distributed.net Client - 2.9010.495for RC5-72 encryption contest |
|
||||||||||||||
|
|||||||||||||||
Feedback Summary:
| Version 2.9010.495: | |||||
| Overall Rating: | Not rated (0.0) | Features: | Not rated (0.0) | Support: | Not rated (0.0) |
| Ease of Use: | Not rated (0.0) | Quality / Stability: | Not rated (0.0) | Price: | Not rated (0.0) |
Key to Types of Feedback:
Reviews
Troubleshooting
Usage Tips
Developer Notes
Commentary
Featured Reviews
Actual changes! - Version: 2.9012.497, 4/18/2006 01:42PM PST
paulc
Whomever posts these should mention what the newer version is all about. Since 2.901.495, the only change for Mac users is this version is reported to correctly recognize dual core "G5" chips (970MPs).
Why dnetc? - Version: 2.9010.495, 6/30/2005 12:21AM PST
(2 of 4 users found this comment useful)
lullabud
I used to participate heavily in RC5-64. When it was cracked I kept doing the OGR (which I personally think is a more beneficial project, but has no purse) until the RC5-72 client came out. When it came out it wasn't optimized for Altivec. Altivec has a hardware bit-rotate function which makes it 4x faster than a Pentium 3, clock for clock, and 16x faster than a P4. Since this RC5-72 core wasn't optimized for Altivec I was getting terrible speeds. I e-mailed the guy who developed the RC5-64 core and asked him to make an RC5-72 core with Altivec tweaks. What he told me was very enlightening... He told me that there were better things to spend my CPU cycles on. Cracking encryption does have its applications, but RC5-64 took 4 years to complete. Each additional bit doubles that time. According to Moore's Law we will double CPU speed every year. Take those two points into consideration and you come to the conclusion that if we all waited 6 years after RC5-64 was cracked we'd then need roughly 3 years to finish RC5-72. Otherwise we'd need several more years to finish it. So, assuming we'll crack it in a few years, then what? ... My darling, sweet, beautiful two year old niece was recently diagnosed with leukemia. ... There are distributed computing projects that are dedicated to finding cures for diseases. Projects like Folding@Home, Genome@home, and other's which help research cures for diseases that plague mankind himself, the very fabric that makes us live, not the pitfalls of the things which we have created and the disagreement about actions we take on each other. I agree that restrictions on the freedom to secure your digital information and technological assets is a problem, but I think anybody who is participating in distributed.net needs to rethink the power of any statement that could be made by succeeding in cracking a specific encryption algorithm as opposed to the discoveries and progress which could be made towards overcoming global afflictions and understanding the building block which we are made from and the weaknesses within them.
Not optimized for G5 yet. - Version: 2.9005.484, 9/19/2003 04:54PM PST
J. Charles Holt
Here's the benchmark results on my G5 1.6 as supplied by Crunching Cows:
[Sep 20 00:07:24 UTC] OGR: Benchmark for core #0 (GARSP 5.13 Scalar)
0.00:00:16.35 [14,991,058 nodes/sec]
...
[Sep 20 00:08:08 UTC] RC5-72: Benchmark for core #5 (KKS 604e)
0.00:00:16.99 [3,469,951 keys/sec]
I found mention of a version 2.9006 on distributed.net that said it had been optimized for the G5, but it wasn't in the download list. And unfortunately, due to the slewing on the G5 it gets quite noisy when crunching due to the increase in fan speed.
[Sep 20 00:07:24 UTC] OGR: Benchmark for core #0 (GARSP 5.13 Scalar)
0.00:00:16.35 [14,991,058 nodes/sec]
...
[Sep 20 00:08:08 UTC] RC5-72: Benchmark for core #5 (KKS 604e)
0.00:00:16.99 [3,469,951 keys/sec]
I found mention of a version 2.9006 on distributed.net that said it had been optimized for the G5, but it wasn't in the download list. And unfortunately, due to the slewing on the G5 it gets quite noisy when crunching due to the increase in fan speed.
Most Recent Replies: View All 2 Replies
- Not optimized for G5 yet.