Wikipedia

User:ScotXW/amdgpu

< User:ScotXW
Illustration of the support for AMD's GCN under Linux.

amdgpu is half of the device driver for AMD's GCN-based GPUs. There are two other "halfs" available, radeonsi, that is part of Mesa 3D and a proprietary half called AMDGPU-PRO. Originally the UMD radeonsi was written to work on top of the KMD radeon and AMDs proprietary UMD to work on top of fglrx.ko.

There is also RADV, which implements the Vulkan 1.0 specifications. RADV is strongly based on Intel's ANV driver but works on top of amdgpu (like radeonsi does, only that radeonsi implements OpenGL).[1]

History edit

  • On January 9th, 2012 the first GCN-based product, the "Tahiti XT"-codenamed GPU (4313 million transistors organized as 2048:128:32) found on the Radeon HD 7970 graphics cards was released. Support for this card was added to the proprietary fglrx.ko as part of the AMD Catalyst for Linux driver package and later to /drm/radeon, serving as basis for the new UMD radeonsi in Mesa. Linux kernel 3.2 was released on January 4th, 2012 and 3.3 on March 18th, the same year.
  • On March 22nd, 2013 the first GCN 2-based product, the "Bonaire XT"-codenamed GPU (2080 million transistors organized as 896:56:16) found on Radeon HD 7790 was released.
  • On October 24th, 2013 the second GCN 2-based product, the "Hawaii XT"-codenamed GPU (6200 million transistors organized as 2816:176:64) was released.
  • On October 7th, 2014 at the XDC 2014 Alex Deucher made the very fist announcement of AMD's new Linux support with a shared open-source kernel-mode driver[2] This was confirmed at the GDC 2015.
  • In April 2015, the initial release of the new amdgpu-stack was announced on the dri-devel mailing list.[3]
  • On August 30th, 2015 Linux kernel 4.2 including drivers/gpu/drm/amd/amdgpu was released; an own mailing list "amd-gfx" has been created for this development.
  • DAL (Display Abstraction Layer) is part of amdgpu and was initially a huge set of patches (about 93,000 LoC) to the Linux kernel; but DAL has not been accepted into the Linux kernel; DAL is required to be distributed as part of AMD's new proprietary Linux driver "AMDGPU-Pro" and is being maintained and developed out-of-tree
    • the DAL code was initially made available to the public in April 2015 on cgit.freedesktop.org/~agd5f
    • On September 17th, 2015 at XDC there was another talk by Deucher and Zhou[4], the slides mention a new display component called "DAL" and a new new power component called "Powerplay" (though "powerplay" was replaced with "powertune")
    • On 2016-02-11 Harry Wentland proposed to upstream DAL Enabling new DAL display driver for amdgpu on Carrizo and Tonga and was rejected
    • On 2016-09-23 Harry Wentland presented DAL at the XDC2016: DAL presentation
    • On 2016-12-08 Harry Wentland wrote a RFC regarding the acceptance of the DAL patch set
    • The issue with out-of-tree code is that its developers need to rebase it from time to time and align it with current upstream. This consumes a lot of time, that might be used for fixing actual issues. To save time, DAL/DC hasn't been re-based from 4.4 to newer kernels. Though it seems that it has finally been moved to 4.14-devel to be mainlined into 4.15.
  • On February 16th, 2016 Vulkan 1.0 was released Vulkan press-release
  • In 2017-H1 Vega-GPUs will be available: hardware-wise "rasterisation and render output units" were changed and a much improved energy efficiency is expected from that.

