Intel Kernel Flaw: Will this effect GW2?

For those who don't know: https://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/

The question is, will this effect GW2, and if so how much? Locally, GW2 is known to be CPU heavy, so I am expecting a hit. This is also expected to hit servers HARD, so what kind of issues can we expect from the server side?

It will affect everything, just a question how much. Personally I deactivated Windows Update already. I lived 10 years with this security risk and I am not willing to sacrifice even 1% performance to fix one out of hundreds of security issues.

Yep I turned off all updates for my win 7 box, and killed windows update itself on my win 10 laptop. There's just no way I'm taking a performance hit like that.

Yeah, pretty sure there's no such thing as a "Forcefully Unmap Complete Kernel With Interrupt Trampolines"...

Nothing to worry about for games. However if you're doing something else with your computer, you may feel the difference.

I do think a 5% performance drop is tolerable. A 30%performancedrop however should not and would issue reasons for a recall by intel.I also would advice afgainst turning off windows update. Even if you only play GW2 on your pc. It is a compromise to your security . Why would I care? well it increases the risk of your account being hacked and being used for botting gold, selling gold and/or gold seller spam.If I forget and nobody else does it, please poke me and I'll post the update number and instructions how to uninstall a specific update.

Instructions are (for windows 10):

• Right click start menu
• pick control panel
• go to programs (check if the category view is on)
• click 'view installed updates'
• uninstall the update that causes the performancedrop. (as sad, I'll try and remember to post the update number when it is out there. for now it is not to be found yet.

@Arya.3841 said:Nothing to worry about for games. However if you're doing something else with your computer, you may feel the difference.That is simply wrong, especially for CPU bound games like GW2. These kind of games will take a bigger hit than single player games for sure.The exact impact of this hotfix will depend on many things and even within an application there can be different kind of impacts. For example a story instance in GW2 will feel less impact than a WvW zerg vs zerg battle since during the WvW battle a lot more networking and loading new textures, effects, animations and models will happen.

And considering that some people spend hundreds of dollars for a few extra frames per second (eg. water cooling and fast, overclocked memory), even the 3 frames per second measured in a single player game is a big impact.

Hopefully lawsuits will happen and Intel will be forced to replace all CPUs they sold over the last few years with error free steppings.

@mercury ranique.2170 said:I do think a 5% performance drop is tolerable.

Just no, a 5% hit is huge for enthusiast pc users. They spend hundreds of dollars to get 5% more performance.

Manually uninstalling the update is also just a short term fix. Microsoft will eventually include it in larger patches as always. So blocking all updates will be the only way to be sure until you buy a CPU without these kind of errors. Silly me bought a 8700k recently, wish I knew early (too late to return)

@"mercury ranique.2170" said:I do think a 5% performance drop is tolerable.

BOGGLES You have not been around when I spec out a new machine to buy.

I pointed out this to someone earlier, it will be interesting as about the only major desktop/laptop it does not affect are chromebooks. I can see the adds for things in the pipeline now: "our new 2018 model .... 30% slower than our 2017 model". Yeah, that bird will not fly.

As to the original question though this will affect pretty much everything that runs on top of the operating system. How much depends on many factors but it is safe to say a lot of people will notice it when every machine in their office starts running slower.

Locally, shouldn't be much of an issue. The slowdown occurs at the point when a program asks the kernel to do something like networking or disk IO, so the scare stories about "Upwards of 30% slower!!" are usually referring to benchmarks that repeatedly write small amounts of data to disk, or making networking calls to your local machine (for instance to access a database server you're running). When you're talking about a task that usually takes a nanosecond jumping up to like 100 nanoseconds, then doing a bunch of really fast operations in quick succession adds up to a noticeable performance hit. On the other hand, when you're talking about normal game stuff like communicating with a server or loading a texture off storage, where the time spent actually performing the task is on the order of tens or hundreds of milliseconds, then the performance hit will be fairly minor.

Your local PC will be fine and you won't realize much. I am more worried about servers literally all around the globe and how well their IT departments take care of the risk.I am happy that, as "content creator" and semi-professional filmer, I need many threads, so an Octacore was my intention, which, back then, was too expensive by intel holidng up their monopoly and charge ridiculously high prices for anything above 4 cores. Then, thank god, AMD introduced the Ryzen and that is easily the best CPU upgrade I have experienced so far. I mean, actual speed boost instead of intel's "oh, a new quad core with a slightly different model number - for $500".

