Encryption Projects as SU group

[sdrfgs]

4140588288 vs my 4072284032 for 2 copys combined seems right to me since my 3060ti is slower card
for you running 2 copys as the same time i would estimate 2200-2500 each

 

[Lak7]

There is no need to run any longer than the first report of "crunching along at ..."

 

[Sunrise999]

I read at Tom's Hardware Blackwell 5090 in about 10 or 12 month to be release that will be a game changer.
Your Exe worked very fast in my view.
Thanks

 

[moonbase]

I read at Tom's Hardware Blackwell 5090 in about 10 or 12 month to be release that will be a game changer...

The RTX 5090 is supposed to have 24,576 CUDA cores and release date of Q4-2024
Massive jump up in cuda core count and if clock speeds take their expected increase this card could be an absolute biss ripper.

My guess estimates based on latest tests with an RTX 4090 and comparison with supposed spec of RTX 5090 suggest the RTX 5090 might process a full biss range in three hours via multiple instances.
It could be even faster as it will probably be PCIe 5.0 capable compared to the RTX 4090 which is only PCIe 4.0 capable, next generation PCIe will boost speeds.

Definitely on the shopping list, will wait for an upgrader to have buyers remorse and try to pick one up a few months old.

 

[sdrfgs]

I think the way forward for brute cuda for now and for the future. Would be a master slave server format pooling resources of users together , some of those old mining setups are cheap these days. Imagine the speed if 20 mixed gpus are working on the one project spread over multiple users.

e.g you get get allocated a range from a pool, based on your gpu compute capabilitys

 

[moonbase]

I think the way forward for brute cuda for now and for the future. Would be a master slave server format pooling resources of users together , some of those old mining setups are cheap these days. Imagine the speed if 20 mixed gpus are working on the one project spread over multiple users.

e.g you get get allocated a range from a pool, based on your gpu compute capabilitys

It is not impossible but something like this is highly unlikely to happen, it would require collaboration from users and someone to initiate, coordinate and manage the project.
The other issue is global time zones and availability of multiple users across time zones. For example, person in USA wants CW from 89W in the evening, Euro people and possibly Asia people are not around to run their allocated range from the pool.

Anyone serious about this with aspirations to get a brute CW for a satellite feed in time enough to use it on the feed would setup their own multi GPU rigs and remain independent.