This site may earn affiliate commissions from the links on this folio. Terms of utilize.

One question that's been popping up with increasing frequency when we talk near high-end graphics cards is whether 4GB of RAM is plenty to power electric current and next-generation gaming. When nosotros initially covered AMD's Fury X launch, I promised to return to this topic and embrace it in more particular. Earlier we can hitting the data, all the same, we need to talk about how VRAM management works and what tools are bachelor to evaluate information technology in DirectX 11.

FuryCard2

While it might seem straightforward to test whether or not any given title uses more than 4GB of RAM, the tools for doing this are rather inexact. The GPU itself does not control which information is loaded into memory. Instead, retentivity management is handled by the operating system and the GPU driver. The GPU tells the Bone how much memory it has, but it doesn't brand any decisions about how data is loaded or which data is loaded start.

One style that game developers handle retentivity management in software is by creating game presets that assume a detail amount of VRAM is nowadays on the card. Low item might be tuned to run on 512MB cards, while ultra particular assumes you have at least 4GB of VRAM. If y'all choose a detail level that calls for more VRAM than is nowadays on your menu, you'll likely see a heavy performance hit as the arrangement is forced to load data from main memory.

Memory usage in Shadow of Mordor

Retentiveness usage in Shadow of Mordor

SoM2

Shadow of Mordor frame rate scaling past resolution between all three cards

Some games won't use much VRAM, no matter how much you offering them, while others are more than opportunistic. This is critically important for our purposes, because there'south non an automatic link between the amount of VRAM a game is using and the corporeality of VRAM it actually requires to run. Our first article on the Fury X showed how Shadow of Mordor actually used dramatically more VRAM on the GTX Titan X every bit compared with the GTX 980 Ti, without offering a higher frame charge per unit. Until we hit 8K, there was no performance advantage to the huge memory buffer in the GTX Titan 10 — and the game ran so slowly at that resolution, information technology was incommunicable to play on any card.

GPU-Z: An imperfect tool

GPU-Z claims to report how much VRAM the GPU actually uses, but there's a significant caveat to this metric. GPU-Z doesn't actually report how much VRAM the GPU is actually using — instead, it reports the amount of VRAM that a game has requested. We spoke to Nvidia'south Brandon Bong on this topic, who told us the post-obit: "None of the GPU tools on the market study retentiveness usage correctly, whether information technology'south GPU-Z, Afterburner, Precision, etc. They all report the amount of memory requested by the GPU, not the bodily memory usage. Cards will larger memory will asking more retentivity, just that doesn't mean that they actually employ it. They simply request it because the memory is available."

GPU-Z

There are other tools, similar Procedure Explorer, which can also capture GPU memory requests — only they don't confirm actual retentiveness usage, either.

Our own testing backed up this claim; VRAM monitoring is subject to a number of constraints. Resolution switching or visiting more than than 1 area before beginning testing can significantly increase full memory "in use" without actually impacting performance at all. In that location's also a moderate amount of variance between test runs. Nosotros can say that the GPU requested effectually four.5GB of RAM, for instance, but one test run might bear witness a GPU topping at 4.3GB while the adjacent showed a maximum RAM consumption of iv.5GB. Reported VRAM consumption tin also vary during the game; logging and playthroughs must exist advisedly managed.

Finding >4GB games

When we started this process, I assumed that a number of high-end titles could readily be provoked into using more than 4GB of VRAM. In reality, this proved a tough nut to crack. Enough of titles meridian out effectually 4GB, merely nearly don't exceed it. Given the lack of precision in VRAM testing, we needed games that could unambiguously break the 4GB limit.

Nosotros tested Assassin's Creed Unity, Battlefield 4, BioShock Infinite, Culture: Beyond Earth,  Company of Heroes 2, Crysis 3, Dragon Age: Inquisition, The Evil Within, Far Cry 4, Yard Theft Auto V, Metro Last Lite (original), Rome: Total War 2, Shadow of Mordor, Tomb Raider, and The Witcher 3: Wild Hunt. Out of those xv titles, just four of them could exist coaxed into significantly exceeding the 4GB limit: Shadow of Mordor, Assassin'due south Creed: Unity, Far Weep four, and Thou Theft Auto V. Fifty-fifty in these games, nosotros had to use extremely loftier detail settings to ensure that the GPUs would regularly report well over 4GB of RAM in utilise.

Our testbed for this project was an Intel Core i7-5960X with 16GB of DDR4-2667 running Windows eight.1 with all patches and updates installed. While Windows 10 has just recently launched, we began this project on Windows 8.one and wanted to stop it there. Transitioning operating systems would have necessitated a complete retest of our titles. We tested AMD's Radeon R9 Fury 10, Nvidia's GTX 980 Ti, and the top-finish GTX Titan 10. With Titan X, we were curious to see if we'd see any benefits to running with a 12GB RAM buffer over and above the 6GB buffer on the GTX 980 Ti.

In every case, the games in question were pushed to maximum detail levels. MSAA was not used, since that incurs its ain operation penalty and could warp results, but the highest non-GameWorks settings were used in all standard menus. GW-specific implementations available just on Nvidia hardware were left disabled to create a level testing field. The i exception to this was Yard Theft Auto V, where we used Nvidia'south PCSS shadows for its cards, and AMD'due south preferred CHS shadows for the Fury Ten.

GameWorks, operation, and 4GB VRAM

At that place'due south 1 common factor that ties three of our iv >4GB titles together — GameWorks. Iii of the 4 games that we've tested (Far Weep iv, Assassin's Creed Unity, and One thousand Theft Auto Five) are Nvidia GameWorks titles, meaning they make apply of Nvidia-provided libraries to provide key DirectX eleven functions like ambient occlusion, soft shadows, and tessellation. AMD cannot optimize for these games to the same degree and AMD GPUs tend to perform significantly worse in GameWorks titles than in other games. Nosotros've discussed GameWorks, its implications, and its bear upon on various titles at multiple times in the past few years.

One affair I desire to stress is that while nosotros'll be looking at performance data in this article, its primary purpose isn't to compare how the Fury X stacks up, performance-wise, against Nvidia's highest-end GPUs. Such comparisons are inevitable, to some extent, but this isn't a standard review. We've created specialized exam cases designed to examination a theory and used settings that significantly depart from what nosotros consider playable or advisable for 4K testing. As such, the 4K performance results in this story should not be treated as typical results for Nvidia or AMD. The goal of these tests is to create a worst-example scenario for a GPU with 4GB of VRAM and see what happens as a outcome.

We covered Shadow of Mordor in our initial Fury X coverage, so this article will business organisation itself with the three new games we tested. Let's boot things off with Far Cry 4.

  • one of five