https://www.one-tab.com/page/kWr8HPkRTE-PQsLDY31vKg
This is a companion discussion topic for the original entry at https://level1techs.com/video/level1-news-may-2-2018-cryptohypocrisy
This is a companion discussion topic for the original entry at https://level1techs.com/video/level1-news-may-2-2018-cryptohypocrisy
Good one… The Village Idiot approved :)~
1 Like
(chip dev)
I smell and architecture change in the mid term. It’s obvious that x86 is being milked for all it’s worth. Moore’s Law just isn’t being satiated any more. There’s not really enough to go on to make any confident predictions; but the small amount of competition that exists is producing some interesting results. The trends are in the parallel architectures of NEC’s parallel super computers and Adapteva’s parallel super PCs. That’s what is in development and producing usable machines.
Of course ARM is funding a lot of the development with sales in mobile devices; and it seems like the consumer electronics industry would like to focus on mobile tech, for various reasons. This is particularly interesting because Adapteva’s parallel chip is a parallel accelerator for an ARM chip. Adapteva has been fairly silent concerning development lately though. I don’t want to be too presumptuous; but this began with development of 64 bit chips. It may be that they’re struggling with their 64 bit chips.
The main struggle with architecture changes is probably rooted in financial risk. It seems that now the one taking the risk on the chin would be the first one to offer the change. Now that the hype train is beginning to slow significantly; that point may turn 180 degrees. If a new prospective architecture were to show a great deal of promise over the current architecture, it may be that the one not to offer the change is playing a risky game.
Since the development of a chip still has a price tag of upwards of 100 million USD, it’s probably not likely that a startup could be the bearer of solutions. It’s just not likely that a startup would have the funding to produce a viable proof of concept. It may be that some middle ground is needed to solve the issue.
Adapteva has theory and math for 4,096 cores. Keep in mind, it’s an accelerator chip for ARM processors.
(cryptocurrency)
ICOs may be the reason that crypto is becoming more accepted. The big players love derivatives; and ICOs are novel in that sense. Many have suggested that ICOs would be treated as securities and there may be some silent scheming (there definitely is) for usage of ICOs. Licensing ICOs could solve much of the speculation and volatility issues that exist now.
ICOs were developed as a more common platform to mitigate extraction. Aggregation is a huge problem right now; and the big players may be more concerned about legislation that might be more inconvenient. For instance Dylan Ratigan has been suggesting blending; which no self-respecting financeur would allow. ICOs would allow big finance to distribute their influence.
Keep in mind, this is a very energy and computation hungry endeavor.