Hi - I was just wondering two things:
The first one is: how many workunits should I expect to crunch before I crack any of the code?
And the second one is: If I do crack any of the code, how will I find out?

Thanks for any answers you may have,
From Chris

---

1. We have actually cracked about 1/3 of available workunits (you can check that in server status page). Every one of the rest workunits can be the result (chances that the workunit you're cracking hides the original message are 1/151392644 at the moment), but if there's a transcription error, we can simply never found the original message. As I calculated, we will finish cracking current available workunits base in the end of 2017.

2. The project administrator (TJM) will post that there at the forum.

Thanks jaro, and It'll be interesting to see what'll happen at the end of this batch - I suppose we'll just start a new badge, and ok, so if we cracked it, it would be well publicised :D

---

It would be awesome if we'll find the original message as quick as possible and we won't have to pass through all workunits base. And, indeed, there will be solution for this cracked message.

Moreover, I really regret that this project can't be computed on GPU's, cause this would really boost project speed and so shorten the time to go through all workunits.

It would, wouldn't it?!

And I agree, (I don't have a GPU capable of working on projects myself), but I agree that it would make it A LOT faster...

---

And I agree, (I don't have a GPU capable of working on projects myself), but I agree that it would make it A LOT faster...

Not necessarily, it depends on the application being used and the type of computation.

---

Click Here to see My Detailed BOINC Stats

True - I didn't think of that...

---

It would still be faster than working on CPU, moreover, many of users could engage their unused cards, and so computation on the CPU+GPU configuration is many times faster.

I agree, jaro, but in a way, I like it CPU only, because it means you have to work for your credits, but if you got a GPU coming in, it's easily finish a task every 15 - 20 minutes...

---

That's the deal. But we want to make it as fast as possible, it's good that GPU does computing much faster. You don't do any work, your PC does :) I understand that collecting points would be quicker on GPU, but this can only force users to use it in their computations, and that's good.

It's completely normal that quicker device will guarantee you more points. When GPU does 2 workunits at the same time as CPU does one, it only means that GPU should get 2x points. That's the deal there.

It get we want it as fast as possible, but I also want as many credits as possible :D

As for the bit when I said you have to work for the credits, you knew what I meant ;) and it is more than 2x quicker - my 4 core laptop can do (on average) 1 task per core per 2 hours (8 per 2 hours), but if a GPU can do a task between 15 and 20 credits (probably even faster on some GPUs), it is 4 - 8 times faster...

---

When I came back to BOINC last year, I decided that when there was a CUDA client for a project, I would only do CUDA (not OpenCL) work for that project, since GPUs are easy to replace if/when something bad happens.

The only two CPU-based projects I do are Enigma and Yoyo OGR and that's because they are long-time favorites of mine. It would be great if I could take at least one of them off the CPU.

I agree with Jaro in that we should get a GPU client for Enigma. CPU-based distributed computing is dying out because of both mobile devices and the increasing power and influence of GPUs on desktop machines. We can't rely on 2017 as an end to this project, we should develop a GPU client or it will be much longer.

---

-Dan Q

danq.co

@danq

Workunits left: 149 315 484
Workunits done last hour: 6766

That means that project will be done in:
22068 hours = 920 days = 2,5 years

2nd April 2015 + 920 days = 8th October 2017

So, the end of the project is predicted to happen somewhere in 4th quarter of 2017 (using CPU only) . Of course, it doesn't mean that we can't find the answer somewhere earlier.

By adding GPU power, we could crack the whole workunits base somewhere in 2016, but ending this project in the end of 2017 is semi-guaranteed.

Workunits left: 149 315 484
Workunits done last hour: 6766

Where are you getting the "workunits done last hour"? Is it "workunits validated last hour"? In less than 2 months, that number is now only 951.

---

-Dan Q

danq.co

I see 5499 at the moment. Yes, I'm talking about validated ones.