Table edit

Version Release date browse diff to prior mails: drm-next pull request Notes Kernel version in Linux distributions:
Stable Rolling
Old version, no longer maintained: 4.2 2015-08-30 4.2 diff 4.2/4.1 pull drm-next-4.2 Basic support for GCN 3rd (Tonga, Fiji, Carrizo, Bristol Ridge),
Experimental support for GCN 2nd (Bonaire, Hawaii, Kaveri, …),
NO support for GCN 1st (Oland, Cape Verde, Pitcairn, Tahiti) 		Ubuntu 15.10 "Willy Werewolf", Fedora 23, … Debian "Testing",
Arch Linux,
Ubuntu ppa, …
Old version, no longer maintained: 4.3 2015-11-01 4.3 diff 4.3/4.2 pull drm-next-4.3 Support for Fiji, Added drm/amd/scheduler
Older version, yet still maintained: 4.4 LTS 2016-01-10 4.4 diff 4.4/4.3 pull drm-next-4.4 Support for Stoney, enable scheduler by default, Ubuntu 16.04 LTS "Xenial Xerus"
Old version, no longer maintained: 4.5 2016-03-13 4.5 diff 4.5/4.4 pull drm-next-4.5 Added drm/amd/powertune-modules for tonga, fiji, carrizo, stoney Fedora 24, …
Old version, no longer maintained: 4.6 2016-05-15 4.6 diff 4.6/4.5 pull drm-next-4.6
Old version, no longer maintained: 4.7 2016-07-24 4.7 diff 4.7/4.6 pull drm-next-4.7 Support for GCN 4th (Polaris 10, Polaris 11), current display stack on par with other ASICs;
for advanced features DAL (Display Abstraction Layer) is required
Old version, no longer maintained: 4.8 2016-09-25 4.8 diff 4.8/4.7 pull drm-next-4.8 Over 20% of the patch is documentation updates, due to conversion of drm and media documentation from Docbook to the Sphinx doc format.[5]
https://www.kernel.org/doc/html/latest/ 		Ubuntu 16.10 "Yakkety Yak", Fedora 25, Ubuntu 16.04.2 LTS with HWE (hardware enablement stack)
Older version, yet still maintained: 4.9 LTS 2016-12-11 4.9 diff 4.9/4.8 pull drm-next-4.9
pull drm-next-4.9 		Experimental support GCN 1st (Oland, Cape Verde, Pitcairn, Tahiti), SI support for amdgpu Debian 9 "Stretch", SteamOS 31, …
Old version, no longer maintained: 4.10 2017-02-19 4.10 diff 4.10/4.9 2016-Oct-27
2016-Nov-23
2016-Dec-08 		Moving towards full support for GCN 1st
explicit fencing mainlined lwn.net 		Ubuntu 17.04 "Zesty Zapus", …
Old version, no longer maintained: 4.11 2017-04-30 4.11 diff 4.11/4.10 2017-Jan-27
2017-Feb-09
2017-Feb-16
2017-Feb-23 		… (please look it up)
Old version, no longer maintained: 4.12 2017-07-02 4.12 diff 4.12/4.11 Support for GCN 5th (Vega 10),
Old version, no longer maintained: 4.13 2017-09-03 4.13 diff 4.13/4.12 Fedora 27, Ubuntu 17.10 "Artful Aardvark"
Older version, yet still maintained: 4.14 LTS 2017-11-12 4.14 diff 4.14/4.13
Old version, no longer maintained: 4.15 2018-01-28 4.15 diff 4.15/4.14 2017-Sep-27 The driver for the display controller was (finally) mainlined: drm/amd/display Ubuntu 18.04 LTS "Bionic Beaver"
Old version, no longer maintained: 4.16 2018-04-01 4.16 diff 4.16/4.15 Fedora 28, …
Old version, no longer maintained: 4.17 2018-06-03 4.17 diff 4.17/4.16
Old version, no longer maintained: 4.18 2018-08-12 4.18 diff 4.18/4.17 2018-05-24
2018-05-31 		Vega20 support, clock and power gating for VCN, … Ubuntu 18.10 "Cosmic Cuttlefish", Fedora 29, …
Older version, yet still maintained: 4.19 LTS 2018-10-22 4.19 diff 4.19/4.18 2018-06-21 add initial amdgpu documentation, add initial GPU scheduler documentation, add support for the JPEG engine on Video Core Next (VCN), …
Old version, no longer maintained: 4.20 2018-12-23 4.20 diff 4.20/4.19 2018-06-21
2018-09-14
Old version, no longer maintained: 5.0 2019-03-03 5.0 diff 5.0/4.20 2018-12-07
2012-12-12
2012-12-29 		Ubuntu 19.04 "Disco Dingo", Fedora 30, …
Old version, no longer maintained: 5.1 2019-05-05 5.1 diff 5.1/5.0 2019-01-25
2019-02-01
2019-02-08
Current stable version: 5.2 2019-07-07 5.2 diff 5.2/5.1
Latest preview version of a future release: 5.3 2019-09-??
Legend:
Old version
Older version, still maintained
Latest version
Latest preview version
Future release

1 SteamOS is based on a Debian stable release though a newer Linux kernel is later back-ported. Current version of SteamOS is at Linux kernel 4.1. It seems likely that SteamOS 3 will be based on Debian 9.

Other half edit

amdgpu is only "half" of the graphics driver stack. Other halfs are Mesa and AMDGPU-PRO. AMDGPU-PRO Driver 17.30 (released 2017-07-27) is available for:

  • RHEL 6.9 / CentOS 6.9 (Linux 2.6.32)
  • RHEL 7.3 / CentOS 7.3 (Linux 3.10)
  • SLED/SLES 12 SP2
  • Ubuntu 16.04.2 (Linux 4.4, but also 4.8 HWE until 16.04.3 scheduled for August 2017)

Wild guesses edit

Deus Ex: Mankind Divided (2016-aug) – using the "Dawn Engine", reportedly based on the Glacier 2 game engine, which itself was used in Hitman: Absolution (2012) – scored significantly higher frame rates on the 4.11-drm-next code. Why? It's would be interesting to compare the two git-branches and see what kind of mechanics achieve twice the performance!

