Topic: Large Bitcoin Collider Thread 2.0 - page 11. (Read 57158 times)
Congrats on finding the private key. I don't think you know but you can also claim BitCore and Bitcoin GOLD with that private key. I see you or somebody already claimed Bitcoin Cash.
I sweeped both your BitCore and BitcoinGold funds before somebody else got to it. Post a BitCore and BitcoinGold address in this thread and I will send them back to you.
Here are those transaction ids
https://chainz.cryptoid.info/btx/tx.dws?172551.htm
http://btgexp.com/tx/f78617a5e6bf438c7a310639f6d8fe10c194a74be6ac98b2c329024a4d1df745
EDIT: Please post your addresses where you want the coins sent in the 32btc contest thread since its unmoderated.
https://bitcointalksearch.org/topic/bitcoin-puzzle-transaction-32-btc-prize-to-who-solves-it-1306983
Thanks
Congrats! 0.54 BTC so about $4000 right now. I see your are estimating 40 to 135 days for the next one. Happy hunting!
Rico, what about 2 or evern 3 GPUs and how it depend on GPU speed (how it's better to have 1080ti compared to some medium GPUs? does it have sence) ?
Arulbero wrote already something about this topic in this thread (scroll back for "CPU limited", "GPU limited" situation).
At the moment a 1080Ti maxes out with some 45 Mkeys/s
a 1050Ti maxes out at some 37 Mkeys/s
"maxes out" = GPU limited situation. The GPU can't deliver more.
In this case - if your CPU is stronger - a 2nd GPU will help.
Let's say you have a 14 core Haswell ~2GHz and two 1050Ti, then you will get around 70 - 74 Mkeys/s
Theoretically, if you have a mainboard with two such Haswells and 4 x 1050Ti, you will get double that keyrate.
Same with other GPU/CPU combinations.
7820X + 2 x 1080Ti gives you at the moment 85/86 Mkeys, because each 1080Ti maxes out at some 43 Mkeys/ with that CPU.
I'm trying to find both time and funds to improve the generators (https://lbc.cryptoguru.org/crowdfunding), but interest is low, so for the time being we have to cope with these CPU/GPU limitations.
For everyone else to estimate speed:
A 4GHz Kaby Lake core does about 1 Mkeys/s
A GPU client gives you a 7x speedup compared to a CPU-only client. (Now - soon this may be 14x)
So let's say you have a quad-core Kaby Lake 4GHz and a GPU -> 28 Mkeys/s
It's only a rough estimate, but should help you to assess what to expect from your hardware.
A 4GHz Kaby Lake core does about 1 Mkeys/s
A GPU client gives you a 7x speedup compared to a CPU-only client. (Now - soon this may be 14x)
So let's say you have a quad-core Kaby Lake 4GHz and a GPU -> 28 Mkeys/s
It's only a rough estimate, but should help you to assess what to expect from your hardware.
Now I'm slecting MB and CPU only. Not GPUs. But I limited myself only with 1151 socket as 2066 CPUs/MBs overload my budget and also I don't like AMD. The second step will be GPUs but it's a little later.
What you say about https://ark.intel.com/ru/products/126686/Intel-Core-i7-8700-Processor-12M-Cache-up-to-4_60-GHz with MB https://www.msi.com/Motherboard/B250-KRAIT-GAMING.html ? I'm not very familiar with modern HW so maybe there are better options. Any more advices ? Thanks.
What you say about https://ark.intel.com/ru/products/126686/Intel-Core-i7-8700-Processor-12M-Cache-up-to-4_60-GHz with MB https://www.msi.com/Motherboard/B250-KRAIT-GAMING.html ? I'm not very familiar with modern HW so maybe there are better options. Any more advices ? Thanks.
So you're from Russia - or so? ;-)
Cannot speak for the mainboard - didn't even look it up. The CPU is ok, my guesstimate would be 5.5 Mkeys/s.
Guys, can you advice what is the best speed/price ratio from Core i5-i7 CPU family for launch LBC soft? Im going to participate and luckly it's time to upgrade my HW. Thanks.
Depends on your budget.
The https://ark.intel.com/products/123767/Intel-Core-i7-7820X-X-series-Processor-11M-Cache-up-to-4_30-GHz is a nice CPU with lots of potential.
Seems a 1950X Threadripper is even better.
Don't go for Vega{56,64} yet - lot's of trouble - AMD guys seem to need more time to provide sane drivers.
If I was to choose some MAX configuration, still in reasonable reach for the well funded individual, I'd probably go for a
1950X system with two 1080Ti GPUs - should give me 90-100 Mkeys/s now and XXXXX in the future.
If you want better price/performance ratio now, at the expense of less future potential, replace the 1080Ti with 1050Ti.
Rico thanks for answer.
Now I'm slecting MB and CPU only. Not GPUs. But I limited myself only with 1151 socket as 2066 CPUs/MBs overload my budget and also I don't like AMD. The second step will be GPUs but it's a little later.
