{Hardware} Accelerated GPU Scheduling Efficiency Revisited

Home windows 11 consists of an optionally available OS graphics function referred to as {Hardware}…

Home windows 11 consists of an optionally available OS graphics function referred to as {Hardware} Accelerated GPU Scheduling (abbreviated as HAGS), and this efficiency evaluation will showcase gaming utilizing the Gigabyte AORUS RTX 3080 MASTER. We are going to examine the efficiency of 26 PC video games with HAGS on versus HAGS off utilizing our newest really useful GeForce Recreation Prepared driver, 512.15, and Home windows 11 model 21H2.

Home windows 11 “Default graphics settings” – The best way to allow or disable HAGS

The Home windows 10 Might 2020 Replace (model 20H1) added this function, and we examined it utilizing this Home windows model in November 2020. Sadly, we noticed that the HAGS function led to efficiency inconsistencies on Home windows 10. That made us not suggest it for basic use on the time. We’re now revisiting HAGS efficiency utilizing the newest Home windows 11 public construct to see whether it is lastly price it.

You possibly can allow HAGS by means of the Home windows 11 “Default graphics settings” menu so long as you fulfill its necessities: Home windows 10 v20H1 or newer, a supported GPU (Pascal onwards from NVIDIA, and RX 5600 onwards from AMD), and HAGS suitable GPU drivers.

This time, our testing platform is a latest set up of Home windows 11 64-bit Professional Version, an i9-12900K with inventory clocks, an ASUS PRIME Z690-P D4 motherboard, and 32GB of T-FORCE XTREEM ARGB WHITE DDR4 3600MHz. The video games examined, settings, {hardware}, GeForce drivers, and Home windows 11 construct are similar apart from the HAGS’ states we examine (on vs. off).

Our AORUS RTX 3080 MASTER, rev. 1.0 (Ampere GPU micro-architecture).

Earlier than providing the efficiency knowledge and charts of every completely different evaluation state of affairs, it’s essential to explain each the {hardware} and software program configuration utilized in our testing, in addition to the evaluation methodology.

Benching Methodology

Check Configuration – {Hardware}

  • twelfth Gen Intel Core i9-12900K (Hyper-Threading/Turbo increase on; inventory settings)
  • ASUS PRIME Z690-P D4 motherboard (Intel Z690 chipset, v.1008 BIOS)
  • T-FORCE XTREEM ARGB WHITE 32GB DDR4 (2×16GB, dual-channel at 3600 MHz CL14 XMP), equipped by TeamGroup
  • Gigabyte AORUS GeForce RTX 3080 MASTER 10GB (rev. 1.0); v.F4 VBIOS, inventory clocks
  • 1 x Samsung 500GB SSD 960 EVO NVMe M.2
  • 2 x WD Blue 1TB SATA SSD
  • 1 x TeamGroup MP33 M.2 PCIe 1TB SSD
  • Corsair RM850x, 850W 80PLUS Gold energy provide unit
  • ASUS TUF Gaming VG289Q 28″ IPS UltraHD (3840×2160) 60Hz 5ms FreeSync Monitor for testing video games at 2160p decision.
  • ASUS ROG Swift PG279Q 27″ IPS QuadHD (2560 x 1440) 165Hz 4ms G-Sync Monitor for testing video games at 1440p decision.

Check Configuration – Software program

GeForce Driver Suite-related

  • We use DCH Recreation Prepared drivers.
  • The show driver is put in.
  • We set up the newest model of PhysX.

Recreation Benchmarks-related

  • We use the corresponding built-in or customized benchmark sequence.

Frametimes Seize & Evaluation tool-related

  • We use CapFrameX for capturing and analyzing the related efficiency numbers obtained from every recorded built-in or customized benchmark sequence.
  • Consecutive runs till detecting 3 legitimate runs (no outliers) that may be aggregated by CapFrameX utilizing the next technique:
    • ‘Combination excluding outliers’:
      • Outlier metric: Third, P0.2 (0.2% FPS percentile).
      • Outlier proportion: 3% (the % the FPS of an entry can differ from the median of all entries earlier than counting as an outlier).
  • We examine and worth the outcomes and aggregated information by way of percentages of achieve/loss, by setting the next thresholds to think about a sure % worth as vital (not inside the margin of error) for our benchmarking functions:
    • Rating/FPS Avg > 3% when valuing hybrid and non-synthetic benchmarks;
    • FPS Avg > 3% when evaluating uncooked efficiency;
    • P1/P0.2 > 3% when evaluating body time consistency; after making use of our customized method

{[(LowPercentileFPS_2 / AvgFPS_2) / (LowPercentileFPS_1 / AvgFPS_1)] – 1} x 100

Benchmark Suite: 26 PC Video games

DX11 Video games

  • Borderlands 3 (DX11)
  • Chernobylite (DX11)
  • Days Gone (DX11)
  • Deus Ex: Mankind Divided (DX11)
  • Far Cry New Daybreak (DX11)
  • God of Battle (DX11)
  • Neon Noir (DX11; Loop Mode)

