Discussion in 'Article Discussion' started by bit-tech, 15 May 2019.
finally game to a head early?
Ah, the G key is so close to the C key. Well, it's not really, but... Look, the Goodyear Blimp!
<runs and hides>
Intel just love racking up those vulnerabilities.
How the heck did the 4690/K dodge that one...?
Okay, so it that i haven't read the entire source material and have very little time to invest in doing so (along with little technical knowledge in this arena) I wanted to ask a question.
In respect to finding flaws in products more than 10 years old and the items were manufacturered with no prior knowledge of said flaws, how much emphasis can we put on this?
I guess my question is, how long after production should Intel (and other manufacturers) be responsible for their products performance and security when they operate in such a fluid and changing technology industry?
Props to someone in the TPU comment section for this gem:
The i9 9900k is also affected...
Apologies if my article didn't make it clear, but the flaws are found in processors from 2008 to the present day. Not prior to 2008, 2008 upwards. So while you might not expect Intel to support an Atom chip it stopped making a decade ago...
There's also that. And pretty much every other chip it's made in the last eleven years.
Basically, do you have an Intel chip made after 2007? Then you're probably vulnerable.
No Hyper-Threading. Can't steal data from a thread running on the same physical CPU core if you can't run on the same physical CPU core...
... or should that be "Has no-one with a 4690/4690K found or been affected by this?"
You get 10 year limit for faults with cars why should a CPU be any different, I guess it is difficult for Intel as it would require computer manufacturers to provide BIOS updates, unlikely to happen as many companies don't exist long enough in the tech world.
Leaves me and my Arrandale in use right this second out in the cold, hopefully software patches will mitigate and I don't have to concern myself.
...the 4690[K] doesn't have hyperthreading enabled, the compenent vulnerable to these exploits?
Any word of AMD SMT being affected by these or similar?
IIRC so far only Intel's CPUs have been demonstrated to be vulnerable these specific exploits. Doesn't necessarily mean AMD's SMT isn't vulnerable to similar exploits, just not these specific ones.
So far, it's looking Intel-specific - they're primarily variants on the Meltdown theme. The old vulnerability checker script I'm using has been updated to check for it and claims my AMD system is vulnerable, but I think it's only looking for the mitigations and not finding them - actual exploitation should, in theory, fail.
EDIT: Ninja'd by @RedFlames!
The only 2 CPU's in that (sub-)family w/o HT - what are the chances...?
... still ran WU, as best practice...
Odd that the i5-4670/K is/are on the list, but that doesn't have HT either; maybe I'm overthinking this, but it's still odd.
It could also be a simple oversight... Intel make/made a ****load of various models and it's easy to lose one or 2 in what will be a lengthy list.
... fair. Glad my rigs' next iteration will be Ryzen, itc.
FWIW the 'Products formely Haswell' list on ARK is 256 items long... and that's just Haswell... not including Haswell-E or anything of that ilk.
Separate names with a comma.