Wednesday, August 18
Daily News Stuff 18 August 2021
Splat Goes The Weasel Edition
Splat Goes The Weasel Edition
Top Story
- Well, that's one way to handle things. PolyNetwork - the crypto site that got hacked recently and lost $600 million - has its money back and has offered the hacker half a mil and a job as Chief Security Advisor.
- Meanwhile down south Reichskaren Ardern of New Zealand just locked down the entire country over one case of Bat Flu.
And Tasmania in turn has declared New Zealand to be a high-risk zone.
No matter how bad your political leadership may be, there's worse nearby.
Tech News
- Memory prices are starting to come back down. (Tom's Hardware)
Still well above the all-time low of A$3 per GB, but at that price half the world's DRAM manufacturers went out of business.
Good news for me because I'm looking to buy 128GB pretty soon.
- Package specs for AMD's Zen 4 server CPUs have leaked. (WCCFTech)
Max instantaneous power draw on the new server parts is 700W, up from 450W. That lines up with the leaked increase of maximum continuous power from 280W to 400W, and that lines up with the increase from 64 to 96 cores.
The I/O die by itself can use up to 126W, but it's driving 12 Infinity Fabric links, 12 channels of DDR5 (actually 24, but let's not go into that), and 128 lanes of PCIe 5.0 so that's not really a surprise. A lot of it seems to go to the Infinity Fabric links, since the smaller parts have all the PCIe and memory channels active but fewer links, and have much lower I/O die power.
- Zen 3 Threadripper and Threadripper Pro parts are expected to arrive soon. (WCCFTech)
I considered getting a low-end Threadripper system for my new PC, but I think I'll be better off with two Ryzen systems - running Windows and Linux respectively.
- Scientists at LLNL triggered a fusion reaction that released 10% as much energy as all the sunlight hitting the Earth. (NYT)
More than enough to power human civilisation.
For 0.1 nanoseconds, anyway, which is how long it lasted.
- A critical security flaw affects 83 million IoT devices. (Bleeping Computer)
The flaw is called IoT.
Disclaimer: Get out! Get out now! The security flaws are coming from inside the house!
Posted by: Pixy Misa at
06:42 PM
| Comments (4)
| Add Comment
| Trackbacks (Suck)
Post contains 382 words, total size 4 kb.
1
From the fusion article: "On Aug. 8, the burst of energy was much greater — 70 percent as much as the energy of laser light hitting the hydrogen target"
So still extremely energy negative. Wake me up when they at least reach the Wright Flyer level of capability. Given the massive infrastructure involved in all fusion experiments, reproducibility by independent third parties is limited and so any announcements have to be taken with a big grain of salt.
Sadly, too many "scientists" out there disdain the scientific method in favor of putting out results to keep the grant funds flowing that even potentially legitimate breakthroughs have to be looked at with a higher degree of skepticism.
So still extremely energy negative. Wake me up when they at least reach the Wright Flyer level of capability. Given the massive infrastructure involved in all fusion experiments, reproducibility by independent third parties is limited and so any announcements have to be taken with a big grain of salt.
Sadly, too many "scientists" out there disdain the scientific method in favor of putting out results to keep the grant funds flowing that even potentially legitimate breakthroughs have to be looked at with a higher degree of skepticism.
Posted by: StargazerA5 at Wednesday, August 18 2021 11:54 PM (cOJsr)
2
Some magnetic confinement setups have briefly achieved positive returns. They're just unstable.
70% sounds like good progress until you realise they need well something like 1000% to be practical.
70% sounds like good progress until you realise they need well something like 1000% to be practical.
Posted by: Pixy Misa at Thursday, August 19 2021 05:37 AM (PiXy!)
3
Thing is, this is a situation where you will probably get a lot of terrible results for a while before you start being /able/ to get good results.
If continuum mechanics is valid for this, you have a lot of data to gather before you can really know what you are doing. It seems likely that the mechanical part will be at least as bad as statistical mechanics. If even statistical mechanics is not good enough, you are going to need a lot of tests. And, the tests will probably need to be done close to the design space you are aiming for.
So, work out the nitty gritty of the physics, then figure out stability, and /then/ design principles? Gonna be a while.
That is assuming that the other messes aren't contaminating this one.
If continuum mechanics is valid for this, you have a lot of data to gather before you can really know what you are doing. It seems likely that the mechanical part will be at least as bad as statistical mechanics. If even statistical mechanics is not good enough, you are going to need a lot of tests. And, the tests will probably need to be done close to the design space you are aiming for.
So, work out the nitty gritty of the physics, then figure out stability, and /then/ design principles? Gonna be a while.
That is assuming that the other messes aren't contaminating this one.
Posted by: PatBuckman at Thursday, August 19 2021 09:41 AM (DHVaH)
4
From the practical side: the point at which they make some serious breakthrough, they'll all die from the positive feedback. Sorry guys! I hope you made good notes and that they're off-site.
Posted by: normal at Thursday, August 19 2021 11:43 AM (obo9H)
53kb generated in CPU 0.016, elapsed 0.1218 seconds.
58 queries taking 0.1097 seconds, 344 records returned.
Powered by Minx 1.1.6c-pink.
58 queries taking 0.1097 seconds, 344 records returned.
Powered by Minx 1.1.6c-pink.