Explanations edit

 
Generic block diagram of a GPU
code name abbr. version number fab concrete chips/dies
"southern islands" si GCN 1 GCN1st GCN 1.0 28 nm Oland, Cape Verde, Pitcairn, Tahiti
"sea islands" ci GCN 2 GCN2nd GCN 1.1 Bonaire, Hawaii; cik refers to the GPU on some of the APUs
"volcanic islands" vi GCN 3 GCN3rd GCN 1.2 Tonga, Fiji; uses GCA (Graphics and Compute Array) version 8.0
"arctic islands" ai GCN 4 GCN4th GCN 1.3 14 nm Polaris 10/11/12; uses GCA (Graphics and Compute Array) version 8.0
"vega" vega GCN 5 GCN5th GCN 2.0 Vega 10/11

There are three different Device Dependent X (DDX)-drivers available:

  • xf86-video-ati = the old proprietary DDX driver; works on top of fglrx.ko in parallel with the proprietary OpenGL driver.
  • xf86-video-radeon = the FOSS DDX driver in Mesa; works on top of /drm/radeon
  • xf86-video-amdgpu = the FOSS DDX driver by AMD; works in top of /drm/amd/amdgpu in parallel with proprietary OpenGL driver and with Mesa; uses Glamor!

The interface libraries for amdgpu are libdrm_generic and libdrm_amdgpu, cf. https://lwn.net/Articles/654542/.

cf. Free_and_open-source_graphics_device_driver#ATI.2FAMD, Graphics Core Next, commons:Category:Diagrams illustrating AMD technology

 
The System Management Unit (SMU) handles power management (and other tasks).

For more statistics see e.g. heise.de and of course the man page for git.

 
3 Rings.

Commands are sent to the GPU by putting them in rings:

  • Graphics ring
  • Compute rings
  • DMA rings (seem to be only GCN 1.1 aka GCN 2nd aka Bonaire, Hawaii and newer)

Commands are processed by the GPU Command Processor. It contains multiple sub-units (ME (Micro Engine), PFP (Prefetch Parser), CE (Constant Engine), DE (Dispatch Engine)), each of which is a custom ‘F32’ CPU running microcode firmware. Rings can call out to Indirect Buffers (IBs) with more commands

GCN 1:

All resource descriptors (sometimes called “fetch constants” on previous hardware) are read from memory instead of registers. In order to improve performance, the CP block adds a new constant engine. The constant engine (CE) runs in parallel to the 3D engine, and allows constants to be written to memory ahead of the main command stream.

Updating Shader Resource Descriptors (SRDs) using the CPU would be too slow. The CP provides a constant update engine to accelerate the process.

PM4 is the packet API used to program the GPU to perform a variety of tasks. The driver does not write directly to the GPU registers to carry out drawing operations on the screen. Instead, it prepares data in the format of "PM4 Command Packets" in either system or video (a.k.a. local) memory, and lets the Micro Engine to do the rest of the job.

There are two CP engines on SI: the Constant Engine (CE) and the Drawing Engine (DE). Previous ASICs only had a DE (previously referred to as the Micro-Engine or ME). Certain packets are only allowed on the DE or the CE. Packets can be executed in the DE via the ring or via an INDIRECT_BUFFER packet. Packets can only be executed on the CE by using the INDIRECT_BUFFER_CONST packet.

  • Draw Engine (DE): The standard graphics engine is now referred to as the Draw Engine. Most PM4 commands continue to be submitted to the DE command buffer.
  • Constant Engine (CE): The constant engine uses a second, separate command buffer to control constant uploads. The engine runs in parallel with the draw engine, allowing constant updates to get sufficiently ahead of the draws/dispatches that will use them. Additionally, CP has 64KiB of on-chip RAM (CE RAM) that acts as a staging buffer for constant updates. Shaders cannot read directly from the CE RAM.

The 64KiB is carved up between the 3 rings, with ring-0 (gfx) having 32KB. The driver is responsible for further subdividing the partitions to store an on-chip copy of the most up-to-date copy of every SRD.

There are two CP engines on GCN 1.0 (SI, Southern Island): the Constant Engine (CE) and the Drawing Engine (DE). Previous ASICs only had a DE (previously referred to as the Micro-Engine or ME).

