Best video card for battlefield 4: Battlefield 4 — The NVIDIA GeForce GTX 1080 & GTX 1070 Founders Editions Review: Kicking Off the FinFET Generation

Test stand configuration
CPU	Intel Core i7-3960X @ 4. 6GHz (100×46)	Intel Core i7-3970X @ 4.6 GHz (100×46)
Motherboard	ASUS P9X79 Pro
RAM	DDR3 Kingston HyperX 4×2 GB @ 1600 MHz, 9-9-9
ROM	Intel SSD 520 240 GB
Power supply	Corsair AX1200i, 1200W	Seasonic Platinum-1000, 1000 W
CPU cooler	Thermalright Archon
Body	CoolerMaster Test Bench V1.0
Operating system	Windows 7 Ultimate X64 Service Pack 1
AMD card software	AMD Catalyst 13. 14 Beta
Software for NVIDIA cards	335.23 WHQL

CPU power-saving technologies are disabled in all tests. The PCI-Express bus operates in 3.0 mode. To activate PCI-E 3.0 on GeForce 600 and 700 series video cards in a system based on the X79 chipset, a patch from NVIDIA is applied.

In the NVIDIA driver settings, the CPU is always selected as the processor for PhysX calculation. In the AMD settings, the Tesselation setting is always moved from the AMD Optimized state to Use application settings. In CrossFire configurations, the Frame Pacing option remains enabled.

The scripted scene at the beginning of the Tashgar mission was used for the benchmark. Testing took place in three quality modes — Low, High, Ultra. Medium is dropped from the list because the performance difference between it and High mode is relatively small. Each mode has tests with resolutions of 1920×1080 and 2560×1440.

The participation of a single adapter in tests with different quality was determined as follows. If the frame rate reached 60 FPS at a resolution of 1920×1080 in the most demanding mode, the testing was over. If the frame rate was less than 60, but more than 30 FPS, an additional test was carried out in a lower quality mode. At less than 30 FPS with a resolution of 1920×1080 adapter results are not shown in this mode.

⇡#Test participants

Video cards on GPU AMD

• AMD Radeon HD 7990 (1000/6000 MHz, 6 GB)
• AMD Radeon R9 290X (1000/5000MHz, 4GB), Uber Mode
• AMD Radeon R9 290 (947/5000MHz, 4GB)
• AMD Radeon R9 280X (1000/6000MHz, 3GB)
• AMD Radeon HD 7950 w/Boost (925/5000MHz, 3GB)
• AMD Radeon R9 270X (1050/5600MHz, 2GB)
• AMD Radeon R9 270 (925/5600MHz, 2GB)
• AMD Radeon R7 260X (1100/6500MHz, 2GB)
• AMD Radeon HD 7770 (1000/4500 MHz, 1 GB)
• AMD Radeon R7 250 (1050/4600MHz, 2GB)

NVIDIA GPU graphics cards

• NVIDIA GeForce GTX TITAN Black (989/7000MHz, 6GB)
• NVIDIA GeForce GTX 780 Ti (875/7000MHz, 3GB)
• NVIDIA GeForce GTX 780 (863/6008MHz, 3GB)
• NVIDIA GeForce GTX 690 (915/6008 MHz, 4 GB)
• NVIDIA GeForce GTX 770 (1046/7012MHz, 2GB)
• NVIDIA GeForce GTX 670 (915/6008MHz, 2GB)
• NVIDIA GeForce GTX 760 (980/6008MHz, 2GB)
• NVIDIA GeForce GTX 660 (980/6008MHz, 2GB)
• NVIDIA GeForce GTX 650 Ti (928/5400MHz, 1GB)
• NVIDIA GeForce GTX 650 (1058/5000MHz, 1GB)
• NVIDIA GeForce GT 630 (902/1800MHz, 1GB)

⇡ # Test results: Ultra

1920×1080

• No wonder that video cards based on two GK104 or Tahiti chips became the kings of the test. And even though the Radeon HD 7990 is no longer being produced, and the GeForce GTX 690 is inadequately expensive, no one bothers to assemble a similar configuration from two separate Radeon R9 280X or GeForce GTX 770.
• Among video cards with one GPU, the GeForce GTX TITAN Black predictably won, which is currently awaiting a full review on 3DNews. The GTX 780 Ti also outperformed AMD’s flagship video adapter.
• The minimum acceptable frame rate (30 FPS) among NVIDIA adapters was provided by GeForce GTX 660, and among AMD adapters by Radeon R9 270.

