Who the hell is this??
Union of Cracking Gods - they had over 1M blocks/day for about a week now and only 6 team members!!
Oh, that's just me.
Ever heard of Elbrus? Its something I caught when I was looking at the stats for UCG that made me think a bit. If Elbrus does not ring any bells, think "Russian" and "x86 Processor"...
Well, actually, it is conceivable it really could be a hack.
Russian Team Coordinator sends an open letter to d.net
There's also a couple interesting threads at Anandtech and Slashdot.org discussing it.
I tend to believe that d.net should keep an eye on them at least simply because the details are fuzzy at best, perhaps even do a bit of proactive research of them.
Of course, it could just be AnandTech getting jealous.
I finally got curious enough to dig up something on this. This seems to be the best bit of information regarding the UCG RC5 team: http://rc5.aha.ru/english/rc5whs.shtml
Very interesting investigation. Consider the numbers for a moment. At the time of this post, D.net shows 3,437,294.27K keys for the overall rate of the team over the last day. My 300A @ 464Mhz gets around 1200K keys at idle so using that as a rough estimate, it would take around 2900 equivalent machines running the client to get such a keyrate. Not exactly the type of resources I would expect to find in Stavropol, Russia or any other part of Russia for that matter. The third link Joel posted above does an interesting bit of computing involving a p200 and power consumption that also illustrates the unbelievability of this team.
Honestly though, until now I had not given that team much thought, but I think I've been pretty thoroughly convinced it is a hoax. We shall see. My previous post mentioning some relation to the Elbrus E2K is also probably quite wrong, as that product is still in the 'vapor' stage by all indicators.
"Eliminate the improbable and all that remains is the truth"
Having browsed the above links, it seems several options exist:
1) the Union of Cracking Gods have organised their OWN distributed computing network, piping results from thousands of computers to a half-dozen email addresses and one team.
2) there exists a new standard in CPUs, one which surpasses all before it and makes current efforts look pitiful.
3) it's a hack.
Allow us to analyse each in turn (comment welcome - I'm playing Devil's Advocate)
1) the "many-to-one-email theory"
A post on slashdot would have us believe that the Russian city of Stavropol has pooled its computing resources in an effort to crack RC5-64.
I quote from an Iowa website:
"Stavropol Krai became a Sister State to Iowa because of the many similarities the states share: geographic size, economies based on agriculture or ag-related industries and businesses, and nearly equal populations with a few large cities and many small towns. With so many things in common it seemed a natural fit to link the two peoples and time has proven the assumption to be correct. Both states have benefited through exchanges in the fields of agriculture, the arts, business, education, government, law, medicine, sports and economic development."
Hmmmm doesn't sound like a high-tech hub, does it?
Am I mistaken, or does Russia currently have a problem paying its academics, soldiers, nuclear physicists? If this is the case, I'm sure they have better things to worry about than proving that US encryption export laws suck.
2) The Elbrus super-processor theory.
An interesting one this, and proof perhaps that a little information is a dangerous thing.
There was a recent press release (NEC or something like that) that had created the first integrated PC-on-a-chip solution. This received widespread press and media coverage.
A couple of years ago, Texas Instruments announced it had effectively jumped a generation in its transistor compression techniques. Bet you heard about that too. (What DID happen to that?)
Anyone hear stuff about Elbrus? A processor that runs faster than anything in existence, and that more than one person has access to?
Nope, didn't think so. If you had created this processor, wouldn't you want your name on it?
"RC-5-64 Elbrus Russian People's Laboratory Team" perhaps? Get the idea?
I do not doubt or belittle the Russian capability - there is a long history of solving problems in a different way to the West - the Mig29 for example - but to create a working processor to these (supposed) specs would take massive investment, and you would make **** sure that someone other than people checking RC5 stats saw how cool it was, especially when Russia needs so much inward investment at the moment.
Also, why RC5-64? Why not SETI@HOME, surely a higher profile endeavour to prove a sh*t-hot piece of kit on?
3) the hacking theory.
The one thats left, most probable, and easiest to believe. SETI@HOME is a widely-known and popular distibuted computing effort. RC5 seems to be one strictly for the techies. Who do hackers wish to impress/annoy the most? Their peers.
My money would be on this option. And they probably come from Iowa
[Footnote 22/7 :800XL posted his notes while I was writing mine, so I was unaware of it at the time. Nice to see that the same conclusion is reached by different methods ]
[This message has been edited by U-96 (edited 07-22-99).]
It apears that their 15 min. of glory may be over!
It also looks like D.net will need to plug the security holes quick before anyone else tries this.
Check out 800XL link above for the latest news.
They released large parts of the client source code publicly. That can't be good for security.
New Security Features Planned for Firefox 4
Another Laptop Theft Exposes 21K Patients' Data
Oracle Hits to Road to Pitch Data Center Plans
Microsoft Preps Array of Windows Patches
Microsoft Nears IE9 Beta With Final Preview
Simplified Analytics Improve CRM, BI Tools
Android Passes RIM as Top Mobile OS in 2Q
VMware Updates Hyperic System Management
File Monitoring Key to Enterprise Security
LinkedIn Snaps Up SaaS Player mSpoke