I relize this is the wrong Thread to ask but same topic....Does Amd have any cpu vulnerabilities of there own or this is only for Intel? Reason I ask is I will be upgrading soon and its good to know these things. Thx Guys. Nz

nz3777:

I relize this is the wrong Thread to ask but same topic....Does Amd have any cpu vulnerabilities of there own or this is only for Intel? Reason I ask is I will be upgrading soon and its good to know these things. Thx Guys. Nz

AFAIK, this is only for Intel processors.

TieSKey:

Home Alone movie, the kid uses that scene from a movie on tv to freak out the hotel staff. Not sure if the other movie was a real one or not.

Merry Christmas you filthy animal. And a happy new year. Probably the best scenes from those movies.

The gift that keeps on giving!

This is not good for overlcockers and those who undervolt their laptops via Intel Extreme Tweaking Utility to improve battery life. Lucky for me only 1 out of my 3 PCs are effected which is my gaming laptop. But I don' use SGX or have it enabled so I have nothing to worry about.

blkspade:

The thing about that is the independent security analyst have generally been testing AMD CPUs as well. Beyond the initial wave that effected even ARM CPUs, AMD has seemingly been in the clear.

Ahh yes, only would make me sleap easier if I had an AMD CPU (yet).

nz3777:

I relize this is the wrong Thread to ask but same topic....Does Amd have any cpu vulnerabilities of there own or this is only for Intel? Reason I ask is I will be upgrading soon and its good to know these things. Thx Guys. Nz

On the paper they state this affects the last 5 generations of Intel Core processors only. ARM TrustZone similar fault discovered in the 2017 on certain ARM SoCs is not applicable to Intel SGX and vice-versa (and the Intel version is the more serious). AMD SEV looks like is not affected at all to this issue.

angelgraves13:

Intel Inside will be just as frightful as the radiation symbol in a hundred years.

Apparently those Intel Inside processors last for millions of years... 😱

http://i.imgur.com/fSBtASk.jpg

angelgraves13:

Intel Inside will be just as frightful as the radiation symbol in a hundred years. We might need to figure out where we’re going to stockpile them. Never forget about Skynet...

Well, their new depth camera may fulfill that prediction of yours. Good for logistics? Good for DARPA.

nz3777:

I relize this is the wrong Thread to ask but same topic....Does Amd have any cpu vulnerabilities of there own or this is only for Intel? Reason I ask is I will be upgrading soon and its good to know these things. Thx Guys. Nz

Yes, AMD has vulnerabilities. Just not this one.

angelgraves13:

You can only destroy them in Mount Doom. They can take a lot of heat lol

You meant At Mount Doom :p

If they truly take away voltage controls, it will be the first BIOS update I am not installing. I'm relatively OK with Meltdown&Spectre, at least on the 8700K I have, their impact is 100% negligible, outside some NVMe SSD benchmarks. But voltage controls, no, that I actually need. I am using my delidded CPU undervolted, it makes for a fully quiet system while I do most of my work, browse and read the web, with temps barely jumping in the 40-50s under load and no fans spinning, case, CPU, GPU, even PSU - all 0 rpm at idle. I like the PC like that. Undervolting is an important part. Not to mention it would also kill overclocking too, which would lead to riots and refunds for people with X and K chips. Oh, and pretty much nobody uses SGX? I have never installed drivers for it and no app asked for it to be enabled. So weird and tiring, all of this is. For those that don't know where the locked voltage thing is coming from, it's Intel. https://i.imgur.com/Yf1Y0sj.png

toyo:

Oh, and pretty much nobody uses SGX? I have never installed drivers for it and no app asked for it to be enabled.

any application capable of reading uhd blurays uses the sgx unit during the drm process.

Intel releasing an update, that stops overclocking. [youtube=l4yA87VmyhQ] :p

Astyanax:

any application capable of reading uhd blurays uses the sgx unit during the drm process.

What a failure... How do you manage to exclude one of the 2 big CPU manufacturers in one small stupid step, just to get some "better" DRM? I mean, Bluray is already niche and in general disks are kinda going away, in my main PC I didn't install a DVD RW for 4-5 years now, although I do keep one handy, just in case. In the MSI UEFI SGX seems to default to Software Controlled, from my understanding that means apps can require its enabling if needed, with a reboot. We'll see when the new BIOS gets released. Intel must know what sort of reaction they will get if they mess with voltage controls, and it's been quite the bad 2 years for them, I don't think they want to make it even worse. https://i.imgur.com/LGHCJSX.png

This is a funny story about SGX from back in February. https://arstechnica.com/gadgets/2019/02/researchers-use-intel-sgx-to-put-malware-beyond-the-reach-of-antivirus-software/

1 month later, bumping this. MSI has released a new BIOS for Z370, with new microcode (specifically what AIDA64 detects as CAh). It does seem to be the new microcode for fixing Plundervolt. I am mostly writing this to recommend against updating to the series of upcoming BIOSes containing this microcode, although your mileage may obviously vary. On my PC with a 8700K, this BIOS is quite horrible (first bad one I've seen in more than 2 years). Some symptoms: - uncore/ring multi was reduced from 44 to 43 on default settings (this might be MSI's fault alone) - Cinebench r15 scores dropped from 1455-ish to 1408, well outside the margin of error - the voltage settings were messed up. I would use the CPU with a -0.085V Adaptive+Offset undervolt, which would result in something like a 1.11V vcore under full load with p95/SmallFFTs and 118W package power. After the new BIOS/microcode, the same setting is now running at 1.16-1.17V and 127W. I've had to go as far as -0.125V to achieve the same voltage as with previous BIOS/microcode. I assume this is part of the Plundervolt fix. Luckily it seems stable for now, just 1hr in stresstesting tho' - the VRMs are 5-10C hotter for some reason (at similar voltage/frequency/power draw). Don't know why. Only tested for now at the default clocks. Rather annoyed with the loss of Cinebench score and lowered ring/uncore muli. Will see about OC later.