Started By
Message

re: Build the Best Gaming PC Your Money Can Buy: A Detailed Guide (Updated Sep 2014)

Posted on 9/29/13 at 2:10 am to
Posted by ILikeLSUToo
Central, LA
Member since Jan 2008
18018 posts
Posted on 9/29/13 at 2:10 am to
------------------------
++++ALERT: You are reading an out-of-date version of the guide and wasting your time. Read the PDF for the most accurate up-to-date info.It's best to download the PDF and use a proper PDF reader. Google's formatting of PDFs breaks all of the links. Link to directly download the PDF. I have stopped updating the text in the thread because the forum's limited code makes it far too time-consuming to change images and add text.++++
------------------------

Before we move on, here are a few things to consider about the current state of video cards:

> New cards are just around the corner – If you’re willing to wait for the latest and greatest (and you have the extra cash to spend), NVIDIA will be releasing a new generation in early–mid 2014. NVIDIA is also set to release an upgraded version of the 780, dubbed the 780 Ti, in November. The 780 Ti is rumored to be faster than the Titan and priced around $700.

> Video Memory (VRAM) is becoming a stronger point of consideration – I can’t argue with the fact that most of today’s games, even at 1440p, don’t necessarily require more than 2GB of VRAM, but I do know that with the release of the PS4 and XBox one and HD gaming becoming the norm, games are gearing up to be memory hogs by design. Remember, many popular PC titles are console ports, which were originally designed to use 512MB VRAM max. The next-gen consoles are going to have access to several GB of memory for their GPUs (even after you factor in the memory for the console’s OS, networking, and the other background functions). Those PS4 and Xbox One ports will be part of our list of things to play.

====///====Choosing the Right AIB Partner====\\\====
Oh, you thought we were done? Once you’ve chosen between NVIDIA and AMD, it’s time to choose the right brand. AMD and NVIDIA sell their GPUs, as well as a reference design of the full card, to their Add-in-Board (AIB) partners. Major partners include ASUS, EVGA, Galaxy, Gigabyte, HIS, MSI, PNY, PowerColor, Sapphire, VisionTek, XFX, and ZOTAC. Not every company is known to sell both AMD and NVIDIA.

AMD and NVIDIA leave it up to those companies to decide how they want to market their video cards. In general, you can expect the following variables:

====Cooling====
Video cards come with a wide range of pre-mounted cooling options, depending on the brand. In many cases, the cooler is the only variable worth considering when deciding between multiple cards with the same GPU.

###Reference Coolers###



As shown above, reference coolers are usually a GPU heat sink and a single fan housed in a plastic shroud. It’s called a reference cooler because it’s fabricated based on a design by the GPU manufacturer (AMD or NVIDIA), meant to provide adequate cooling to keep your GPU at the specified “safe” temperature. For the most part, it can be an effective cooler. The main drawback is that the fan can get loud at higher temperatures, and turning it down means potentially bringing your GPU to a temperature threshold where throttling (automatic GPU speed reduction) kicks in to protect the GPU from overheating. Thus, performance suffers. The fan noise is less of a problem if you game with headphones, so don’t let the noise deter you if you can find a good deal on it.

###Aftermarket/Custom Coolers###
For a better/similar cooler and less noise, you may want to opt for a model that ships with an aftermarket cooler, such MSI’s Twin Frozr, shown below:



It should be noted that no matter what cooler you choose, you will often have to set the fan profiles yourself. GPUs have target load temperatures set by default (usually around 80C), and your GPU cooler fan(s) will speed up and slow down accordingly. You might still see your load temperatures at 80 with an aftermarket cooler, but the cooler maintains those temperatures much more quietly.

You can use various software solutions to adjust your fan profile to set a lower load temperature goal, such as 65-70—the cooler will work harder and louder, but will still be quieter than a reference cooler trying to maintain those same temperatures.

There is a tradeoff to using custom coolers. Reference coolers often use a loud, fast blower fan that exhausts the GPU heat out the back of your case. This method is very effective, but only if you crank up the fan. The custom coolers, while quieter and overall more effective at cooling your GPU, usually dissipate heat inside the case rather than outside. This, of course, can add heat to other components such as your CPU, but an effective CPU cooler and proper case fan arrangement can offset this. See the diagram below:


This diagram represents a typical ATX case containing an aftermarket CPU cooler and three 120mm case fans for front intake, rear exhaust, and top exhaust.
This post was edited on 3/20/14 at 3:34 pm
Posted by ILikeLSUToo
Central, LA
Member since Jan 2008
18018 posts
Posted on 9/29/13 at 2:10 am to
------------------------
++++ALERT: You are reading an out-of-date version of the guide and wasting your time. Read the PDF for the most accurate up-to-date info.It's best to download the PDF and use a proper PDF reader. Google's formatting of PDFs breaks all of the links. Link to directly download the PDF. I have stopped updating the text in the thread because the forum's limited code makes it far too time-consuming to change images and add text.++++
------------------------

====Printed Circuit Boards (PCBs)====
This is what I mean by a PCB:


Technically, the PCB is the actual board that holds everything else, including capacitors, mosfets, GPU, memory modules, conductive pathways, etc. But, for the purpose of this discussion, PCB refers to the board plus everything on it except the GPU and VRAM.

For simplicity sake, there are two types of PCBs to consider in video cards: Reference and Non-Reference.

A reference PCB is one that was designed by the GPU manufacturer (AMD or NVIDIA) specifically to run your GPU and VRAM—the factory standard. This design is provided to their Add-in-Board (AIB) Partners, which include the companies mentioned earlier. Some companies simply sell the GPUs with reference PCBs and reference coolers, merely slapping their brand name on the factory design. Other companies design better air coolers and sell them with reference PCBs. And then there are those that design their own PCBs for these GPUs. These “non-reference” PCBs may have added enhancements such as better power phase designs and custom coolers, all of which lead to better overclocking potential (and some PCBs are simply redesigned to accommodate the AIB partner’s custom cooler).

On the other hand, there is a dirtier side to non-reference PCBs, because some AIB partners design them with the exact opposite in mind—rather, the goal is to cut costs, make those PCBs as cheaply as possible, the minimum required for the card to run at its stock specification. When AMD and NVIDIA design their reference PCBs, every component on the card has a definite purpose. But an AIB attempting to be highly competitive will put the cards through a “component reduction” process, where the corporate bean counters make decisions on what components (voltage regulators, mosfets, capacitors, etc.) aren’t completely necessary, or which ones can be replaced by cheaper and inferior components.

You see more and more of these types of cheap non-reference cards pop up as a generation of GPUs get older, so always be diligent about checking reviews of specific models when looking for a deal, especially ones that appear to significantly undercut the price of the others. This “cost-cutting” extends to the cooling solution as well, and these cards tend to have heat and stability problems under extended workloads; they certainly aren’t going to overclock very well, if at all. In fact, quite a few non-reference cards come with locked voltage.

Reference PCBs also overclock fairly well and accommodate a variety of after-market air coolers. If you only plan to air cool, a non-reference card with an enhanced PCB and better cooling would be a good choice, as long as the price isn’t too far above reference designs.

====Warranty====
Video card warranties range from 1 year to lifetime. A typical warranty will be 3 years. Don’t overpay just because a card has a longer warranty, as not all companies honor their warranties as readily as others. Do, however, be wary of cards with shorter warranties, as they may be one of those “cost cutter” designs. In general, you can expect a defective card to show symptoms within the first 90 days anyway.

====Overall Reputation====
How well does the company honor its warranties? How poor is their customer service, phone support, etc.? How long does it take to get a replacement from them if your card fails? Here’s a hint: They’re all pretty bad. I guarantee you there are an endless amount of horror stories posted on the internet about every single company and their customer service. Your best bet is to read reviews and consider the rest of the criteria mentioned above, so that you buy a card with statistically a lower chance of needing warranty replacement. Companies who have to deal with fewer replacements may be more likely to provide a faster replacement if something does happen.

====///====My Video Card Picks====\\\====
$1,000–$1,100 budget—AMD Radeon HD 7970 (3GB)
Why the 7970: It comes in at the right price, around $300-$320 for decent brands as of August 2013. NVIDIA’s 2GB 770 may be a slightly better performer, but it doesn’t justify the price. Look for a 7970 under $320 (after rebates and promos) that has the best reviews. Even better if it’s a factory overclocked (GHz) model, but you may not find one at this budget. If you have the extra money, I’d recommend spending as much as $330-340 on a GHz model with well-reviewed custom cooling. If you opt for NVIDIA, I’d recommend the 4GB 770 over the 2GB version. It’s a $440 card, so you’d have to extend your budget by over $100. I honestly cannot make an educated assessment on whether or not it’s worth it.

$800 budget—AMD Radeon HD 7950 (3GB)
Why the 7950—While not as powerful as a 7970, it’s remarkably close. And given its price tag of $240 and below, it easily holds the title of best bang for buck. There have been some recent sales putting some of these cards under $200, including good brands such as the MSI Twin Frozr. If you prefer, $250 will get you an NVIDIA 760 (2GB). These cards are about even in current game performance, so the 7950 was chosen due to price and the extra GB of VRAM.

$600 budget—AMD Radeon HD 7870 (2GB) or AMD Radeon HD 7950 (3GB)
Why the 7870—Quite simply, it’s the only current gen card that can compete for performance at this budget. Many of them are comparable to a 7950’s performance. On the other hand, with the recent sales of 7950s, there is barely a $20 difference in price. You might as well opt for a well-reviewed 7950 at a $200 price point if possible. If the 7870 prices fall further to create a larger price gap between it and the 7950, I have no problem recommending the 7870.

Yes, I realize the above recommendations make it seem like I’m AMD biased. Believe it or not, I was looking for a reason to recommend the NVIDIAs first and foremost in the $1000 and $800 budget categories. However, the current pricing does not justify the performance increase obtained through a mere refresh of old architecture. They simply do not fit within these predefined budgets without making other hardware compromises that I believe to be unacceptable.
This post was edited on 3/20/14 at 3:34 pm
first pageprev pagePage 1 of 1Next pagelast page
refresh

Back to top
logoFollow TigerDroppings for LSU Football News
Follow us on Twitter, Facebook and Instagram to get the latest updates on LSU Football and Recruiting.

FacebookTwitterInstagram