Folding Forum

http://www.intel.com/buy/us/en/product/ ... tech_specs

looks like it acts as a processor in itself, with operating system and all. correct me if im wrong, but this seems like it would work on FAH.

point wise I would not expect much from it; seems a 4 Core-1.4GHz Atom.

But actually what would worry me more is the heat. My company once tried "Windows-To-Go"-USB Sticks with Windows 8; those got really HOT and failed quite quick.

If you want to use FAHControl on it that sure could work well.

Yeah, no reason why not - I thought about getting an 8 way HDMI switch with a row of these connected, but not really worth it given where cpu folding is with respect to even budget GPUs.

I've got similar 'net-top' systems running BOINC with these low power quad core atom processors and they are surprisingly potent for what are basically phone/tablet SoCs, but sometimes need a bit of help with cooling to maintain their boost frequencies and performance. I'd imagine the 'stick' format would be even more susceptible to throttling and need a bit of air being forced over them.

I'm not sure that's an USB key ... it looks more like a HDMI key, like a Chromecast ...

Yea, my bad, it definately is a hdmi. the 99 dollar one, or cheapest one with windows has the same cpu as the 150 dollar one. The linux one is like 80 bucks. Besides that, definately could help if your looking at saving electricity versus every other option on the market. One question I would like answered, as I could be as wrong as when I thought MSI helps write code for distributed computing projects. But my number one question would be is there a major difference between the amount of time spent reading CPU and GPU results as in on the folding @ home side? I thought I read this at a certain time, but honestly if their isnt a difference it would alter my perception on FAH. thanks.

I didn't understand your question, but think I can answer it just the same. The Folding CPU client is 15+ years old. It is about as efficient as it can get, with the exception of adding AVX support, which is in development. And FAH had the very first Distributed Computing GPU client, and is well experienced there also. Sorry, we don't do a lot of comparisons to other DC projects. The volunteers here focus primarily on helping FAH donors.

Thats not exactly it, What I was getting at is this,

Is there a difference in the time spent on the Stanford side reading already finished work unites. As in difference between cpu and gpu completed work units on Stanfords end at all. Hell, I would be interested in even a difference in time it takes to compile them at all.

Green wrote:http://www.intel.com/buy/us/en/product/ ... tech_specs

looks like it acts as a processor in itself, with operating system and all. correct me if im wrong, but this seems like it would work on FAH.

It looks like it's a complete system.

The question about whether it would work on FAH is a good one, and the heat question is important, but the Atom series of CPUs are typically used in laptops so they're a low power-low performance choice. I can't find anything in the specs mentioning SSE or even FLOPs, which are essential to FAH.

Sony's Android client might be adaptable to it but i doubt that it can compete favorably with, say, an i5.

The price is nice, but I'm not quite sure how I'd use one.

Since meaningful results takes months and sometimes years to collect and then process the completed work units, I'm not sure it really matters if a completed work unit sits for a few hours or a few days before it gets combined and studied. But it doesn't sit unstudied until all the work units are completed in that particular project number. Help us understand what prompted the question or how it affects your folding choices.