Intel's Chip Bug

[Evildeffy]

so if AMD are such good guys, why are they allowing the release of a Kaby Lake based Meltdown affected product featuring their own tech?

Not exactly sure how a Graphics product has anything to do with Intel's issue.
AMD is simply selling their GPUs to Intel for them to use in their CPUs, it's not AMD's issue as their GPU isn't affected by the meltdown bug.

Also not exactly sure how that is related to being "a good guy"?

[Artorius]

so if AMD are such good guys, why are they allowing the release of a Kaby Lake based Meltdown affected product featuring their own tech?

Because they aren't. They're a company like other companies, and companies generally do what is best for them. AMD selling GPUs to Intel is great for them.

[Demona3]

What OS do you run and what do you use your PC for? There is a very good chance this won't really impact you.

Winblows 10 and mainly gaming. I also use it for drawing.

[schwarzkopf]

Thought I'd post bullshit and random rubbish and use colours cleverly like a child.

Not sure why you'd want to do that.

[pansertjald]

... AMD is cheaper. Not sure you were making the point you wanted to make. Not that there is anything wrong with Ryzen.

- - - Updated - - -



Child, youre so deluded its painful. I posted DOZENS of links to testers everywhere. You were wrong. Youre still wrong. I stopped responding because i categoricaly proved you wrong a dozen times and you never once posted anything to refute it. You had to go back to a six year old benchmark with shitty testing methodology (didnt even let the AIO run long enough to heat soak and reach equilibrium)

Tildeer: AIOs arent any better than good air coolers, even on Threadripper. Every modern test done by every reputable tech site shows it. Welcome to reality. (Next answer prediction: “you cant trust any of those sites, they are all paid shills”.)

What a load of BS. You say you posted a dozen links. LOL you posted 2 links and both links are cooling the Threadripper wich again are being cooled by AIO's that doesn't support the TR4, but the air coolers used in the test do. Even Nexus points this out and say's that if the AIO would cover the whole CPU it would cool better.

So little Child i think it's time for you to just learn to accept when you are wrong.

From the Nexus link you posted "Liquid is still better when the plate contacts the whole die, as you’d expect, but the smaller contact area coldplates of the Asetek CLCs (represented by the 280mm X62) can’t keep up. That contact area is key. As more CLCs break into the TR4 market with properly sized coldplates, the AIO options should expand beyond Enermax and open loops."

[Vash The Stampede]

so if AMD are such good guys, why are they allowing the release of a Kaby Lake based Meltdown affected product featuring their own tech?

Because this is AMD's graphics tech. The CPU is still Intel's. Intel has done a lot to make sure you never see the AMD label, just the Radeon graphics. Essentially AMD is selling them a part, not building their CPU.

- - - Updated - - -

Felt like retorting that exactly as you have, without colors.

Cause I see a lot of idiots putting out false information to make Intel look good. If anyone needs to be brought down from their pedestal, it's Intel. Not only does AMD's Ryzen run faster at productivity and runs cooler, but is now the most secure processor. AMD is the only CPU manufacturer who is ONLY effected by one of the Spectre vulnerabilities. ARM chips are effected by two, and Intel is effected by three.

So not only does AMD have a lower TDP and was already faster at productivity, but is now ~20% more faster thanks to Intels fix. Intel needs to take a hit, and they deserve it. It's only good for us consumers. It's not like Intel won't recover.

[Sturmbringe]

Intel needs to take a hit, and they deserve it. It's only good for us consumers. It's not like Intel won't recover.

TBH I couldn't agree more. I would also love to see NVIDIA brought down a dozen notches or so.

Hopefully, this story with Meltdown/Spectre will help AMD CPU Dept to survive and thrive.

[Zenny]

Winblows 10 and mainly gaming. I also use it for drawing.

Nah, you don't really need to worry about performance. Gaming and drawing should have very little difference, -2% at most.

[Demona3]

Nah, you don't really need to worry about performance. Gaming and drawing should have very little difference, -2% at most.

Good to know, thank you. Honestly, I think they should be replacing all the chips they ever sold, but that would cause them to go bankrupt.

[Vash The Stampede]

Someone on Reddit did some tests and it seems Haswell is slower in some games. Anywhere from 1.5% to 8% loss in performance. Newer Intel CPUs seem to have much less of a performance hit from all this.

https://www.reddit.com/r/pcgaming/co...tch_impact_on/

[mrgreenthump]

Someone on Reddit did some tests and it seems Haswell is slower in some games. Anywhere from 1.5% to 8% loss in performance. Newer Intel CPUs seem to have much less of a performance hit from all this.

https://www.reddit.com/r/pcgaming/co...tch_impact_on/

AFAIK, these are missing the microcode update. At least he is specifically mentioning Windows versions and saying nothing about BIOS updates.