What you say about https://ark.intel.com/ru/products/126686/Intel-Core-i7-8700-Processor-12M-Cache-up-to-4_60-GHz with MB https://www.msi.com/Motherboard/B250-KRAIT-GAMING.html ? I'm not very familiar with modern HW so maybe there are better options. Any more advices ? Thanks.
Guys, can you advice what is the best speed/price ratio from Core i5-i7 CPU family for launch LBC soft? Im going to participate and luckly it's time to upgrade my HW. Thanks.
Depends on your budget.
The https://ark.intel.com/products/123767/Intel-Core-i7-7820X-X-series-Processor-11M-Cache-up-to-4_30-GHz is a nice CPU with lots of potential.
Seems a 1950X Threadripper is even better.
Don't go for Vega{56,64} yet - lot's of trouble - AMD guys seem to need more time to provide sane drivers.
If I was to choose some MAX configuration, still in reasonable reach for the well funded individual, I'd probably go for a
1950X system with two 1080Ti GPUs - should give me 90-100 Mkeys/s now and XXXXX in the future.
If you want better price/performance ratio now, at the expense of less future potential, replace the 1080Ti with 1050Ti.
Guys, can you advice what is the best speed/price ratio from Core i5-i7 CPU family for launch LBC soft? Im going to participate and luckly it's time to upgrade my HW. Thanks.
The collider generated (and checked) over 10000 trillion keys, which is double the addresses.
I released a new version of the LBC client (1.195) - this fixes one or two bugs in Skylake and Skylake-avx512 detection.
Everyone who has either a Skylake CPU or a Skylake-avx512 CPU should update to this version, because later today I plan to put the Skylake-avx512 generator binaries for download. If you do have a Skylake CPU and an older version than 1.195, your LBC will break by downloading a skylake-avx512 binary although you only have a skylake.
Same applies if you have a Kaby Lake (as these are virtually the same as Skylakes and therefore want to use the same binary)
I released a new version of the LBC client (1.195) - this fixes one or two bugs in Skylake and Skylake-avx512 detection.
Everyone who has either a Skylake CPU or a Skylake-avx512 CPU should update to this version, because later today I plan to put the Skylake-avx512 generator binaries for download. If you do have a Skylake CPU and an older version than 1.195, your LBC will break by downloading a skylake-avx512 binary although you only have a skylake.
Same applies if you have a Kaby Lake (as these are virtually the same as Skylakes and therefore want to use the same binary)
Code:
$ LBC -u
New client '1.195-LBC.bz2' found.
Finished update run - system up to date.
Oh ok I didn't realize it was complaining about OpenCL since it was running on CPU only mode. I will fix it when I get to enable the GPU. For the 3000 Gkeys I am working on it 
It's a tribute to the unified generators (earlier we had different generators for CPU and GPU, now the binary supports both), it needs to dynamically link against a OpenCL library - but you can install a dummy one:
https://github.com/OCL-dev/ocl-icd
Most linux distributions have that as a package.
Quote
Code:
./kardashev-sandybridge: /usr/local/cuda-8.0/targets/x86_64-linux/lib/libOpenCL.so.1: no version information available (required by ./kardashev-sandybridge)
Still it is able to compute keys at 0.79 Mkeys/s (it's a AMD Athlon X2 340 ultra cheap cpu) but I guess it's complaining because the cpu is AMD?
No, it's complaining because your OpenCL installation is slightly broken.
Reinstalling OpenCL helps. clinfo should work, if you have a Nvidia GPU, nvidia-smi should work, then you are probably good to go.
The performance you see is because you run a CPU-only generator.
In this project, you have to earn your merits and perks. One of them being to be allowed to use a GPU. Which comes after you delivered 3000 Gkeys CPU-only.
Oh ok I didn't realize it was complaining about OpenCL since it was running on CPU only mode. I will fix it when I get to enable the GPU. For the 3000 Gkeys I am working on it
I was wondering if this will run on ARM cpus, I tried to set it up on a Raspberry Pi and I have an error during benchmarking:
ARM is not supported, only 64bit x86 compatible (x86_64 a.k.a. AMD64)
Quote
Code:
./kardashev-sandybridge: /usr/local/cuda-8.0/targets/x86_64-linux/lib/libOpenCL.so.1: no version information available (required by ./kardashev-sandybridge)
Still it is able to compute keys at 0.79 Mkeys/s (it's a AMD Athlon X2 340 ultra cheap cpu) but I guess it's complaining because the cpu is AMD?
No, it's complaining because your OpenCL installation is slightly broken.
Reinstalling OpenCL helps. clinfo should work, if you have a Nvidia GPU, nvidia-smi should work, then you are probably good to go.
The performance you see is because you run a CPU-only generator.
In this project, you have to earn your merits and perks. One of them being to be allowed to use a GPU. Which comes after you delivered 3000 Gkeys CPU-only.
