Intel users: prepare to lose 5-30% of performance

[dandyp]

I've updated 2 Ivy bridge chips, and not really noticed much (if any) real-world performance impact. The hardest part was having to wait, and chase-down the motherboard BIOS updates. Both companies (Gigabyte and Asrock), had the new BIOS, but still haven't uploaded them to their websites. It seems that it's very much a case of ask and you shall receive.

[julian98]

how long did it take to get back to you, i have an asus on ivy bridge and thought they weren't being updated?

[Millennium]

I used to work for MSI and if you contact the slightly (At the time) hilarious support function you can indeed get beta bios'

Anyone wanna let me know if there's a BIOS for Ryzen 1700's from MSI? Doesn't appear in the readme.txt on their website that they've implemented anything relevant in a march update yet.

[shaithis]

I've updated 2 Ivy bridge chips, and not really noticed much (if any) real-world performance impact. The hardest part was having to wait, and chase-down the motherboard BIOS updates. Both companies (Gigabyte and Asrock), had the new BIOS, but still haven't uploaded them to their websites. It seems that it's very much a case of ask and you shall receive.

Didn't think IVB had new microcode for them yet? (or ever..)

[CAT-THE-FIFTH]

Didn't think IVB had new microcode for them yet? (or ever..)

Forget BIOS level updates,just the software ones,had an impact in my modded FO4 playthrough in any areas where there is lots of assets which needed to be loaded,and there is more stuttering,and even with newer CPUs,DF found the biggest drops in performance in open world games like The Witcher 3.

This is where the patches hurt performance. For other games I didn't notice much of a performance drop.

[ik9000]

Didn't think IVB had new microcode for them yet? (or ever..)

everything back to lynnfield/clarkdale (s1156) is patched by intel in microcode: https://newsroom.intel.com/wp-conten...e-guidance.pdf

beyond that, nope. The odd one being the way they ditched nehalem entirely, though that could be because it isn't vulnerable - but they haven't clarified.

[kalniel]

beyond that, nope. The odd one being the way they ditched nehalem entirely, though that could be because it isn't vulnerable - but they haven't clarified.

Tools suggest it is. I think it's back to the fact it doesn't support cheapish mitigation that you can use on later processors, so the cost is high.

[ik9000]

Tools suggest it is. I think it's back to the fact it doesn't support cheapish mitigation that you can use on later processors, so the cost is high.

but isn't inspectre just checking against a processor ID code list rather than testing for the exploit itself?