Go Back   OC3D Forums > [OC3D] Off Topic > Off Topic Chat > Rants
Reply
 
Thread Tools Display Modes
 
  #1  
Old 08-11-13, 05:22 PM
Zoot's Avatar
Zoot Zoot is offline
OC3D Elite
 
Join Date: Jul 2008
Location: Ireland
Posts: 1,547
When PC Hardware Sites Don't Understand Electronics

Initial Disclaimer: I actually am an Electronic Engineer so this kind of stuff does indeed bug me.

Most PC hardware users/enthusiasts on forums/sites like this and elsewhere are really just PC users with above average knowledge of hardware/software, but not really electronics. This is fine for things like a CPU or graphics card since you just run benchmarks and compare the results, you don't actually need any electronic knowledge of the internals of either. It becomes a problem in other areas though.

One example, which is sadly quite frequent of pc enthusiast site reviewers not understanding electronics is PSU reviews. You really need a proper knowledge of power electronics to understand the internals of a PC PSU. Take this review for instance:
http://www.hitechlegion.com/reviews/...-xigmatek-p700

The authors are using a traditional pc as a load to test the voltage regulation performance. There's a number of problems with this, one of them being that traditional PC hardware isn't going to nearly load all the rails of the PSU to properly test their performance, also it's not going to provide a continuous load.

Normally to test the performance of a certain rail of any power supply what you do is put a current load on it, or rather pull current from it. Lets say your PSU is rated for rated for say 60A on a single 12V rail; you'd test it by loading that rail with up to 60A and investigating how close the voltage is to the target (12V), you'd do the same with the other rails too; namely the 3.3V and the 5V rail up to their rated current load. There's a certain point whereby it'll go out of regulation and the voltage will drop away quite fast from the 12V target as you increase the current, but if it's designed correctly and it's rated for 60A, then that point should be above 60A.

In the above review, because they're using traditional PC hardware to load the PSU, and measuring the power draw at the wall, we've no idea what kind of current is being pulled from the power rails, which makes their voltage regulation measurements worthless. What's the performance of the 3.3V rail like? How about the 5V rail? What happens if you load the 12V rail with 90A? Does the unit shut down, or blow up? Also not to mention there's actually two 12V rails, what's the performance like for both of them? We don't know from the review, and therefore it's useless.

The way to solve this is by using active load testing that actually does what I say above. Most of the big tech sites actually do this (including OC3D) and to be fair to the authors of the above review, they actually do acknowledge the lack of active load testing, but sadly this is something you'll see in a lot of PSU reviews thanks to a lack of understanding, which only tells a small fraction of the full story. It's a problem, given that somebody could buy a PSU and have it blow up despite getting "Gold" reviews for the reason it wasn't tested properly.

This is even BEFORE you get to stuff like noise and ripple. Stick to easier stuff guys...

Here's another example of a lack of understanding of electronics, the guy in the below article is talking about the Radeon 7990 and the GTX 690, and he says both have two GPUs on a single piece of silicon. I can't believe someone would not know the difference between the actual silicon die (the GPU) and the PCB it resides upon. Both have two GPUs yes, but on the same PCB, not the same die. Of course there's also the fact that getting two of those GPUs onto the same die would currently be impossible.
www.techoftomorrow.com/2013/pc/shootout-amd-hd-7990-vs-nvidia-gtx-690-performance-results/

Another example:
http://www.bit-tech.net/hardware/gra...-290x-too-hot/
The author states the limit silicon can operate at is 100-105C. This is completely false, I can point to ICs (chips) that are rated all the way up to 190C junction temperature. Actually, most of the electronics in your car have to be rated to operate at these kind of temperatures to guarantee functionality (125C, 150C etc.)

I could go on and on, since there's many examples of this out there.

TL;DR If you don't understand electronics, don't try and pretend you do.

/rant

__________________
Desktop: AMD Ryzen 7 3700X | X570i Aorus | 32GB Corsair DDR4-3200MHz | RX 5700 XT | Aorus PCIe 4.0 1TB M.2 | Fractal Define Nano S
Home Server: AMD Ryzen 3 2200G | Asus Prime B450M-A | 16GB Crucial DDR4-2400MHz | 39TB | Fractal Design Define R5
HTPC: AMD Athlon 5350 | ASRock AM1H-ITX | 4GB Corsair DDR3-1333MHz | MiniBox M350






Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump










All times are GMT. The time now is 12:18 PM.
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.