New Hardware Vulnerability Found in Intel Processors

Published by

Click here to post a comment for New Hardware Vulnerability Found in Intel Processors on our message forum
#5506576
https://forums.guru3d.com/data/avatars/m/246/246921.jpg
Roger That
#5506584
https://forums.guru3d.com/data/avatars/m/266/266726.jpg
Man thats amazing, intel went from having a 50% performance lead to possibly -25% in a year. 10/10 would design cpu again.
#5506590
data/avatar/default/avatar24.webp
How much the speculation fails? Does it effects SLAT too? (I guess yes since they are gonna to patch all azure clients too.. and this is a good change for EPYC.. toc toc AMD marketing?).
#5506593
https://forums.guru3d.com/data/avatars/m/138/138684.jpg
Ouch, this will have a big impact on performance. Would be nicer at OS level if apps could be set as "Trusted" instead of incurring a performance hit on everything.
#5506595
https://forums.guru3d.com/data/avatars/m/239/239175.jpg
Isn't The Register one of the worst news sites out there? Is this confirmed somewhere else? (The low quality of that site is immediately evident, since they don't actually state which chips are affected. So they claim every Intel chip ever made is affected, which sounds like complete BS.)
#5506602
data/avatar/default/avatar28.webp
RealNC:

Isn't The Register one of the worst news sites out there? Is this confirmed somewhere else? (The low quality of that site is immediately evident, since they don't actually state which chips are affected. So they claim every Intel chip ever made is affected, which sounds like complete BS.)
A paper "KASLR is Dead: Long Live KASLR" talks a lil about performance and architectures involved. At least Skylake CPUs should be affected. However the overhead looks pretty lower than claimed in a host machine, which is good. Not sure how this affected virtualization performance. EDIT: according to "Breaking Kernel Address Space Layout Randomization with Intel TSX" Haswell, Broadwell and Skylake are all affected. edit2: according to "Prefetch Side-Channel Attacks: Bypassing SMAP and Kernel ASLR" Ivy Bridge and Sany Bridge are also affected.
#5506604
data/avatar/default/avatar17.webp
I
RealNC:

Isn't The Register one of the worst news sites out there? Is this confirmed somewhere else? (The low quality of that site is immediately evident, since they don't actually state which chips are affected. So they claim every Intel chip ever made is affected, which sounds like complete BS.)
t is confirmed m8.. from ibm from amd from KASLR... intel messed up big time
#5506608
https://forums.guru3d.com/data/avatars/m/266/266726.jpg
RealNC:

Isn't The Register one of the worst news sites out there? Is this confirmed somewhere else? (The low quality of that site is immediately evident, since they don't actually state which chips are affected. So they claim every Intel chip ever made is affected, which sounds like complete BS.)
Well some shit is going down at the very least, probably under NDA atm, the bread crumbs are real , https://lwn.net/Articles/742404/ , https://lkml.org/lkml/2017/12/27/2, https://www.fool.com/investing/2017/12/19/intels-ceo-just-sold-a-lot-of-stock.aspx
#5506621
https://forums.guru3d.com/data/avatars/m/239/239175.jpg
OK, this looks legit :-/ Not sure if this is bad news for Intel though. Apple had to artificially slow down their devices to make people buy new ones. Intel is getting this for free. :P
#5506636
https://forums.guru3d.com/data/avatars/m/247/247876.jpg
I am skeptical about the claimed performance loss.
#5506638
data/avatar/default/avatar26.webp
If this is legit, Ryzen could be at least equal in many scenarios. Game changing factor, just before Zen refresh launch. A good year for AMD is coming indeed.
#5506654
data/avatar/default/avatar38.webp
mbk1969:

I am skeptical about the claimed performance loss.
The specifics will be benchmarked in short order but in worst case scenarios it looks pretty bad. How common those worst cases are will vet out in the post update benchmarks.
#5506684
https://forums.guru3d.com/data/avatars/m/247/247876.jpg
I think that hype was created not by IT guys but by journalists just to increase their pages` counters.
#5506699
https://forums.guru3d.com/data/avatars/m/239/239175.jpg
Some first tests indicate that this issue will have zero impact on gaming. However, the tests are performed on Linux, and the article is minimal at best; no mention on whether these benches were CPU-bound or not: https://www.phoronix.com/scan.php?page=news_item&px=x86-PTI-Initial-Gaming-Tests It's also not known whether the issue can affect frame pacing. Depending on how much worse a context switch performs now, it stands to reason that this could introduce latency spikes that won't show up in an FPS bench. Maybe Guru3D should prepare some FCAT testing?

Next Page