• Should we try cooling the planet with sulphur dioxide? (Japan Times)
  
  I mean, we know that it works. What's the catch?

  "The whole notion of spraying sulfur compounds to reflect sunlight is arrogant and simplistic," Canadian environmentalist David Suzuki said. "There are unintended consequences of powerful technologies like these, and we have no idea what they will be."

  Yeah, we have no idea what would happen if large amounts of sulphur dioxide were suddenly released into the upper atmosphere because such a thing has never happened before.

  Raymond Pierrehumbert, an atmospheric physicist at the University of Oxford, said he considered solar geoengineering a grave threat to human civilization.
  
  "It's not only a bad idea in terms of something that would never be safe to deploy," he said. "But even doing research on it is not just a waste of money, but actively dangerous."

  I'll just pause here to say that this is a wonderful case of nominative determinism, because this is precisely what you would expect to hear from someone named Raymond Pierrehumbert.

  Opponents of solar geoengineering cite several main risks. They say it could create a "moral hazard," mistakenly giving people the impression that it is not necessary to rapidly reduce fossil fuel emissions.

  In other words, I don't want to solve the problem. I want global communism.

• The Starbook from Starlabs is a 14" burger with the lot. Starlabs)
  
  A 4k screen, the Four Essential Keys, a 16 core Intel 165H processor, up to 96GB of DDR5 RAM, an M.2 2280 SSD (there's also what looks like an M.2 2230 slot but that may be for wifi), two Thunderbolt 4 ports, three USB ports, HDMI, a micro SD slot, and a headphone jack.
  
  It comes with a choice of six flavours of Linux preinstalled, or you can load Windows onto it yourself.
  
  
  
• The Ars commentariat is absolutely foaming at the mouth at the thought that Congress might have to write laws that actually say what they mean. (Ars Technica)
  
  This is probably the craziest I have ever seen them.
  
  
  
• Intel knew about its CPU oxidation issue in late 2022. (WCCFTech)
  
  That's 18 months before the company told customers about it.
  
  Some people have asked if Intel CPUs can be considered reliable once the new microcode update is out. Unfortunately, we don't know, because information from Intel itself has not been reliable.
  
  
  
• Though Puget Systems says it has seen more failures with AMD chips than with Intel. (Tom's Hardware)
  
  Sort of. Worth taking a look at the article if you plan to buy.
  
  Failures in Ryzen 5000 CPUs seem high, but those are approaching four years old now, much older than Intel's 13th and 14th gen chips. Failures in Ryzen 7000 are also high, but those failures are almost all before systems are sold to the customer.
  
  The best reliability on this chart comes from Intel's 12th gen chips, which you basically can't get anymore.
  
  
  
• Zen 5 benchmarks show that hyperthreading provides 18% better performance for 2% more power consumption. (Tom's Hardware)
  
  Intel is removing hyperthreading from its own chips to make the performance cores smaller. Instead the plan is to add more efficiency cores - which never had hyperthreading - in the space saved to get back that performance.
  
  That will certainly work, but it's also part of why Intel's consumer chips don't have AVX-512 support right now. The efficiency cores don't have it, only the performance cores, and that made software support just too complicated. Moving a task from one core to another in the same chip could cause it to fail.
  
  
  

Posted by: normal at Sunday, August 04 2024 11:00 PM (bg2DR)

Posted by: PatBuckman at Monday, August 05 2024 02:32 AM (rcPLc)

Hide Comments | Add Comment