Wonder if this project could have WU's more light, something which could be crucnhed in around 6 hours at the most.

It is a very important project, with very clear potential to help in short time the researches. But since was started in WCG, became the last one in the crunchers preferences due it restrictions (pre requisites limits, hard usage of machine mainly in upload/download conections, and the veryb long time to crunch the most of WU's in any kind of machine).

If possible change the scheme, I guess that the researches would take a look in the ways to do it asdap, and if done, for sure the project will gain a lot of adepts, and will increase very quickly your share in the total WCG participation and preference of crunchers. Think about it.

Wish look and best regards.

Dear GIBA and dkt,
Our impression is that the "opt-in" status is the biggest problem for the project in terms of participation and support. Many casual users are scared or just don't want to be bothered with a project which is not entirely standard and hence decide to pass on it, although their hardware (except for very low-end machines) could easily handle CEP. We talked to the IBM team about this, but there does not seem to be a favorable resolution. We don't want to cause problems which may make people leave WCG entirely.
Ok, and now once more to the structure of the wus: We could have cut down the number of jobs but once calcs on a particular molecule are running on a host it is very efficient to keep it running there. The later jobs in a wu are not as important as the earlier ones (although they do improve our overall computational characterization), but the point is that they run in only a fraction of the time and without any extra network traffic when we run them in sequence within the wu. The 12h limit was introduced as a compromise for users who don't want to keep a job on their machine for too long and cash in on points regularly rather than in big chunks.
We have started working on the next generation of wus which will target different properties and which may have fewer restrictions...
Best wishes
Your Harvard CEP team

Dear andzgrid,
we are sorry to hear that you seem to be unhappy with CEP2 (although we don't quite see from your badges how you can claim much first-hand experience...). Please let us comment on a few points you raised:

Yes, indeed we believe that. Due to the ‘opt-in’, no cruncher who joined WCG before the CEP2 launch participates automatically and many may not think about or want to be bothered with actively changing their settings. Newbies on the other side usually want to get WCG going to test it out - they will often leave out the one ‘hard project with the warning sign’. Many of those may equally not come back later to opt-in to CEP2. So, unlike participant of other projects, CEP2 crunchers had to at some point decide to join and deliberately change the necessary settings, which many casual crunchers will realistically not do.

No, we are not calling for that. As we wrote, we contemplated this and other scenarios with our friends at IBM/WCG and came to the mutual conclusion to leave things as they are. But that also means that the relatively low CEP2 participation is still an open issue.

Your implied charge that CEP2 is a hard-drive killer is frankly silly. The Q-Chem code is a sophisticated research software which has been in development for 20 years and which runs in labs and on clusters around the world. No IT department would touch it, if it were a danger to their equipment. If it makes you feel better: CEP2 is also running on all the personal computers (including the laptop this post is written on) of the CEP team and our friends and families…

Just to make this entirely clear: We certainly don’t blame our friends at IBM for anything either, if that's what you implied! They are a great team and we are extremely grateful for their help and support. They are awesome in accommodating our research and finding solutions for complicated problems. We could not wish for better partners.
Best wishes
Your Harvard CEP team

The extremely small kernel of truth behind the hard drive wear is simply that the hard drive is being used. Distributed computing in any form for any science will shorten the life of the system it is running on.

But this is negligible at best. Over 99% of the people will have discarded their computer for other reasons, such as being too slow, before they will see any impact from decreased life expectancy. As for SSD hard drives, there may be a concern with the increased read/write cycles, but they are still too new to have any accurate data.

TLDR version: don't worry, you'll toss your computer long before it wears out.

CEP2-related wear is not something I'm worried about.

Me, I've had hardware failures on systems running CEP2 - 'cuz I accidentally cocked the ATX power cable to a mobo and it eventually vibrated loose enough to arc. (Motherboard, power supply, and hard and SSD drives survived...although I did have to chisel some burnt plastic out of the way to reconnect everything). But that had nothing to do with the workload CEP2 places on a system!

Then, I lost an SSD out of a RAID stripe set on that same system running CEP2 - because when I was taking care of my blunder with the power cable, I forgot to plug a data cable back into one of that RAID set's disks...then when I plugged it back in, I took the lazy way out and didn't pull the unit off the rack...causing me to unintentionally cock the data cable to the other drive. When it vibrated loose, the RAID set dropped out.

But again, nothing to do with CEP2!!

In summary, I've not had any hardware failures attributable to CEP2, and I always run at least 7 tasks on hyperthreading 4-core CPUs, 11 tasks on hyperthreading 6-core CPUs, and all cores on non-hyperthreading CPUs (i.e., if the thing has 4 cores, I run 4 tasks). When ambient temperatures drop in the winter, I'll go back to running all cores, all threads on all CPUs.

I.e., I worry far more about heat than mechanical wear; watch your system temps with something like CoreTemp, and if you see Tj temperatures hitting 65 or 70 degrees Celsius, either increase/make cooling more efficient or reduce the number of tasks that you're running.

I'm motivated , you see. In light of the fact that every single [insert unit of currency] spent on energy by a scientific research effort is a [insert unit of currency] taken away from that research effort, I personally would call the search for cheap and freely available energy the single most important existing scientific research effort on the face of the planet - perhaps even the single most important task all of humanity currently faces. Particularly since individual humans spend money on energy that they could be spending on medicine, food, shelter, education, etc. etc. etc. Not to mention that the current energy sources are each vulnerable to the random extortion of the incredibly variable and absolutely uncontrollable private taxes called "profits" by any or all of the energy harvesters, the energy distributors, the energy converters, and - worst of all - the speculators. Not to mention entire economies and populations have already been negatively affected by precisely such extortion.

So I'll carefully (lolll...sorta...obviously) watch how my hardware behaves while it quite happily and obliviously crunches CEP2 units.