DX12 Video games

  • Murderer’s Creed Valhalla (DX12)
  • Borderlands 3 (DX12)
  • Name of Obligation: Vanguard (DX12)
  • Cyberpunk 2077 (DX12)
  • DIRT 5 (DX12)
  • F1 2021 (DX12)
  • Far Cry 6 (DX12)
  • Forza Horizon 5 (DX12)
  • Ghostwire: Tokyo (DX12)
  • Hitman 3 (DX12)
  • Horizon Zero Daybreak (DX12)
  • LEGO® Builder’s Journey (DX12)
  • Metro Exodus PC Enhanced Version (DX12)
  • Shadow of the Tomb Raider (DX12)
  • Tom Clancy’s The Division 2 (DX12)
  • Watch Canine: Legion (DX12)

Vulkan Video games

  • DOOM Everlasting (VK)
  • Quake 2 RTX (VK; v.1.5.0)
  • Unusual Brigade (VK)
  • Wolfenstein: Youngblood (VK)

NVIDIA Management Panel settings

Listed here are the worldwide NVIDIA Management Panel settings:

NVIDIA Management Panel International 3D Settings (AORUS RTX 3080 MASTER).

Each ‘Excessive-High quality’ values for texture filtering-quality setting and ‘Want most efficiency’ for energy administration mode are set on a per-game or program profile foundation by way of Handle 3D Settings > Program settings tab.

The Efficiency Abstract Charts with 26 Video games

Under are the abstract charts of 26 video games used to check the video games’ efficiency adjustments with HAGS enabled and disabled utilizing the AORUS RTX 3080 MASTER.

You possibly can see the record of graphics settings on the charts, and we run every built-in or customized recreation benchmark’s sequence at 2560×1440, apart from Days GoneFar Cry New Daybreak, and Far Cry 6 examined at 3840×2160 decision. You might click on on every chart to open a pop-up for the very best viewing.

Outcomes give common framerates and larger is best. We show the low FPS percentiles (P1 and P0.2) beneath the corresponding averages. We use CapFrameX to document frametimes over time and to visualise and convert them into their corresponding common FPS and P1 and P0.2 FPS percentiles values.

There are additionally columns displaying percentages of achieve/loss in each uncooked efficiency (common FPS) and, when relevant, in frametimes consistency or stability between the completely different driver variations. We utilized the next customized method to calculate the soundness good points or losses:

{[(LowPercentileFPS_2 / AvgFPS_2) / (LowPercentileFPS_1 / AvgFPS_1)] – 1} x 100

We mark vital efficiency adjustments (larger than 3%in daring and use purple for the numerous enhancements or orange font for regressions.

{Hardware} Accelerated GPU Scheduling Efficiency Charts (On vs. Off)

hardware accelerated gpu scheduling
DirectX 11 Video games – Constructed-in benchmarks (besides Days Gone, Chernobylite, and God of Battle, examined utilizing BTR customized sequences).
hardware accelerated gpu scheduling
DirectX 12 Video games – Constructed-in benchmarks (besides Name of Obligation: Vanguards, Ghostwire: Tokyo, and LEGO® Builder’s Journey, examined utilizing BTR customized sequences).
hardware accelerated gpu scheduling
Vulkan Video games – Constructed-in benchmarks (besides DOOM Everlasting, examined utilizing a customized sequence).
Hardware Accelerated GPU Scheduling
DirectX Raytracing Video games – Constructed-in benchmarks (besides Ghostwire: Tokyo and LEGO Builder’s Journey, examined utilizing BTR customized sequences).
Hardware Accelerated GPU Scheduling
Vulkan Raytracing Video games – Constructed-in benchmarks (besides DOOM Everlasting, examined utilizing a customized sequence).

Notes on Home windows 11 {Hardware} Accelerated GPU Scheduling efficiency (On vs. Off)

For the sport benchmarks, though most video games present no vital efficiency variations with HAGS enabled, we see vital enhancements in Ghostwire: Tokyo (DX12), Metro Exodus PC Enhanced Version (DXR), and Wolfenstein: Youngblood (RTX) by way of frametime consistency; but in addition noteworthy efficiency regressions in Ghostwire: Tokyo (DX12) by way of uncooked efficiency, LEGO® Builder’s Journey (DX12 & RTX modes) and Quake 2 RTX (Vulkan Ray Pipeline & Vulkan Ray Question modes) by way of frametime consistency. Each LEGO® Builder’s Journey and Quake 2 RTX regressions are placing and are related to extreme and constant frametime stability and frame-pacing points throughout testing and gameplay.

Additionally, we noticed inconsistent and low GPU load in Tom Clancy’s The Division 2 (DX12) and skilled random crashes in Horizon Zero Daybreak (DX12) with HAGS enabled.

Disclaimer

Please remember that the next outcomes, notes, and the corresponding HAGS’ state advice are legitimate for comparable Ampere gaming rigs utilizing GeForce Recreation Prepared 512.15 driver and Home windows 11 v21H2. Its representativeness, applicability, and usefulness on completely different GPU architectures, testing benches, GPU drivers, and MS Home windows variations could fluctuate.

Conclusion

Primarily based on our earlier outcomes and findings, though we acknowledge that HAGS efficiency outcomes are higher on Home windows 11 than we confirmed up to now utilizing Home windows 10, we nonetheless see vital efficiency inconsistencies that stop us from recommending HAGS for basic or international use for GeForce avid gamers.

Anyway, the HAGS function continues to be fairly promising and may enhance efficiency in some instances, so we additionally suggest doing all of your testing to see the way it works together with your gaming rig and set of favourite video games.

See also  NVIDIA Resizable BAR Efficiency Revisited