Tech myth debunk thread

[Boinbo]

I don't know if we have something like this, but I'd like to start one. What are some common PC myths that many people believe to be true, but aren't? Please post below.

Edited June 28, 2019 by Boinbo

[Dedayog]

8 minutes ago, Boinbo said:

I don't know if we have something like this, but I'd like to start one. What are some common PC myths that many people believe to be true, but aren't? Please post below. I'll give one:

Air cooling is usually better than liquid cooling in almost every way. 

Is that a myth you're posting or the debunk?

[SenKa]

10 minutes ago, Boinbo said:

Air cooling is usually better than liquid cooling in almost every way. 

In terms of cooling capacity CLC's only now are reaching the capabilities of large air coolers, so that's not a myth. Custom loops do often perform better when using a 360MM radiator or more, but the cost is prohibitive to most people (myself included).

 

A good waterblock costs more than my Dark Rock Pro 4, and I simply don't have $500 to spend on my cooling.

[Boinbo]

1 minute ago, jstudrawa said:

Is that a myth you're posting or the debunk?

Debunking the myth water-cooling is better, should have posted that first

[Dedayog]

5 minutes ago, Boinbo said:

Debunking the myth water-cooling is better, should have posted that first

It's not a myth, nor is it a debunk.  Some AIO's perform better than Air Coolers, while most do not, true.  It depends on the AIO and the Air Cooler.

 

Just because you asked this earlier and got some answers doesn't mean it's a definitive debunk.

[Boinbo]

Just now, jstudrawa said:

It's not a myth, nor is it a debunk.  Some AIO's perform better than Air Coolers, while most do not, true.  It depends on the AIO and the Air Cooler.

 

Just because you asked this earlier and got some answers doesn't mean it's a definitive debunk.

Kay.

[Arika]

this thread is a mess already...

[Boinbo]

3 minutes ago, Arika S said:

this thread is a mess already...

Agree

[Ravendarat]

There is no way this doesnt just dive into becoming a big fight. Debunking myths are only interesting when the myth is widely considered to be fact, and when you tell people what they know is wrong shit goes sideways

[imreloadin]

I'm calling it now, I'm debunking the myth that this thread will effectively debunk any PC myths...

[WoodenMarker]

3 hours ago, jstudrawa said:

It depends on the AIO and the Air Cooler.

It also depends on the system in question and how the cooler is installed. 

An AIO drawing fresh unfiltered air will perform much better than one drawing in warm air or when installed next to restrictive panels / meshes. 

[mr moose]

3 hours ago, Arika S said:

this thread is a mess already...

that's a myth.  This thread like many internet forums is well arranged with thoughtful posts. 

 

 

[Evanair]

Pads on the power supply keep it from shorting out the system.... .... ...

 

Thanks Verge.

[JovanD]

That static can kill your PC...
Many times when i was static charged i would get zapped when trying to plug in usb, the spark would go from my finger to usb port metal shield/ground, nothing would happen to pc, id get a unpleasant shock on the other hand, happened many times in the office.

[Boinbo]

@JovanD @Evanair @mr moose @WoodenMarker @imreloadin @jstudrawa @SenpaiKaplan and especially @campy, thank you I much for your thoughtful/information/funny responses!

[Mira Yurizaki]

A few things that stick out in my mind

 

Incremental CPU improvements area a recent thing

A common complaint I've heard about Intel's recent generations of CPU is that they're only doing incremental IPC improvements because they were enjoying a comfortable lead over AMD. In this case, incremental IPC improvements equate to somewhere around 10% better than the last generation.

 

So taking this I decided to go look at benchmarks of previous CPUs throughout time. What I found is that incremental IPC improvements have almost always been a thing. What kept us going for most of the 90s and 2000s was a healthy improvement in clock speeds. To put things in perspective, the Pentium, Pentium II, and Pentium III all had models that at the tail end ran at twice the speed as one of the earlier models. Basically, this is a 200% improvement within the same generation. And one time this happened within a year. The only exceptions to this pattern were Intel's Core 2 and AMD's Ryzen. However, it's only this way because they were both coming from lackluster architectures to begin with.

 

I looked into this at https://linustechtips.com/main/blogs/entry/1323-on-the-complaint-of-incremental-cpu-improvements/

 

 

Intel's "toothpaste" TIM

I encourage people to read this article because it offers plausible explanations, outside of unverifiable claims, for why Intel used thermal paste for TIM: https://medium.com/@OpenSeason/soldered-cpu-vs-cheap-paste-59fb96a4fca7

 

ARM in general can't match x86

While I'm hoping there's more data available about testing between the two architectures, the ISA doesn't make a processor faster or slower, it's the implementation of it.

 

Background applications in their entirety can eat up a significant amount of CPU resources, so you should reduce the amount you have

While background applications still eat up CPU resources, it's tiny and it's most likely because it set a timer to wake up periodically to check for something. Unless the background application actually has something to do, it's not using the CPU to any meaningful amount at all. I looked into making Windows 10 lean to see if it significantly improved performance, which has a similar effect on eliminating background processes: https://linustechtips.com/main/blogs/entry/1231-does-making-windows-10-lean-do-anything-for-gaming-performance/

 

 

Cloning is dangerous/error prone/etc, you shouldn't do it

I've cloned my OS from one storage to another various times, and sometimes in roundabout ways. Now I'm not saying it's 100% effective, however treating it as 100% not effective is just bad. This also leads to another semi-related myth: that going from an HDD to SSD if you're using Windows is bad because Windows won't know it went onto an SSD. Windows, at least since 8, can identify it's on an SSD even if its cloned from an HDD.

