5 Common Enthusiast Myths, Misconceptions, and Misunderstandings

Transitioning from being a PC enthusiast to a reviewer and then eventually to working for one of the companies that makes those products has given me an interesting perspective on hardware and system builds that doesn’t often reach the outside world. End users are typically being informed by reviews, forum chatter, and anecdotal experiences when they go to put together their builds, and this is all incredibly useful information. The problem lies in that there’s also a lot of misinformation, myth, and misunderstanding that gets spread, either because a misconception takes root in a community or something simply doesn’t get discussed.

With that in mind, shamelessly cribbing from the Cracked style, here are five myths that simply aren’t true.

1. Silent Cases Mean Silent Builds

There’s a niche market for cases like the Carbide Series 330R and the Obsidian Series 550D along with competing cases like NZXT’s H series, Nanoxia’s Deep Silence enclosures, and Fractal Design's Define series. These cases include acoustic dampening foam and tend to be much more closed off than other cases to prevent sound from escaping.

 

The Obsidian Series 550D is a fantastic case, but you have to choose the right parts to really take advantage of it.

Two corrections need to be made here. First, a “non-silent” case (one with lots of airflow and no dampening foam) can, with proper fan optimization and smart building, be incredibly quiet. My Carbide Air 540 build at home is audible but not especially so, and that owes to judicious use of speed controlled fans (and a liquid cooling loop managed by Corsair Link). Second, “silent” cases can actually be louder if quiet parts aren’t selected and airflow isn’t optimized. A “silent” case doesn’t take normal components and hush their noise, it takes already quiet components and makes them borderline inaudible.

2. Always Get a High Wattage Power Supply

Users routinely overestimate their power supply needs by a large margin, and this owes partially to the glut of poor quality, inefficient, high-wattage power supplies in the market at low price points. Buying a cheap power supply invites disaster, full stop, and the reality is that even a 400W power supply from a reputable brand is likely to perform better and last longer than a 600W power supply from a junk brand.

 

An efficient power supply like our CS650M will often be more than enough for a basic user's build.

But you need to understand how much wattage your system is going to want as well. My desktop system runs three SSDs, Corsair Link, a blu-ray burner, an overclocked i7-4770K, two GTX 780s, a watercooling loop with nine fans, 32GB of RAM, and a high-end PLX-enabled motherboard. Under sustained, extreme load, it pulls close to 700W. That’s an incredibly beefy system, and it could live pretty happily on just an AX760i, or an AX860i on the safe side. Most systems are going to run a single high performance graphics card and a single CPU; for most users, 500W is actually going to be fine, and 600W/650W will be plenty for room to grow.

With all that said, there are corner cases where you’d want a massive power supply. Our AXi and RM Series power supplies are very efficient and designed to run without active cooling at low loads; it’s only after they hit about 50% of their capacity that the fan even starts to spin up. They’re very quiet even at peak load, but if you want absolute silence, take your peak consumption (add up the TDPs of your parts) and boost it by about 1/3. That AXi or RM power supply will likely never need to spin up the fan.

3. DDR3-1600 is Fast Enough

Once common wisdom becomes entrenched in a community, it’s very difficult to unseat it, even if the science proves otherwise. 1600MHz DDR3 has been the standard for PC building for an eternity in tech years, but for systems built off of Intel’s Haswell architecture or one of AMD’s FM2/FM2+ APUs, it can actually bottleneck performance. We’ve investigated this on Kaveri, on Haswell, and in Battlefield 4.

The graphics cores on AMD’s APUs are notoriously memory bandwidth starved, but the CPU cores are essentially fine even at DDR3-1333 because they’re just not powerful enough to take full advantage of the bandwidth available. The same is not true of Intel’s Haswell chips.

 

The Dominator Platinums are fancy and look great, but even Vengeance or Vengeance Pro DIMMs are fine...as long as you're at DDR3-1866 or better.

Depending on the application being used or where the bottleneck in your system is, DDR3-1600 may very well be fine for an i7-4770K or i5-4670K. But as you overclock, the applications and games that are bottlenecked will only become more bottlenecked and may prevent you from seeing as much of a performance improvement as you’d hoped. Bottom line: for modern chips, you’re best off at least bumping up to DDR3-1866.

4. Bigger is Better For CPU Coolers

We have competitors releasing liquid coolers with 360mm radiators for the CPU soon, but we haven’t announced one and honestly, I wouldn’t expect us to. Why? Frankly, because 360mm is excessive and offers virtually no benefit for the cost. If a 120mm radiator is more than adequate for keeping a 200W+ GPU under 60C, why would we need 360mm for a 130W+ CPU? Even heavily overclocked, the CPU isn’t liable to push that level of capacity. To give you a clearer understanding, I’ve invited one of our thermal engineers, Bobby Kinstle, to explain:

“Heat transfers most efficiently when there is a large temperature difference between two regions.  As DeltaT diminishes, then heatsink area must increase to increase heat transfer.  Liquid cooling works very well because it has a very large radiator with large surface area.  It works so well, in fact, that in many cases the coolant is able to drop all of its heat into the air and return to the loop at or near ambient.  When the radiators reach a certain size, then the coolant temperature drops to the point where it’s no longer able to transfer heat into the radiator efficiently.  At this point, making the radiator larger won’t help at all.

 

The Hydro Series H110 is already overkill for most processors outside of an LGA2011 chip or one of AMD's 220W FX parts.

Heat transfer is non linear.  For this reason a small radiator (40mmx120mm) could easily cool a 300W GPU with enough airflow and a high enough coolant temperature.  In the case of most our coolers, the coolant temperature is less than 10C above ambient, so doubling the size of the radiator often only translates in a few % points increased performance.  Tripling the size of the radiator under these circumstances only diminishes the returns even further.”

5. “Why Didn’t <Manufacturer> Include <X, Y, Z>?”

This is an incredibly tough one to grok if you’ve never worked inside the industry. I’ve reviewed a lot of hardware where I was mystified by certain design decisions. One of my biggest ones was with the Corsair Hydro Series H90; that cooler performs much, much better with two fans, so why wasn’t anyone shipping a 140mm dual fan liquid cooler? The answer is simple: it would be too expensive. Not just too expensive for the company, but too expensive for the end user. At that point, the manufacturing cost gets perilously close to the 240mm Hydro Series H100i.

 

A second fan radically improves the H90's performance, but including it actually winds up being too costly.

Honestly, “prohibitively expensive” winds up being the answer to a lot of the “why” questions. Around here we’re typically willing to spend up if the investment is worth it, if it gets us features or performance that we know our users want. Sometimes the increase is just too great, though; you end up with a fantastic product that has to be priced essentially out of competition. Every design that isn’t an AX1500i or Obsidian Series 900D is essentially an exercise in compromise so that we don’t wind up releasing a “Homer" (for all you Simpsons nerds out there.)

The other possibility is that we simply weren’t certain how well a product would be received. People have asked for different sizes of the Carbide Air 540, but the Carbide Air 540 was an experiment. Development takes a remarkably long time, and we can’t start planning a successor until we’ve gotten feedback from our customers. By the same token, our K series keyboards and our fans all sold well beyond expectations, leaving us playing catch up.

The important takeaway is that the people designing the products you enjoy are smart, and there are usually very good reasons why they didn’t come out the way you wanted or expected.