Announcement

Collapse
No announcement yet.

The Performance Impact Of MDS / Zombieload Plus The Overall Cost Now Of Spectre/Meltdown/L1TF/MDS

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • carewolf
    replied
    Originally posted by xfcemint View Post

    Peak MIPS is a bad way to measure performance.

    Realistically, a 486-DX4 100 MHz is about 100x slower than a modern single-core.
    Let's see, 40x times higher clock-frequency, and an average IPC these days of two something vs single-cycle architecture's fixed 1 IPC.. Yeah, checks out. Almost exactly 100 times slower on average.

    Edit: Though if you measuren float-point performance, and SIMD optimized benchmark scores, then a modern processor can get up to being a few thousand times faster.
    Last edited by carewolf; 19 May 2019, 03:23 PM.

    Leave a comment:


  • angrypie
    replied
    Originally posted by F.Ultra View Post

    Good luck with that 486 of yours, if you are lucky you can cram about 70 MIPS out of the old beast while a modern i7 can give you aprox 100000-300000 MIPS depending on model and clockrate still with all these mitigations enabled. Anyone claiming that the mitigations here brings the performance back to or even worse than a 486 have never actually used one or have a really rose tainted memory.
    It's enough to run Doom and that's all I care about.

    Leave a comment:


  • RussianNeuroMancer
    replied
    Cybersecurity researchers at the Vrije Universiteit Amsterdam, also known as VU Amsterdam, allege that Intel tried to bribe them to suppress knowledge of the latest processor security vulnerability RIDL (rogue in-flight data load), which the company made public on May 14. Dutch publication Nieuwe Ro...

    Leave a comment:


  • Cybmax
    replied
    Originally posted by xfcemint View Post
    Gamers are a minority, but a very vocal minoritiy. As usual, their priorities seem to be revolving about producing greater integers in applications called 'benchmarks'. The problem is that some people are going to listen to their brilliant logic about what's important and what's not.
    How come it is "ok" to loose performance due to security patches that fix some hypothetical scenario in the computing world... and even being called somewhat of a dumbass for pointing that out... While VW owners loosing power and torque after the "dieselgate fix" is applied have a perfectly valid reason to demand their money back?

    I mean, in the world of IT almost anything is acceptable, while when it comes to most other stuff in the world you would never ever get away with something like that. How would it look if you got to your VW dealer and complains about loosing power, and the response was "Oh we are sorry we messed that up, but you could always buy the new 2019 model!"? This is something that seems totally oki when it comes to computers...

    Leave a comment:


  • F.Ultra
    replied
    Originally posted by angrypie View Post

    A 486 is faster now, since it isn't affected by any of those flaws. Time to dust off my DX4-100.

    As for those flaws being "highly unlikely" to be used in the wild: there's a thing called "off-the-shelf malware." And let's not forget that Spectre is a flaw that runs deep inside the uarch, so it's highly likely there are more exploits to be discovered. With those flaws all over the press there's more incentive to exploit it.

    Of course Intel knew it all along. If a much smaller company like AMD could foresee the nastiness and take measures to secure their CPUs, so could Intel.
    Good luck with that 486 of yours, if you are lucky you can cram about 70 MIPS out of the old beast while a modern i7 can give you aprox 100000-300000 MIPS depending on model and clockrate still with all these mitigations enabled. Anyone claiming that the mitigations here brings the performance back to or even worse than a 486 have never actually used one or have a really rose tainted memory.

    Leave a comment:


  • angrypie
    replied
    Originally posted by xfcemint View Post
    I would mostly prefer if the greedy bastards at Intel & friends did their job properly on the processor that I paid. Why can't they replace my CPU with a propery working model? I mean, it didn't say "full of holes" on the box I bought. Why am I not entitled to a CPU replacement? What a scam.
    Never gonna happen, they can "fix" everything in microcode now. They begrudgingly replaced FDIV-bugged Pentiums because that wasn't a possibility at the time.

    I really support Quinn Norton's motion of holding those companies liable in court for vulnerabilities like this. (Never gonna happen either, sadly.)

    Leave a comment:


  • Cybmax
    replied
    So, is there any good proof-of-concept webpage that i can visit that would instantly rip all my passwords and install some rootkit on my Ubuntu machine? Just so i can test the awfullness of using "mitigations=off" vs. default.. using regular browser with all adds disabled.

    Cos having a exploit that requires A and B and C to run, while version X of program D is used.. ONLY if E is alligned with jupiter, and its the 29th or february is kinda vague even for those in the tin-foil-hat departement

    It's kind of a moot point to have 20 locks on your front door and spending 20 minutes locking yourself in and out of your house... if your back porch door is open - ie. Actually logging on to things like Facebook/phoronix/twitter++ from your main computer... which kind of is rather plausible if you ditched the tin-foil-hat a while back. There will probably be a multitude of ppl rambling about that "You should not use the internet from your main computer anyway", but the reality is that 99.9% of ALL internet users, they actually do. Learning to NOT click on every "You need to update your contact details on your bank by clicking this link that points to http:/iwill.steal.yourdetails.cn/" would be a better advise in the long run, and having all the mitigations in the world wont save you from being stupid.

    Some questions then remains: How many uses "mitigations=off" (or equivalent) + surf the web in a "normal" way? Is it only weird ppl that actually cares for reaching 144+ fps (for that 144Hz monitor), and NOT settling for 121fps, when you KNOW your hardware is good for it? Are those ppl total whack-jobs? Is it THOSE ppl that ruin the interweb for the rest of the world?

    Will it just boil down to every religion's statement of "Well.. you have free will, so if you WANT to experience eternal damnation, be my guest!"

    Leave a comment:


  • Bsdisbetter
    replied
    Originally posted by xfcemint View Post

    Nice. Will it be backported to older LTS kernels? I hope so. Some or all?
    Run... Run to the hills, mitigation is off!!!!!

    Leave a comment:


  • Bsdisbetter
    replied
    Originally posted by xfcemint View Post

    Good thing that I got that transcedental meditation typeless keyboard for entering passwords and credit card numbers. I just imagine the number in my head, and the meditation keyboard transmits it to any web shop transcendentally.

    If you type the number on the keyboard, it can be stolen. If you copy-paste it, same thing. There can be no workaround, except for transcendental meditation keyboard.



    To do: move my Phoronix forum password from browser store to the wallet. Because if I lose it, my life becomes worthless.



    Um... so I have to decide, and old 386 or mitigations. Let me think, I'm not quite sure...

    I would mostly prefer if the greedy bastards at Intel & friends did their job properly on the processor that I paid. Why can't they replace my CPU with a propery working model? I mean, it didn't say "full of holes" on the box I bought. Why am I not entitled to a CPU replacement? What a scam.
    I'll leave you with this thought:
    You hand your credit card information to ANYBODY whether personally or via entry, you are giving them your information. If you think this potential/now mitigated via microcode exploit leaves you more vulnerable than before, then more fool you.
    You have more chance of having your information hacked from a third-party storage center than your own machine. But, hey, place that tin foil hat on real firm now.

    Oh, and your apparently non-working cpu works just fine, has done and will continue to do. You want a refund because of a bug? LOL, that's hilarious.
    Last edited by Bsdisbetter; 19 May 2019, 06:37 AM.

    Leave a comment:


  • Code Artisan
    replied
    https://make-linux-fast-again.com/

    Code:
    noibrs noibpb nopti nospectre_v2 nospectre_v1 l1tf=off nospec_store_bypass_disable no_stf_barrier mds=off mitigations=off[URL="https://make-linux-fast-again.com/"][/URL]

    Leave a comment:

Working...
X