[Sorshen]

Asus is only providing microcode updates for mainboards from the 200, 300, X99 and X299 series so far.

[Deleted]

I feel like there must have been something going on behind the scenes before it became public as Cisco was already sending out FE's to replace CPU's weeks ago.

Ofc there was something going on behind the scenes. But making changes to CPU design is no small feat - it takes A LOT of time to do so. If it is marginal change it supposedly can take anywhere from 6 months to a year. If it is anything but marginal you are looking at at least a year - info from some dude who worked in intels cpu department for years, cba to look for his tweeter blog about it now. So seeing how they knew about the bug in June the earliest we could hope to see fixes is zen2(but probably not since it was supposedly "taped out" recently) but more likely zen3. Same deal for intel they are unlikely to make it in time with changes for next cpu arch so as I said 2019 probably later.

[Gray_Matter]

Ofc there was something going on behind the scenes. But making changes to CPU design is no small feat - it takes A LOT of time to do so. If it is marginal change it supposedly can take anywhere from 6 months to a year. If it is anything but marginal you are looking at at least a year - info from some dude who worked in intels cpu department for years, cba to look for his tweeter blog about it now. So seeing how they knew about the bug in June the earliest we could hope to see fixes is zen2(but probably not since it was supposedly "taped out" recently) but more likely zen3. Same deal for intel they are unlikely to make it in time with changes for next cpu arch so as I said 2019 probably later.

AMD have a microcode fix out for one of the Spectre issues so that would be included with Zen+. Not sure what's happening with the other one.

[thunterman]

What? Meltdown can be "ignored entirely"? Not everyone plays games. We have 250 processing VM's and I am not even talking about our development machines. Even a 10% drop will necessitate us getting more servers or cutting the number of processing VM's. People who bought Skylake systems for processing could see noticeable drops in some activities. Time will tell how badly the effects will be but it's definitely not something that can be ignored.

Just to prove my point. This is from Epic Games and their servers:

https://www.epicgames.com/fortnite/f...ability-update

I understand the server issues, but talking to 95% of consumers that just play games, browse the web e.t.c they have no reason to be concerned, let alone try and start court cases. The whole thing is ludicrous and whichever outlet it was that leaked it should be ashamed of themselves. It undermines the point of Project Zero.

It's perfectly safe to just not update for the protection, as to actually utilise spectre and meltdown, you still have to infect the target. It's not magic, it's not like you can type an IP address and "meltdown lel" bingo I got your bank details. You have to be an idiot to be affected by these vulnerabilities in the first place, and those people aren't going to even understand what a 0-30% performance loss even means, because they are sat on bloody Facebook. Everyone is loosing their shit over almost nothing. These vulnerabilities aren't even nearly as dangerous for users as the multiple, easy to utilise WPA vulnerabilities, that effect pretty much every home with WiFi across the globe, but hey, there isn't an article saying it might reduce performance slightly about those so noone cares.

Just be more cautious of what your downloading. Give it a few weeks and there will be more than a band-aid fix, that won't gimp performance anyway

[Remilia]

I understand the server issues, but talking to 95% of consumers that just play games, browse the web e.t.c they have no reason to be concerned, let alone try and start court cases. The whole thing is ludicrous and whichever outlet it was that leaked it should be ashamed of themselves. It undermines the point of Project Zero.

Well, considering that it was discerned at first from Linux kernal updates and then later Intel jumped the gun in their press briefing causing Microsoft fast forward their patching because of the press caused by Intel's PR blunder. So... you really either have to blame Linux people doing their job or Intel for jumping the gun on a press release which was supposed to coincide with Microsoft's security patch.

Meltdown's performance degradation doesn't affect us directly but it does definitely impact us indirectly.
Epic's Fortnite server for example showed CPU utilization sky rocket by 30%, which can definitely lead to server congestion. Goes without saying, it can cause a lot of issues if the servers don't have the hardware. Just imagine the worst game server launch ever, and make it 30% worse.

