move to lyra2 was absolutely great as amp required to drive scrypt-N was almost illegal
congrats to Vert team, although wish they would have acted sooner.. in general it would benefit community to study lyra2 with an understanding if an fpga/asic can be built to attack it...
no large community ever would mine scrypt-N on a massive scale (hence no wonder price have plummeted), lyra2 has a chance, amp reduction is significant
I think Lyra2RE can be done on FPGA - just depends how much you want to spend. From my understanding, FPGAs are EXCELLENT at logical operations, but fast memory for them gets expensive quickly. 6KiB is not a lot of memory; I think it would be cheap - but that assumes you only want to run one thread at a time. I think, for best performance, you'd want to run dozens, or maybe even hundreds of threads - then you'd start getting bit by the cost, I think.
Anyone with actual experience doing FPGA work, feel free to correct me/chime in.
theoretically always, but one need an expert , there could be tremendous memory requirement and fpga cost would be prohibitive although a singular manufacturer could create advantage for self
Believe above statement is behind weakening of crypto in general as difficulty level is just not there (unlike bitcoin) to have worry less
It is a tough problem
also explains ripple and nxt rise as asset management tools and rise of POS this year.. over POW.. as far as alt coins are concerned
I also second "Anyone with actual experience doing FPGA work, feel free to correct me/chime in." statement insofar as what and how long and possibly how much would it cost to build fpga for lyra2
any software/hardware prohibitive roadblocks would yield advantage (positively speaking) to an investor(s) I am serious
investors are much more sophisticated than one imagines
---