Yet Another Speculative Malfunction: Intel Reveals New Side-Channel Attack, Advises Disabling Hyper-Threading Below 8th, 9th Gen CPUs

[Manu_PT]

Glad I jumped straight from an i5 4690 to a 9700k wich doesnt have HT.

 

[windwhirl]

"Ah sh*t, here we go again" was my thought when I read the title...

 

[XiGMAKiD]

...significant software changes will be needed to harden systems against this exploit, not only from themselves, but from operating system designers and third party app creators

Buggy updates incoming

 

[Steevo]

You can disable cpu caching and etc.. in bios.

you could also take a hacksaw to the chip because that'll essentially destroy your performance

 

[Caring1]

I disabled HT ages ago.
I doubt I am a priority target anyway.

 

[Totally]

You sir deserve a medal! Yes yes and more yes! Finally, someone with some common sense to hopefully enlighten the masses. Listen to this person! You'll notice I didn't amend anything - it was written perfectly =)

I thought similarly last time until some uni student demonstrated a remote attack via a compromised system on the same network. So yes you can point out that there is an impassable canyon in between that possibility and reality while ignoring the guy building a bridge to the left.

 

[Emu]

While it's still possible to actually do this securely, the pitfalls of SMT will only increase with architectural complexity, and the cost of dealing with this does too, and since the performance gains from SMT are diminishing with increasing IPC, SMT should be abandoned sooner rather than later.

Cinebench would like to disagree with your "performance gains from SMT are dimishing with increasing IPC". My Ryzen 7 2700x scores 3582 in Cinebench Release 20 with SMT enabled and 2074 with SMT disabled. I actually had to run the 8 core / 16 thread benchmark twice because I was only expecting a 30% difference. IPC increases between generations of architectures is usually in the single digit range and increased width in the instruction pipeline benefits SMT performance.

 

[xkm1948]

At this rate i may have to switch to Zen 2 TR sooner than I would like...

 

[Vulpesveritas]

That test in the article is one.



Citation?

The official webpage by the guys who discovered the exploit to begin with, which I linked earlier in the thread. https://mdsattacks.com/

 

[LAN_deRf_HA]

I've been confused by the "newest gen is ok" vs "newest gen is more vulnerable" comments. Is it that the newer chips are more vulnerable just specifically not to the hyper thread exploit?

 

[Vulpesveritas]

I've been confused by the "newest gen is ok" vs "newest gen is more vulnerable" comments. Is it that the newer chips are more vulnerable just specifically not to the hyper thread exploit?

The report from the lab testers say the newest chips are more vulnerable, while Intel is claiming they're not, is what it appears.

 

[Mistral]

Yes, I'm sure all those people that bought i7s are simply delighted that someone is advising them to disable HT... And the "minimal" performance penalty of up to 9% sounds real nice compared to the huge 5% gaming performance advantage Intel enjoys over AMD.

 

[trparky]

I am really starting to regret buying my 8700K right about now. These Intel chips are turning out to have more security holes than Internet Explorer.

 

[R-T-B]

The official webpage by the guys who discovered the exploit to begin with, which I linked earlier in the thread. https://mdsattacks.com/

Must've missed it, thanks.

Now that I can reference the page and tool, it seems interesting to me the tools claims 9th gen is "not affected" by meltdown at all. Considering that was the biggest performance impacting fix from the previous batch of vulnerabilities, it's almost looking like 9th gen may have taken 2 steps forwards, only to fall 2 steps backwards.

 

[Nihilus]

So the once mighty 7700k is basically an 8350k now meaning my $170 at launch 2400g will beat it now in many games.

 

[Cybrnook2002]

And they keep on selling them anyways......... waiting for the next nda to expire so even the next vulnerability goes public. Something to be said about selling knowingly faulty chips under the clock of waiting for nda to lift.

 

[Caring1]

Is Intel going to use this as an excuse to delay the launch of Gen 10?
My bet is on a revision of 9 first.

 

[SIGSEGV]

Ryzen 5 3500U here i come..

 

[Prima.Vera]

At least those are public and not kept secret anymore. I doubt those affect the average Joe user using the PC for mundane tasks...

 

[trparky]

At least those are public and not kept secret anymore. I doubt those affect the average Joe user using the PC for mundane tasks...

But the question that I have is... How many more skeletons does Intel have in their closets? How many more exploits are there that are just waiting to be found?

In some ways, I don't want to know.

 

[TheLostSwede]

AMD seems to at least be partially affected. No patches applied yet on Windows 10.
Edit: Updating Windows today made no difference the list of vulnerabilities...

 

[randomUser]

They are releasing these news now, because they want people with gen 7 and below to go and buy gen8/9 because these gens can work with HT enabled and be safe.

So it the upgare of the cpu not for better performance, but for better protection.

$$$

 

[adulaamin]

They are releasing these news now, because they want people with gen 7 and below to go and buy gen8/9 because these gens can work with HT enabled and be safe.

So it the upgare of the cpu not for better performance, but for better protection.

$$$

With Ryzen 2 coming, most, I think, would upgrade to AMD rather than Intel's 8th oe 9th gen cpus.

 

[craigo]

CONSUMERPOWERUP!

OMG! an security flaw has been published. FETCH MY WALLET, REPLACE ALL THE THINGS!

 

[Assimilator]

New speculative execution bug leaks data from Intel chips’ internal buffers

Intel-specific vulnerability was found by researchers both inside and outside the company.

arstechnica.com

Today a microcode update for Sandy Bridge through first-generation Coffee Lake and Whiskey Lake chips will ship.

...

For systems dependent on microcode fixes, Intel says that the performance overhead will typically be under three percent but, under certain unfavorable workloads, could be somewhat higher.

This isn't sounding nearly as bad as Spectre/Meltdown, either in the ability to exploit or in the performance impact of mitigations.

Rävenlord, someone needs to profo raed yuor hihgspede typnig. Pellscheck?


View attachment 122988

I've long given up on expecting basic editorial standards from TPU.