It's perfectly safe to just not update for the protection, as to actually utilise spectre and meltdown, you still have to infect the target. It's not magic, it's not like you can type an IP address and "meltdown lel" bingo I got your bank details. You have to be an idiot to be affected by these vulnerabilities in the first place, and those people aren't going to even understand what a 0-30% performance loss even means, because they are sat on bloody Facebook. Everyone is loosing their shit over almost nothing. These vulnerabilities aren't even nearly as dangerous for users as the multiple, easy to utilise WPA vulnerabilities, that effect pretty much every home with WiFi across the globe, but hey, there isn't an article saying it might reduce performance slightly about those so noone cares.

Just be more cautious of what your downloading. Give it a few weeks and there will be more than a band-aid fix, that won't gimp performance anyway

I'm not sure how you can see a security flaw, and then proceed to say it's perfectly not safe to protect yourself against it. Local execution can just be done by javascript, so rogue ads or compromised sites. So it's not like they have to install something to make it work. And saying just use noscript or what not isn't a solution, that's an actual bandaid fix and many people do not even know those kinds of addons exist.

[thunterman]

Well, considering that it was discerned at first from Linux kernal updates and then later Intel jumped the gun in their press briefing causing Microsoft fast forward their patching because of the press caused by Intel's PR blunder. So... you really either have to blame Linux people doing their job or Intel for jumping the gun on a press release which was supposed to coincide with Microsoft's security patch.

Meltdown's performance degradation doesn't affect us directly but it does definitely impact us indirectly.
Epic's Fortnite server for example showed CPU utilization sky rocket by 30%, which can definitely lead to server congestion. Goes without saying, it can cause a lot of issues if the servers don't have the hardware. Just imagine the worst game server launch ever, and make it 30% worse.

I'm not sure how you can see a security flaw, and then proceed to say it's perfectly not safe to protect yourself against it. Local execution can just be done by javascript, so rogue ads or compromised sites. So it's not like they have to install something to make it work. And saying just use noscript or what not isn't a solution, that's an actual bandaid fix and many people do not even know those kinds of addons exist.

A website published an article explaining the exploits prior to the embargo lift. That is what caused the shitstorm.

further patches will solve the problems, everyone is threatening legal action over something that will be solved in the coming weeks. This sort of thing happens on a monthly basis but noone really knows the details because they don't care enough to go and find them. This has been plastered all over the web and noticed, that's the only difference between this and the other 1000 exploits that exist. There are many that are far worse than spectre / meltdown.

[Remilia]

A website published an article explaining the exploits prior to the embargo lift. That is what caused the shitstorm.

Which they got from a Linux update. They are not under embargo obligations.

further patches will solve the problems, everyone is threatening legal action over something that will be solved in the coming weeks. This sort of thing happens on a monthly basis but noone really knows the details because they don't care enough to go and find them. This has been plastered all over the web and noticed, that's the only difference between this and the other 1000 exploits that exist. There are many that are far worse than spectre / meltdown.

Care to list them? Cause I don't know of any bug that to truly fix would involve regressing CPU performance two decades. And if you don't know why, it's because it's an inherent problem with OoOE and Speculative Execution.

[Wries]

If this needs a BIOS update.. how many average end-consumers will succeed with that? Now changes to CPU microcode can be loaded on OS level and OS kernel updates can also unsurprisingly be brought in by an update to the OS. So what are they supplying in these BIOS updates?

Rolled back to Windows 8.1, disabled updates, applying all ugly band-aid fixes in existence and am eagerly awaiting my credit card number to get stolen. Bring it!! Also I know this isn't sane, reasonable or responsible, but I'm not either so it works out.

[thunterman]

Which they got from a Linux update. They are not under embargo obligations.

Care to list them? Cause I don't know of any bug that to truly fix would involve regressing CPU performance two decades. And if you don't know why, it's because it's an inherent problem with OoOE and Speculative Execution.

The patches are changing SE to not cache memory, that is where the performance hit is coming from. Theoretically it should be possible to add a encryption layer and re-enable SE.

As I mentioned further up the thread, there are many WPA2 exploits still in existence, that are easy to exploit (which meltdown / spectre is not) In the grand scheme of things, this is not the worst active exploit in computing, but it is the most publicised. I'm more worried that pretty much anyone with a phone can pluck data out the air at will while I'm using public wifi, or even my home network. As long as you are using noscript and adblockers, which most do, it is going to be borderline impossible for anyone to even use the Spectre exploit against you, as in this case, there is only one way to utilise it, which would involve you downloading a virus of some form.

It's a problem for novice users, and they should be applying the patches but at the same time, novice users are going to be wholly unaffected by the performance loss anyway.