DOW-3 had computer probelms, DOW-6 had computer problems (half of episode 5 in season 2 is about a complete computer failure on the DOW-6)

First, I'd like to mention it's supposed to be the latest and greatest tech. (completely inaccurate)

The computers are still socket 370 pentium II/III based (the one being fixed in episode 5). the quick glimpses inside the computer case and the housing area on the truck shows really really oudated hardware, all of which even an already old Pentium D chip would outmatch in processing ability.

I would recommend creating two rack based computers each with an solid state drive to which the os will run off, one computer running on a core i7 860 with a pair of ati 4890's in crossfire (3 monitor support for each card) to run the eqiptment (dopplar radar, ect...
Then build an x58 based comp with an i7 950, 3 ati 5850's, or 3 nvidia GTX 285's running in Crossfire/SLI mode, with a raid 10 array of 8 terabyte hard drives to run as the server/processing center/datacenter for real time processing of data gathered and storage. Have the university's software engineering department create ATI Stream or Nvidia CUDA (depending on which manufacturer's cards were used) api's for the software they use to process the data into useable data/models/ect so the data collected is useable quickly (thats atleast 4+ teraflops processing on a desktop if configured properly). Whatever processing power the i7 860 isnt using can also be used to process data while idle or light tasks the way berkley's seti@home-enhanced program works using BIONIC)

the computer hardware will probably run them around $6000 - $7000 (and im speaking in canadian funds since i live up here in canada)

The 2 computers can be connected together via 2 bridged gigaethernet connections on a wireless N router/switch (to which everyone's laptop would be able to send their data and use as well if within proximity to DOW-6 using whatever wireless ability is built into their laptops (b/g/n)

data can probably sent from each scout/probe via a 3g/4g based cell network in real time to be processed (since distances are well over wireless n range, and would be more cost effective and scalable compared to prices of higher end tech)

For a university based research group, with funds coming from the government, and various private investors, I would have expected the computer systems to be a bit more decent than the 10 year old + hardware from what i've seen, that will no doubt fail not only because of their age, but because that ancient hardware is not as robust as modern hardware


thats my 2 cents. that would be more uptime, and better data processing ability for a small cost (compared to all the other hardware on that truck)

I know what you mean. There were several times during the show when i wondered why they were using such old hardware. That and when the XM link lapsed an hour, they had to radio Justin to get a radar update.. why not have a cellular backup? Also i noticed they were using some web based reflectivity images on a website.. why not just use a program like GRLevel3. Atleast you can zoom and have a better image.

Regardless of what choices they made i still love the show.. and there's always room for improvement.

I would go a step further and set up a co-op data network between willing chaser teams. Each could donate their position and certain data to the collective in order to help the effort. I'm surprised Josh at least doesn't use real-time GPS tracking of his teams to assist with probe placement; I'm sure a cleaver programmer could overlay radar data, road maps, and team location on one screen.

For the amount of effort that goes into these chases, they sure get lost a lot.

This may sound silly, but the reason for the P III architecture in the onboard computers, they can handle being bumped around inside a truck. Also, that compartment Josh is sitting in has some shielding built into it because of the Doppler Radar Dome on the back of the truck. They have to use radios to communicate because the frequencies they work on are not affected by the output of the radar when its on. Also, Josh is in business for himself and relies on grant funding, that means budget priced gear and PIII's are dirt cheap. Its the OS and Software that can get pricey. Now, if you look at the DOW 7 rigs under the Vortex 2 program, they probably have more up to date mobile processors on board. But you are still dealing with semi truck suspension systems.

I think there are some good recommendations here. With a new government funded budget in the millions, a computer set-up between $6000-$7000 should be feasable. I apologize for not being computer savvy enough to dailogue along with you about specifics. Seems to me though that without working computers all efforts for the day could get shut down and therefore makes the computer setup pretty high on the priority list. Maybe you guys should send your resumes to Discovery Channel. Perhaps the comments will encourage them to add one more to their team!

