year showed lots of interest from major mobo makers. Prototypes were shown at various industry shows.

And then... nothing. Even a Google news alert is now silent.

Am I just too impatient or has the AMD DTX bandwagon stalled?

What sort of platform will the forthcoming M$ Windows Home Server run on? Traditional uATX would seem to be a bit too loud as well as too expensive to make Home Server a hit.

jessekopelman
Posts: 1406
Joined: Tue Feb 13, 2007 7:28 pm
Location: USA

Re: Whatever happened to AMD DTX?

Post by jessekopelman » Fri Aug 31, 2007 3:02 pm

dougz wrote: What sort of platform will the forthcoming M$ Windows Home Server run on? Traditional uATX would seem to be a bit too loud as well as too expensive to make Home Server a hit.
Why? It's not like this is a processor intensive application. The main sources of noise is likely to be the hard drives themselves and DTX offers no advantage over ATX in terms of HD placement/noise.

derekva
Posts: 477
Joined: Tue Jun 07, 2005 11:00 am
Location: Puget Sound, WA
Contact:

Post by derekva » Fri Aug 31, 2007 3:25 pm

What sort of platform will the forthcoming M$ Windows Home Server run on? Traditional uATX would seem to be a bit too loud as well as too expensive to make Home Server a hit.
It depends on the manufacturer (or, if you are a 'roll your own' type of person, whatever you have lying about). HP, among others, uses a mini-ITX board in their WHS box. I use a spare Opteron 144 / AsRock 939-Dual SATA2 in a 4U shallow-depth server in a wall-mounted TELCO rack in the garage (I was a beta tester, I have a copy of the release candidate and will be getting the System Builder OEM version of WHS when it becomes available).

And Doug, please don't say M$. It makes you look like a troll. Every company is out to make money (Intel, AMD, NVidia, IBM, Red Hat, Apple, etc.) - some are just more successful or ruthless than others.

-D

dougz
Posts: 317
Joined: Thu Nov 24, 2005 3:03 pm

Post by dougz » Fri Aug 31, 2007 3:38 pm

Why? It's not like this is a processor intensive application. The main sources of noise is likely to be the hard drives themselves and DTX offers no advantage over ATX in terms of HD placement/noise.
Price was cited as a major advantage -- more DTX boards/sheet than uATX boards/sheet of PCB material. Also, DTX boards would fit traditional uATX/ATX boxes, unlike mini-ITX.

Cheaper, smaller, simpler systems with less heat-producing components and lower wattage processors/simpler video. Not thinking terabyte++ multi-disc server, more like single/dual disk, simple video/headless.

No benefit to HD noise, but lower wattage PS, lower heat components, low price point seemed like a nice way to do a quiet home server.

jessekopelman
Posts: 1406
Joined: Tue Feb 13, 2007 7:28 pm
Location: USA

Post by jessekopelman » Fri Aug 31, 2007 4:11 pm

dougz wrote: No benefit to HD noise, but lower wattage PS, lower heat components, low price point seemed like a nice way to do a quiet home server.
Agreed, but I think you get all that with uATX. The huge volume advantage will keep uATX cheaper than DTX in the short run and there are plenty of uATX configurations have very low power consumption. The only issue with uATX is that it sets a minimum box sizes that may be excessive for a one or drive HD server. But if you really want small, worrying about compatibility with existing cases goes out the window.

dougz
Posts: 317
Joined: Thu Nov 24, 2005 3:03 pm

Post by dougz » Fri Aug 31, 2007 4:18 pm

Hmm, looking up the HP WHS box suggests that it may be delayed until the holidays, which may mean that I've been too impatient about DTX.

Contrast the specs of this powerful WHS prototype from Velocity http://gizmodo.com/gadgets/revealed/fir ... 280286.php
with this simpler prototype WHS DTX box
http://www.dailytech.com/article.aspx?newsid=7286

Neither is right or wrong; it all depends on the user's requirements. Clearly, DTX is targeting lower price/heat/performance, which is what I want.

I've considered VIA processors on mini-ITX, but they don't suit. I also considered the Intel mini-ITX
http://www.intel.com/products/motherboa ... /index.htm but it also doesn't suit.
And Doug, please don't say M$. It makes you look like a troll. Every company is out to make money (Intel, AMD, NVidia, IBM, Red Hat, Apple, etc.) - some are just more successful or ruthless than others.
Sorry, derekva, but we'll have to agree to disagree on "M$." I have no problem with the capitalist system; we all benefit from it. I used to work for the company that invented FUD and played extreme hardball. I'm an MCSE and earn my living supporting Microsoft product. But M$ has gone beyond hardball/ruthless and is losing the support of lots of us who used to defend it.

Good example: the current effort to rig the ISO/IEC vote on OOXML.
There is no question that all over the world the competing interests in the Open XML standardization process are going to use all tactics available to them within the rules. - Microsoft's Director of Corporate Standards

as quoted in http://www.consortiuminfo.org/standards ... 8095140796
The Swedish Standards Institute has declared its recent vote in favor of Microsoft's Office Open XML format invalid. It means that Sweden will probably abstain from an important upcoming international vote on whether to make the format a standard.

The reason given by SIS was not the controversial circumstances surrounding the vote, in which Microsoft was found to have offered companies "incentives" if they voted in favor of OOXML. Instead, SIS cited a technicality, saying proper procedures had not been followed....

The original decision to accept OOXML as an ISO standard was taken by the Swedish working group on Monday. Microsoft Sweden was later found to have offered extra "marketing contributions" to its business partners to encourage them to vote for OOXML, according to e-mails seen by Computer Sweden.
http://www.infoworld.com/article/07/08/ ... lid_1.html

FWIW, I think excessive hardball can be counterproductive. In fact, I think that even M$ may be starting to recognize that.

Post Reply