Windows 10 Meltdown/Spectre Patch Performance Impact Assessment

"we must preface this with the fact that AMD systems will not be affected in the same way as an Intel-based system by this update, because AMD CPU architectures are vulnerable to Meltdown".
I think you mean invulnerable, right?
Nice to see the performance hit for the average user is minimal.
 
Just going to a pic dude.

"we must preface this with the fact that AMD systems will not be affected in the same way as an Intel-based system by this update, because AMD CPU architectures are vulnerable to Meltdown".
I think you mean invulnerable, right?
Nice to see the performance hit for the average user is minimal.

Fixed and fixed.

Yeah, I am very happy to know that the performance hit is minimal for most standard users. Only slightly annoying that so much time was spend testing it.
 
Being new here, Hello first of all. Plus thank you for taking the time and effort with the testing Thank you again, regards
 
:DWelcome donp.:D

My main gaming rig (i9-7900X) is so much overkill that I'm sure I will not see any impact on my gaming experience.

The i7-7700K and the i7-6600K are pretty much the same.
I applied Microsoft's fix to them all and I can't tell from normal use that there is a performance hit.

Thanks for posting this analysis. It's good to know about it, even if we can't ~do~ anything about it. :eek:
 
The Question

The question we all should be asking is really how long has intel known about this issue an sat on there ass doing jack about it... If google can find it so why did a highly known company like intel miss this issue or did they ?

I only say this as we all known company's hide so much so they can sell there item's. Intel replied that the processor's in question are all working as they were meant to but you guy left a security hole that big enough to fit my mother in law's ass in an that's one big ass trust me.lol
 
Nice to see a set of results for a chip with PCID (which intel said would reduce the impact) - any chance of testing a pre-Haswell chip? (for us ivy-bridge holdouts)
 
The question we all should be asking is really how long has intel known about this issue an sat on there ass doing jack about it... If google can find it so why did a highly known company like intel miss this issue or did they ?

I only say this as we all known company's hide so much so they can sell there item's. Intel replied that the processor's in question are all working as they were meant to but you guy left a security hole that big enough to fit my mother in law's ass in an that's one big ass trust me.lol

If you have the technical and brilliant brain it requires to create a CPU architecture you know all of its flaws and etc.

I can't do any testing right now but I am going to test M.2 "SATA" and SATA SSDs.

Mark, is it possible you could do that on a regular SSD etc just to put my mind at rest? all you really need to run is ATTO and Crystal tbh, and maybe AIDA.
 
Quick update. I plan to test the impact of security fixing BIOS updates in the future.

Sadly a BIOS update has not been released for my Intel/ASUS Strix X99 motherboard yet, so it may take a while before testing can begin.
 
Why would it affect the SSD?
It affects IO performance so moving data around and especially benchmarks take more CPU oomph to run.

It can be significant on servers with large raid arrays, but to notice anything on a desktop takes one hell of a setup.

So far tech websites have only showed throughput numbers, but recording the CPU usage pre- and postpatch would be interesting.
 
Why would it affect the SSD?

From what I have seen it does not affect SATA SSDs, yet I don't think it has been tested.

https://www.youtube.com/watch?v=JbhKUjPRk5Q&t=14s

It definitely affects NVME/uber fast SSDs.

If you think back to what I posted from Gareth Halfacree when he was explaining what the fix would do to Linux (IE imagine a bike that crashes when you turn left etc) then it seems that NVME is rife to be hit by that logic. IE - if you are re-routing a crap ton of data and instruction then something is going to suffer. And it seems NVME really suffers. Up to 42%, and usually around at least 20%.
 
Back
Top