Hi, I stumbled across this project while researching math behind the process of generating BTC addresses. Very interesting project!
I was wondering if this will run on ARM cpus, I tried to set it up on a Raspberry Pi and I have an error during benchmarking:
I have also tried to set it up on a mining rig I have and it works but I have this warning message when launching the script:
Still it is able to compute keys at 0.79 Mkeys/s (it's a AMD Athlon X2 340 ultra cheap cpu) but I guess it's complaining because the cpu is AMD?
I was wondering if this will run on ARM cpus, I tried to set it up on a Raspberry Pi and I have an error during benchmarking:
Code:
Benchmark info not found - benchmarking... ./kardashev-generic: 1: ./kardashev-generic: Syntax error: ")" unexpected
Generator validity check failed. Expected: 5 or 7, Got: 0
With output:
--
I have also tried to set it up on a mining rig I have and it works but I have this warning message when launching the script:
Code:
./kardashev-sandybridge: /usr/local/cuda-8.0/targets/x86_64-linux/lib/libOpenCL.so.1: no version information available (required by ./kardashev-sandybridge)
Still it is able to compute keys at 0.79 Mkeys/s (it's a AMD Athlon X2 340 ultra cheap cpu) but I guess it's complaining because the cpu is AMD?
Rico gave me a link to show the hooks documentation. It opened up a "secret manual" that I had never seen before!
...
Can anyone else verify the problem?
...
Can anyone else verify the problem?
To quote my answer:
I seem to be old school
When books were a thing
Big technical books. Consisting of several "Parts"
Clicking on manual reveals Parts I to IV
Hooks are documented in Part II - User Manual
Rico gave me a link to show the hooks documentation. It opened up a "secret manual" that I had never seen before!
Here is the problem with the website I am seeing. Steps to reproduce:
1. Click https://lbc.cryptoguru.org/man/user#hooks
2. Verify that you see the hooks information
3. Click trophies from the left bar
4. Click Manual from the left bar
Expected behavior: The hooks documentation is visible
Actual behavior: The hooks documentation is not visible on the page
I am using Chrome on Mac.
Can anyone else verify the problem?
Here is the problem with the website I am seeing. Steps to reproduce:
1. Click https://lbc.cryptoguru.org/man/user#hooks
2. Verify that you see the hooks information
3. Click trophies from the left bar
4. Click Manual from the left bar
Expected behavior: The hooks documentation is visible
Actual behavior: The hooks documentation is not visible on the page
I am using Chrome on Mac.
Can anyone else verify the problem?
At this point the container thinks it is done with it's task and stops. The service will replace it with a new container. This is also how Open Shift, Kubernetes, etc work.
So there was just a "groundhog day" (Bill Murray movie) scenario. A fresh container would be made from the image. It would get the update, but think it was finished.
So there was just a "groundhog day" (Bill Murray movie) scenario. A fresh container would be made from the image. It would get the update, but think it was finished.
This explanation makes perfect sense, and yes, a preceding ./LBC -u solves that particular problem.
On other topics.
We did test yesterday the first AVX512-enabled generator, seems to bring a slight speedup compared with avx2 binaries (Haswell, Syklake)
I'm looking into some other issues of the client and will make one more update to offer a skylake-avx512 generator "soon(tm)".
New client 1.192 (see News in my sig for changes);
I found the cause of the spam. I was using Amazon EC2 Container Service. With the service you configure how many containers you want running at a time. I think I had it set for 8 "desired tasks". The service will deploy 8 containers, execute their entrypoint file, when a container/task stops running, it will start a new one in its place.
I just ran the container locally by supplying "bash" as an override entrypoint.
Code:
Mac$ sh secret/run-local.sh
lbc@22be2fedf198:~$ cat entrypoint.sh
#!/usr/bin/env bash
./LBC --id $LBC_ID --secret $LBC_SECRET --cpus $LBC_CPU --loop $LBC_LOOP
./send-it.sh
lbc@22be2fedf198:~$ sh entrypoint.sh
New client '1.192-LBC.bz2' found.
New generator found. (DL-size: 0.19MB)
Some files were updated - please restart LBC.
At this point the container thinks it is done with it's task and stops. The service will replace it with a new container. This is also how Open Shift, Kubernetes, etc work.
So there was just a "groundhog day" (Bill Murray movie) scenario. A fresh container would be made from the image. It would get the update, but think it was finished.
I just reviewed the LBC -h help and see there already is a flag for update.
I think I fixed in: https://github.com/dcw312/lbc-client-docker/issues/7
The entrypoint is now like this:
Code:
./LBC --update
./LBC --id $LBC_ID --secret $LBC_SECRET --cpus $LBC_CPU --loop $LBC_LOOP
Ok, so, if I see some coins move then come here to reclaim them?
Yes.
Or better yet - directly to our LBC Discord Chat: https://discord.gg/YjD2Sx
where we can solve things more interactively.
Ok, so, if I see some coins move then come here to reclaim them?
Jump to: