Conversation

Notices

1. Bee Yarn Knee 🙊 🇮🇸 🍏 (herrabre@mastodon.xyz)'s status on Friday, 05-Jan-2018 13:38:13 UTC Bee Yarn Knee 🙊 🇮🇸 🍏

   I suspect people trying to find alternate CPU architectures that don't suffer from #Spectre - like bugs have misunderstood how fundamental the problem is.

   Your CPU will not go fast without caches. Your CPU will not go fast without speculative execution. Solving the problem will require more silicon, not less.

   I don't think the market will accept the performance hit implied by simpler architectures. OS, compiler and VM (including the browser) workarounds are the way this will get mitigated.

   • Hallå Kitteh repeated this.
   • The Mysterious Em (themysteriousem@social.tchncs.de)'s status on Friday, 05-Jan-2018 13:57:47 UTC The Mysterious Em

     @HerraBRE
     Yup. It's a workable approach to meltdown which is a specific issue that some other CPUs don't suffer from, but not Spectre which impacts all caching speculative execution. Spectre will take a couple of design->silicon rounds, but newer will generally be better, preferably a couple of years from now.

     Hallå Kitteh repeated this.
   • Hallå Kitteh (clacke@social.heldscal.la)'s status on Friday, 05-Jan-2018 14:00:58 UTC Hallå Kitteh

     in reply to

     • The Mysterious Em
     @themysteriousem @herrabre Preferably yesterday, but probably a few years from now?
   • Bee Yarn Knee 🙊 🇮🇸 🍏 (herrabre@mastodon.xyz)'s status on Friday, 05-Jan-2018 14:18:15 UTC Bee Yarn Knee 🙊 🇮🇸 🍏

     in reply to

     Prediction:

     Software will be used to mitigate #meltdown and #spectre, slowing everything down.

     In 2020, a new line of Intel CPUs hit the market where you pay a premium for "virtual security extensions" (VSE) that let you turn off the mitigations.

     Google, AWS buy truckloads.

     Intel profites.

     Yay!

     Hallå Kitteh repeated this.
   • Hallå Kitteh (clacke@social.heldscal.la)'s status on Friday, 05-Jan-2018 18:33:55 UTC Hallå Kitteh

     • Vertigo
     • Drew DeVault
     • Sean R. Lynch ☑️
     @vertigo @herrabre @sir @seanl
     
     No. We will always want faster cores if we can get it.
   • Sean R. Lynch ☑️ (seanl@social.literati.org)'s status on Friday, 05-Jan-2018 18:42:26 UTC Sean R. Lynch ☑️

     • Drew DeVault

     @HerraBRE @sir But the speedup from speculative execution IS from parallelism. We're just asking the CPU to find it instead of the compiler. So couldn't you move the smarts into the compiler?

     Hallå Kitteh repeated this.