I'm not sure about this one as I don't know the details of the test but first of all, one 64MB block doesn't mean much, it needs to show sustained rate. Secondly, some things don't...
... scale linearly. So just because we can do 33 to 65 doesn't mean that we'll go faster. If you don't parallelize the block verification how will you take advantage of multicore CPUs?
Thanks for the link. I did go through it. So SV is planning to add multithreaded verification of blocks. But it doesn't answer how you can handle a stream of 128mb blocks right now.
Who said anything about a stream? Blocks are produced on average every 10 minutes. Most decent servers can process 128MB block in about 5 to 10 seconds.
... it. But if you say that most servers can already handle it then cool. Will they be able to handle it with DSV as well btw? Is it more complex than normal sigverify or same cmplxty?
It should be the same complexity as normal sig verify but I guess you can put many of them in one TX. Otherwise Ryan's point about DSV being a subsidy doesn't make sense.
Yeah it's a stream of blocks, one per 10min. Ideally you'd want the median server to be able to *easily* handle processing the blockchain so that people can build services on top of...