2560×1440

• Increasing the resolution did not affect the ratio of test participants’ results. The only exception: the GeForce GTX 770 and Radeon R9 280X are swapped.
• A GeForce GTX 760 or Radeon HD 7950 with Boost is sufficient for playable frame rates. Last in testing replaces Radeon R9280, close to him in specifications.

⇡#Test results: High

1920×1080

• Radeon R9 280X leads with a small margin from GeForce GTX 770.
• The frame rate threshold (30 FPS) is already provided by rather weak graphics cards — Radeon HD 7770 and GTX 650 Ti. By the way, the HD 7770 has been discontinued, but is a complete copy of the «new» Radeon R7 250X.

2560×1440

• With the change of resolution, there was no dramatic change in positions in the rating. In order to get the minimum frame rate, you need a Radeon R7 270 or GeForce GTX 660.

⇡#Test results: Low

1920×1080

• Even relatively weak adapters produce a high frame rate with such undemanding settings. Even the GeForce GTX 650 Ti and Radeon R7 250 have a performance margin for a more or less comfortable game.
• Even the GeForce GT 630 frankly slows down, but it tries to somehow pull out Battlefield 4. According to the selection rules, it should have been excluded from the chart, since the frame rate dropped below 30 FPS. But let it remain so that it is clear which video card for BF4 is still not enough under any circumstances.

2560×1440

• The chart is missing a Radeon R7 250 result and this is not an error. There is no DL-DVI or DisplayPort output on the existing sample. As a result, resolutions above 1920×1200 are not available to him.
• The modest power of the GeForce GTX 650 Ti is enough to keep the average frame rate at 30 FPS.

⇡#Mantle vs DirectX 11: Ultra

Let’s get back to the Mantle theme. Recent testing has already answered many questions and generally confirmed AMD’s claims related to the new API. Finally, I would like to check how legitimate it is to transfer the results previously obtained using the Radeon R9290X and R7 260X are graphics adapters based on the GCN 1.1 architecture for the entire modern line of AMD discrete cards, including models with the architecture conventionally called GCN 1.0. It is already clear that the benchmark in the Tashgar mission is not very favorable for Mantle, but still allows you to achieve this goal.

In general, as you might expect, the advantage of Mantle gradually disappears as GPU performance decreases, provided that the CPU performance remains unchanged (and quite high in the case of our test bench). The curiosity is that in some cases the transition from DirectX to Mantle does not increase, but even reduces the frame rate. An example of this is the performance of the Radeon R9270 and HD 7770 in High mode, and additionally — again HD 7770 in Low mode at a resolution of 2560×1440.

Conclusion: Mantle support still needs to be improved on the side of either the driver or Battlefield 4 itself. and very cheap adapters, barely different from the “plug in the case” category. At maximum image quality settings, the game is very demanding — it needs a GeForce GTX 780 or Radeon R9-class graphics card290 to achieve a comfortable 60 FPS at a resolution of 1920×1080. The same task in 2560×1440 mode is only possible with two GPUs in the Radeon HD 7990 or GeForce GTX 690.

If you do not chase the maximum detail settings and sacrifice full-screen anti-aliasing, then the Radeon R9 270 or GeForce GTX 760 will be enough for Full HD resolution Finally, even the GeForce GTX 650 is conditionally suitable for Battlefield 4 with the most benign settings.

⇣ Contents

If you notice an error, select it with the mouse and press CTRL+ENTER.

Related materials

Permanent URL: https://3dnews.ru/811783/page-1.html

Tags:
testing, graphics card, battlefield 4, mantle, radeon, geforce, frostbite 3

⇣ Comments