Many people brought up that the gap in certain benchmarks comparing Intel and AMD might come from this careless saftey hazard. Since these benchmarks are often totally synthetic and unrealistic, I wonder if certain Intel CPU are now below AMDs counterpart which works like intended... Feels like all cases of doping cases of the USSR's athletes who were "so good!" , those files surfaced after the fall of the iron curtain that basically everyone was doped..

But yeah, I am kinda glad I've a Ryzen 1700X here, staying cool at 60°C at 100% usage for a price I couldn't even afford a hexacore from Intel 1 year ago...But also, let's be honest, I think it's not as bad as it is, even though I like the CEO of Intel sold his stocks because he probably knows what his company has done. Before I had a GTX970 with the faulty memory and it was not even remotely as bad as benchmarks showed and I played fine with that card for 3 years.

If you are happy with the performance of a 1700X of course a few % dont seem like much to you. Intel users pay premium for that extra IPC + higher clock speeds and care about every percent.My 8700k does 5,2 GHz no problem while that ryzen stuff is stuck at max 4ghz and lower IPC.

The whole thing sounds like a bunch of BS to me...and why is this in the Register from the UK and not any tech heavy magazines? Not that I care anyways.Note - After reading a much more thorough article in Business Insider, yes, there's a flaw, but the information from the Register is highly inaccurate...most newer computers will not see any slow down at all, only older systems will see a slow down, and that MIGHT reach 30%, but Intel doesn't agree...also there's another security flaw that affects all CHIP's regardless of who makes it...that means Intel, AMD and ARM, but that one is even harder to fix and harder for to be abused. Bottom line is, don't be stupid(stupid as in those that regularly get their computers infested with viruses and malware) and you'll be fine.

@Zaklex.6308 said:The whole thing sounds like a bunch of BS to me...and why is this in the Register from the UK and not any tech heavy magazines? Not that I care anyways.

There is a vulnerability, Intel has made official statements about it, and have said they are working with AMD to develop solutions to close the vulnerability. It takes almost no time to find that information from reputable sources, such as directly from Intel.

@"Malediktus.9250" said:If you are happy with the performance of a 1700X of course a few % dont seem like much to you. Intel users pay premium for that extra IPC + higher clock speeds and care about every percent.My 8700k does 5,2 GHz no problem while that ryzen stuff is stuck at max 4ghz and lower IPC.

I think in relation to AMD the impact will be far less if anything noticeable to the end user, at least based on what internet is saying on the corner.

I found reports that they are affected and others stating they are not. I found other sites quoting AMD saying that on whole they are not really affected based on a difference in architecture but I have no proof of anything other than this but it could just be covering "PR" for now from AMD until they have more concrete answer. Anyone have any good data on the AMD side that can be verified?

https://cnbc.com/2018/01/03/amd-rebukes-intel-says-flaw-poses-near-zero-risk-to-its-chips.html

! Body is 10 characters too short.

@jbrother.1340 said:

@Malediktus.9250 said:If you are happy with the performance of a 1700X of course a few % dont seem like much to you. Intel users pay premium for that extra IPC + higher clock speeds and care about every percent.My 8700k does 5,2 GHz no problem while that ryzen stuff is stuck at max 4ghz and lower IPC.

Also network performance since GW2 is exchanging a lot of data between the client and the servers. And the server performance itself. So Anet might have to reduce the cap for the amount of players per map slightly.

@Malediktus.9250 said:We will not know how much difference it will makeThis is about all we can say right now. We know that some people have run benchmarks, but we also know that benchmarks aren't very helpful for figuring out what will happen for your experience.

There will definitely be some effect, but whether it's slightly noticeable or very noticeable, whether it also affects cloud services... we'll have to wait.

@Zaklex.6308 said:The whole thing sounds like a bunch of BS to me...and why is this in the Register from the UK and not any tech heavy magazines? Not that I care anyways.

@Sojourner.4621 said:

@Zaklex.6308 said:The whole thing sounds like a bunch of BS to me...and why is this in the Register from the UK and not any tech heavy magazines? Not that I care anyways.

Note - After reading a much more thorough article in Business Insider, yes, there's a flaw, but the information from the Register is highly inaccurate...most newer computers will not see any slow down at all, only older systems will see a slow down, and that MIGHT reach 30%, but Intel doesn't agree...also there's another security flaw that affects all CHIP's regardless of who makes it...that means Intel, AMD and ARM, but that one is even harder to fix and harder for to be abused. Bottom line is, don't be stupid(stupid as in those that regularly get their computers infested with viruses and malware) and you'll be fine.

