• 1 Post
  • 141 Comments
Joined 1 year ago
cake
Cake day: June 2nd, 2023

help-circle





  • You don’t have to dig up the roads to fix buried power lines any more than you have to tear up your walls to replace power lines in your house: you install a conduit (basically a pipe) under the road once and if the cable somehow gets damaged and needs to be replaced you can just run new cable through the existing conduit by simply pushing it in on one end and pulling from the other.

    Transformers and other non-cable equipment are typically housed aboveground in little boxes or built in to the house, so they’re actually easier to maintain than if they were installed aboveground on a pole since you don’t need a cherrypicker to access it.

    Obviously in a less wealthy small town with existing overhead infrastructure it doesn’t make much sense to move it all underground “just because”, but if you’re already trenching under the road to install water/sewage/gas mains, it won’t cost much extra to throw down an additional one or two smaller conduits for running power cables or telephone/cable/fiber lines.




  • Traditional graphics code works by having the CPU generate a sequence of commands which are packed together and sent to the GPU to run. This extension let’s you write code which runs on the GPU to generate commands, and then execute those same commands on the GPU without involving the CPU at all.

    This is a super powerful feature which makes it possible to do things which simply weren’t feasible in the traditional model. Vulkan improved on OpenGL by allowing people to build command buffers on multiple threads, and also re-use existing command buffers, but GPU pipelines are getting so wide that scenes containing many objects with different render settings are bottlenecked by the rate at which the CPU can prepare commands, not by GPU throughput. Letting the GPU generate its own commands means you can leverage the GPU’s massive parallelism for the entire render process, and can also make render state changes much cheaper.

    (For anyone familiar, this is basically a more fleshed out version of NVIDIA’s proprietary NV_command_list extension for OpenGL, except that it’s in Vulkan and standardized across all GPU drivers)









  • You’ve made me uncertain if I’ve somehow never noticed this before, so I gave it a shot. I’ve been dd-ing /dev/random onto one of those drives for the last 20 minutes and the transfer rate has only dropped by about 4MB/s since I started, which is about the kind of slowdown I would expect as the drive head gets closer to the center of the platter.

    EDIT: I’ve now been doing 1.2GB/s onto an 8 drive RAID0 (8x 600GB 15k SAS Seagates) for over 10 minutes with no noticable slowdown. That comes out to 150MB/s per drive, and these drives are from 2014 or 2015. If you’re only getting 60MB/s on a modern non-SMR HDD, especially something as dense as an 18TB drive, you’ve either configured something wrong or your hardware is broken.


  • This is for very long sustained writes, like 40TiB at a time. I can’t say I’ve ever noticed any slowdown, but I’ll keep a closer eye on it next time I do another huge copy. I’ve also never seen any kind of noticeable slowdown on my 4 8TB SATA WD golds, although they only get to about 150MB/s each.

    EDIT: The effect would be obvious pretty fast at even moderate write speeds, I’ve never seen a drive with more than a GB of cache. My 16TB drives have 256MB, and the 8TB drives only 64MB of cache.