RealTechNews.com - Independent Tech

Monday, January 17, 2005

What Was Hot in 2004
Over the past nine years we have seen almost every aspect of PC design change except for the case and motherboard --until now. Intel has started again from scratch and designed a new format called BTX (Balanced Technology Extended) to take the heat out of faster computing. The components have all been rearranged so that air can cool them more efficiently, eliminating the areas in ATX cases where there is no airflow. This will also make them operate more quietly. (And the cases all seem to open from the wrong side too.) Intel created the BTX case and motherboard design so there's no obligation for rival AMD to adopt it. But with some products already available we think it is here to stay. Overheating is a major cause of damage and general unreliability in PCs, so anything designed to combat it has to be a good thing.

# Permalink Posted at 6:49 AM 6 comments Email this Link

Comments on this Item:

"Overheating is a major cause of damage and general unreliability in PCs..."

Pardon me, but I thought that Bill-ware was THE major cause of damage and general unreliability...

Followed by poorly written apps, memory leaks, interrupt conflicts, hardware address conflicts, power surges and, only after those variables are accounted for does heat soaking effect the MTBF.

Pulling data out of your butt again, Bill? Cite your sources.

The US Government (I know you don't like them, Bill) publishes a series of standards known as Military Specifications or, MIL-STD to you.

MIL-STD-202F and test method 107G (Thermal Shock (air chamber method)) has published a very nice book-sized set of data for typical CPU/Circuit Board failure rates in thermal cycling tests.

Insofar as the US Government buys a good 50% (want the page number to the OBRA for 2003 showing technology purchases in the "white budget"? Just do a search on Thomas.gov) of the high-end microprocessor output in the US - and nobody wants to tool up a line to produce sub-standard chips, then it is fair to say that the performance curves of all processors fit the spec for the year of manufacture of the device.

Now, aside from the sweeping generalization that nothing changed in case and motherboard design over the last 9 years (What is your problem? Didn't these changes happen: 1/4 size motherboards; various socket arrangements; backplane systems; liquid cooling (standard in G-5 Macs); daughter-board processor mounts; alternative thermal conduction schemes from the passive heat sink to the CPU mounted integral heat sink & fan to ramping up and down processor clock speed to limit heat production for non-processor intensive apps....) do you really have the gall to hype a press release to this degree?

The Huygens probe sits on Titan's surface today, relaying back data with 1990 technology. Somehow, 15-year-old chips withstood the thermal shocks and the physical shocks of the trip to Titan and they are still functioning. Must be because they were MIL-STD components using a series of hard-coded applications - without Windows 3.1 or OS2 muddying up the data collection process.

See, DISR info at http://www.lpl.arizona.edu/~kholso/overview.htm

In the meanwhile, have you thought about drying out a bit? Even Sir Rush requires a bit of downtime from his opiates. Consider taking a week or two to follow the big man's example and check into a nice clinic.

Todd loves you Bill....the rest of us want to save the electrons you put to waste.

# posted by Anonymous 2:29 PM

Shouldn't Todd be out on a ledge somewhere?

# posted by Bob 5:01 PM

Bob, I started out thinking that if you give someone enough rope they'll hang themselves. Then I moved on to, his comments tend to relate so little to the piece in question that it's easy to ignore him. At this point, the solution is the same one many bloggers have used, to block AOL from the blog. But that's not fair to the folk who use that service and actually have a valid point to make.

# posted by Bill 5:29 PM

Nice idea, but I seriously wonder just how much more efficient the new design is...any scientifically measured stats to back Intel's claims?

I mean in this day and age when we're all obsessed with more horsepower, and anything short of liquid cooling isn't going to help much more than adding another fan, especially since geeks want to squeeze every last bit of processor speed they can with overclocking. But even non-overclocked PCs run pretty hot. Hell, even graphics cards need to be hooked up to the power supply now.

# posted by Dolemite 8:26 PM

If you remember that ATX was also a cooling revision, BTX sort of follows logically. The idea then, as now, was to put components in the path of the air flow. Yes, you're correct, components are getting warmer, but enter the 120mm fan in BTX. These demonstrably produce 30% more airflow than the stock 80mm's defined for ATX. If you add the fact that one of the 120's is ducted from the back panel to CPU's heatsink, you also lose the CPU fan --arguably the noisiest of the bunch. So you get cooling and quiet.

# posted by Bill 8:47 PM

If this had come out when the PIV was still prematurely premature it would have made sense. Since it's only now debuting slightly ahead of Intel's Dual Core offerings I'd be afraid, be very afraid.

# posted by Anonymous 4:22 PM

Post a Comment