No slowdown at all is bs, there will be one, just a question of how much. That will vary greatly from application to application and your hardware.

I saw this on reddit: https://www.reddit.com/r/Guildwars2/comments/7o4hl5/fps_drop_after_windows_update/just too bad he did not post anything about his hardware and how much fps he believes to have lost

https://www.reddit.com/r/pcgaming/comments/7o2ctw/benchmarked_intel_security_patch_impact_on/people who say this will have no impact on gaming are most likely wrong. Just 5 games benchmarked there and upto 13% performance loss. Granted it is on a 4 core without hyperthreading, but this is what the average person is probably using.

Games that was running fine just a few days ago is now running horribly... what the hell...

Its probably gonna be after/during the weekend until we get a clearer bigger on it. From what I have seen, gamersnexus (I think it was) tested it on an 8700k/6700k and while there was noticable drops in 4k read benchmarks, games and your average application saw zero difference.

The critical update Microsoft put out this morning, I assume to patch the kernel flaw, hit my 32 bit computer first. It refused to boot past the four pane window symbol after the patch. I restarted and hit the button to interrupt the restart and the computer tried to repair itself and failed. It allowed me to roll back to December 22 and I did so. The computer worked fine after that. I turned off windows update and then turned off Windows update on my 64 bit computer till they fix the problem. Beware

@Malediktus.9250 said:

Hopefully lawsuits will happen and Intel will be forced to replace all CPUs they sold over the last few years with error free steppings.

Yeah, I think you'll find that they never promised anyone any specific performance figures.

Which is good because the cost of offering replacements for all CPUs made since 1995 would either bankrupt the chipmakers (after all Spectre and possibly Meltdown also affect AMD and ARM based systems) or just force them to drastically raise their prices in order to cover the cost of replacements. I don't know about you but I'd like to be able to buy a new CPU when I want one without having to pay a premium in order to cover their losses over this.

Best if everyone just sucks up any performance loss for now. It's not like we have much of a choice anyway. Any fixes will be optimised over time.

@"Ahnog.8795" said:The critical update Microsoft put out this morning, I assume to patch the kernel flaw, hit my 32 bit computer first. It refused to boot past the four pane window symbol after the patch. I restarted and hit the button to interrupt the restart and the computer tried to repair itself and failed. It allowed me to roll back to December 22 and I did so. The computer worked fine after that. I turned off windows update and then turned off Windows update on my 64 bit computer till they fix the problem. Beware

There is a known issue where some AV software blocks some parts of the patch causing BSODs and other issues, perhaps it has something to do with your errors?

http://www.zdnet.com/article/windows-meltdown-spectre-fix-how-to-check-if-your-av-is-blocking-microsoft-patch/

@Zaklex.6308 said:

@Zaklex.6308 said:The whole thing sounds like a bunch of BS to me...and why is this in the Register from the UK and not any tech heavy magazines? Not that I care anyways.

@Sojourner.4621 said:

@Zaklex.6308 said:The whole thing sounds like a bunch of BS to me...and why is this in the Register from the UK and not any tech heavy magazines? Not that I care anyways.

Yeah, I like the register but there are some issues with their article and the whole thing has a hint of a "the sky is falling" article (mind you, I've read worse and more alarming articles over this too). The important thing to remember is there is no known exploit at the moment; the "bad guys" who would write those exploits are just finding out now. The "good guys" have been working on fixes for the past few months. My money's on the "good guys".

It's important at this time to, as the British say, Keep Calm and Carry On.

@"Malediktus.9250" said:No slowdown at all is bs, there will be one, just a question of how much. That will vary greatly from application to application and your hardware.

I saw this on reddit: https://www.reddit.com/r/Guildwars2/comments/7o4hl5/fps_drop_after_windows_update/just too bad he did not post anything about his hardware and how much fps he believes to have lost

https://www.reddit.com/r/pcgaming/comments/7o2ctw/benchmarked_intel_security_patch_impact_on/people who say this will have no impact on gaming are most likely wrong. Just 5 games benchmarked there and upto 13% performance loss. Granted it is on a 4 core without hyperthreading, but this is what the average person is probably using.

You will notice a difference of about 1-3% while playing games, the software patch relates to how sys calls are handled wich most games do little to none of, server side however will notice performance hits upto the 30% wich we are already seeing.