Re: KD5AWS

P3's are not the issue of being able to take "bumps" from truck suspension, the only thing that is really affected by turbulence/vibration would be optical media and hard drives (due to their mechanical basis) hence why I said solid state drives as the MAIN MEDIA (non-active storage which would not have continuous read and write happen like on the hard drive you see housing the os, but just storage space for data would be where the use of mechanical drives would be ideal), also the reason you see anti-vibration mechanism’s in high end cases for media mounting, not for the motherboard and the such. Today’s new hard drives have better error correction, read and write data faster than the old jalopies, and the read/write heads on the end of the arm no longer physically touch the actual surface of the writable media (metal plates in the hard drive) thus again, I reinforce my claim of the newer hardware being more robust than the old (A thing called Head Float was an issue with older hard drives, the more they were used the greater the head float, which over time would lead to drive errors en masse or just failure in general)
I’m also pretty sure that with those P3 based computers had just as dated hard drives. (old 40 or so GB, or even less of the older drives which were sensitive to any form of bumps)
Also, P3's would be more susceptible to excessive vibration damage because the core's back then were not encased in a metal housing like the Tualatin based P3/Celeron and up to today (AMD’s early Athon’s are another example of this), I know people that would irreversibly damage those processors just by putting the heat sink on hastily or incorrectly.
Recently I installed an intel atom n270 based microatx comp into the dash (full custom work) of an autocross track car to maintain its standalone fuel management system, it runs off a 32gb ssd and is running windows xp pro (for tested and true stability), and so far the system hasn’t had one error in its run of 3 months into the track season (which has been over for about a month and a half), and it would see excessive vibrations from the stiff suspension and high reving engine/transmission coupled with high speeds on the track.

RE: hyperlite134
Thanks for pointing it out, I knew I forgot to mention something. Add one more reason to use 3g/4g based phone networks

RE: Max
The co-op idea is an ideal, but would severely downgrade any money that these chasers get from their sponsors. You need to remember that they’re in competition with one another as well, and that the data that they’re going after may not be the same data that the other is after (different usage purposes).
Again, if it was a perfect greedless world, I totally agree with you.

Lenovo sunk a lot of money into Vortex2, see the link below for details and a list of what is being used for what.

http://www.lenovo.com/news/us/.../Vortex-Tornado.html

I agree with some of the points made here, but I also think its worth noting that those old P3 based computers seem to be doing the job ok. Most of the issues seemed to be caused by silly things like a faulty monitor lead, likely due to vibration. Throwing $6000 worth of computer hardware at the problem simply wont fix issues like that, if anything having more hardware would make it worse.

I agree SSD's would probably be a better option in regards to running the OS from, but I don't agree that old computers are more fragile, I'd probably say the opposite is true. If you kick a modern PC there is a chance you'll probably damage something in the PC, if you kick an older PC there is a chance you'll break every bone in your foot, but not the PC.

In my experience older PC's tend to be built a lot stronger than newer PC's, and its not just PC's, most older technology was built a lot more robust than it is today. The chances of you getting 20 years use out of your current PC, TV, dvd player, fridge, freezer or any other electrical product, is slim, but had you bought them 20 years ago, they'd probably still be working now.

Its also worth noting that the computers only seem to be processing information from a doppler radar, I'm not an expert on doppler radar, but given that the P3's seem to be doing the job, I'm guessing you don't exactly need a super computer. Why waste $6000 when $500 gets the job done just as well?

quote:

Originally posted by JSL Enterprises:

The computers are still socket 370 pentium II/III based (the one being fixed in episode 5). the quick glimpses inside the computer case and the housing area on the truck shows really really oudated hardware, all of which even an already old Pentium D chip would outmatch in processing ability.

Call NASA, let them know that the shuttle computers are outdated and need an upgrade from the 1 meg of ram they operate on....

They are running a RADAR, not playing Call of Duty. The older stuff is much more robust and temp tolerant..