 

Task Manager doesn't know how to report things

One I recently looked at was the GPU page, because people have reported the GPU page sometimes is inaccurate as far as utilization is concerned. But the problem is that the GPU page can have a dozen plus things to report and there's a limited space it can do it in. In any case, I explained it at https://linustechtips.com/main/topic/1049404-fun-and-games-with-task-managers-gpu-page/

 

There's also a pervasive one like the CPU page doesn't know how to report clock speed. While I don't know much of the specifics about what Task Manager polls, it's a good guess that it's taking data from the CPUID instruction. The clock speed thing maybe due to the fact there's only one field and cores can be independently clocked.

 

But in the end, Task Manager's primary job, judging by the name of the app, can be implied that it's not a resource monitor. It just happens to be one since it's gathering resource utilization on all processes anyway. Also note that Task Manager is one of the few system tools you can run without elevating account privileges, which limits how much information the app can gather. Notice that every other monitoring tool needs to run with elevated privileges in order to work properly . So I guess on one hand, this "myth" isn't entirely a myth, but it really helps to understand why certain things the way they are, rather than chalk it up to "lol it doesn't work"

 

EDIT: If there's one thing I've noticed about a lot of things I looked at, it's that people make a claim at something but they never provide an answer that holds up to a decent amount of scrutiny or they've never actually went to prove their claims

 

7 hours ago, JovanD said:

That static can kill your PC...
Many times when i was static charged i would get zapped when trying to plug in usb, the spark would go from my finger to usb port metal shield/ground, nothing would happen to pc, id get a unpleasant shock on the other hand, happened many times in the office.

Nothing happened because that's exactly what ground's job is supposed to do: safely dump all excessive current to earth. In fact it's recommended if you don't use a static wristband (which is plugged into ground anyway) to touch the metal chassis of your case or PSU while it's plugged in.

 

Having a static discharge on anything that accepts an input, like V_in or data lines on a chip, can result in damaging something permanently.

 

Edited June 29, 2019 by Mira Yurizaki

[Boinbo]

@Mira YurizakiYurizaki, thanks for the long and thought out response, I really appreciate it!

[Guest]

Macs being bad for the money is a huge myth. 

[Boinbo]

2 minutes ago, floofer said:

Macs being bad for the money is a huge myth. 

Thank you!

[Drak3]

14 minutes ago, floofer said:

Macs being bad for the money is a huge myth. 

Eh, not really.

[Sauron]

1 hour ago, Mira Yurizaki said:

A few things that stick out in my mind

 

Incremental CPU improvements area a recent thing

A common complaint I've heard about Intel's recent generations of CPU is that they're only doing incremental IPC improvements because they were enjoying a comfortable lead over AMD. In this case, incremental IPC improvements equate to somewhere around 10% better than the last generation.

 

So taking this I decided to go look at benchmarks of previous CPUs throughout time. What I found is that incremental IPC improvements have almost always been a thing. What kept us going for most of the 90s and 2000s was a healthy improvement in clock speeds. To put things in perspective, the Pentium, Pentium II, and Pentium III all had models that at the tail end ran at twice the speed as one of the earlier models. Basically, this is a 200% improvement within the same generation. And one time this happened within a year. The only exceptions to this pattern were Intel's Core 2 and AMD's Ryzen. However, it's only this way because they were both coming from lackluster architectures to begin with.

The problem isn't inherently that they've only improved the IPC for about 6 years, it's that the improvements were small. It's simply no longer physically possible to improve IPC or clock speed by 200% in the span of a year so they should have looked for other avenues instead of sitting on their arse. The main reason we have multicore CPUs is that we got to a point where the increase in single core performance was slowing down to a crawl. If we had continued at the rate of the '90s we would have 10GHz single core processors with an IPC multiple times higher than what we have now.

 

Since I'm here I'll debunk a myth too:

"Chrome on my system with 16gb of ram uses >8gb of ram so you can't run Chrome with <x> tabs on an 8gb system"

Chrome, as many programs and most operating systems for that matter, uses as much memory as it can to make the user experience as snappy as possible. This does not mean it requires all that ram to function. In fact, all it really needs is enough ram to show a single page and maintain basic functionality; switching to another tab would simply cause the previous tab to be suspended if there isn't enough ram for both (of course this means it would be slower, but it shouldn't crash).

[Sauron]

18 minutes ago, floofer said:

Macs being bad for the money is a huge myth. 

*intensely disagrees*

[Taja]

30 minutes ago, floofer said:

Macs being bad for the money is a huge myth. 

Its not a myth, but also not entirely true.

 

Every person has different needs and tastes, so it is 100% subjective.

I would say that for A LOT of mac buyers, the mac is not worth it - because of marketing and other stupid reasons, ppl just think its better or are just used to it.

 

some people would say "paying $100 for a book is stupid" and then they go and buy a $300 painting. If you have the money and you spend it wisely, you can get a ton of value from something that others would never understand.

 

[Boinbo]

@Drak3 @Sauron and @Taja thanks for your responses!

[wasab]

Myth: Apple products are superior.

 

Debunked. 

https://www.google.com/amp/s/www.zdnet.com/google-amp/article/apple-is-having-a-bad-year-and-so-many-of-its-problems-are-self-inflicted/

 

Myth: Apple is a benevolent company. 

 

Debunked.

https://en.m.wikipedia.org/wiki/Criticism_of_Apple_Inc.