The purpose of this packet is to provide a generic and flexible way for the CP to write n Dwords of data to any destination to which it has access. As applicable, the writes can be sent from the CE, PFP, ME or DE (Dispatch Engine). The CE (and PFP) are limited to the GRBM (Graphics Register Backbone Manager) and MC (memory controller?) as destinations.

Note: The split between SRBM and GRBM is mostly between "3D engine" and "everything else" – register writes to the 3D engine need to be synchronized with the data flowing through the engine so the register logic is *really* complicated. The rest of the registers don't need to be pipelined and so are managed by a separate block. Most graphics/3D registers sit behind a FIFO, while other (display/asic setup/config) registers do not.


CB Color Block Color buffer
CP Command Processor aka Graphics Command Processor
DB Depth Block Depth buffer
EE Event Engine ?
GRBM Graphics Register Block Manager Registers (0x8000-0x2fffc)?
ME Micro Engine (the things that run the CP/PFP microcode)
PA Primitive Assembler Scissors, Cliprects, Point sprites
SE Shader Engine What Compute Units (CUs) are grouped into. Both Hawaii (44 CUs) and Fiji (64 CUs) have 4. Other have 2 or so.
SPI Shader Processor Interpolator Semantic mapping, interpolators
SQ Sequencer / Shader Shader management, exports, register alloc
SRBM System Register Block Manager Registers (0x0000-0x7ffc)?
SX Shader Export Output mapping, output buffers
TA Texture Addresser ?
VGT Vertex Grouper Tessellator Everthing related to vertex setup, viewport transformation, etc.

Hardware edit

Note: The following tables refer always to the most current stable Linux kernel version, i.e. support for Polaris is available only since Linux kernel 4.7 and not present in earlier versions!

The following table shows features of AMD/ATI's GPUs (see also: List of AMD graphics processing units).

[ VisualEditor ]
Name of GPU series Wonder Mach 3D Rage Rage Pro Rage 128 R100 R200 R300 R400 R500 R600 RV670 R700 Evergreen Northern
Islands 		Southern
Islands 		Sea
Islands 		Volcanic
Islands 		Arctic
Islands/Polaris 		Vega Navi 1x Navi 2x Navi 3x
Released 1986 1991 Apr
1996 		Mar
1997 		Aug
1998 		Apr
2000 		Aug
2001 		Sep
2002 		May
2004 		Oct
2005 		May
2007 		Nov
2007 		Jun
2008 		Sep
2009 		Oct
2010 		Jan
2012 		Sep
2013 		Jun
2015 		Jun 2016, Apr 2017, Aug 2019 Jun 2017, Feb 2019 Jul
2019 		Nov
2020 		Dec
2022
Marketing Name Wonder Mach 3D
Rage 		Rage
Pro 		Rage
128 		Radeon
7000 		Radeon
8000 		Radeon
9000 		Radeon
X700/X800 		Radeon
X1000 		Radeon
HD 2000 		Radeon
HD 3000 		Radeon
HD 4000 		Radeon
HD 5000 		Radeon
HD 6000 		Radeon
HD 7000 		Radeon
200 		Radeon
300 		Radeon
400/500/600 		Radeon
RX Vega, Radeon VII 		Radeon
RX 5000 		Radeon
RX 6000 		Radeon
RX 7000
AMD support    
Kind 2D 3D
Instruction set architecture Not publicly known TeraScale instruction set GCN instruction set RDNA instruction set
Microarchitecture TeraScale 1
(VLIW) 		TeraScale 2
(VLIW5)
TeraScale 2
(VLIW5)
up to 68xx		 TeraScale 3
(VLIW4)
in 69xx [7][8]
GCN 1st
gen 		GCN 2nd
gen 		GCN 3rd
gen 		GCN 4th
gen 		GCN 5th
gen 		RDNA RDNA 2 RDNA 3
Type Fixed pipeline[a] Programmable pixel & vertex pipelines Unified shader model
Direct3D — 5.0 6.0 7.0 8.1 9.0
11 (9_2) 		9.0b
11 (9_2) 		9.0c
11 (9_3) 		10.0
11 (10_0) 		10.1
11 (10_1) 		11 (11_0) 11 (11_1)
12 (11_1) 		11 (12_0)
12 (12_0) 		11 (12_1)
12 (12_1) 		11 (12_1)
12 (12_2)
Shader model — 1.4 2.0+ 2.0b 3.0 4.0 4.1 5.0 5.1 5.1
6.5 		6.7
OpenGL — 1.1 1.2 1.3 2.1[b][9] 3.3 4.5 (on Linux: 4.5 (Mesa 3D 21.0))[10][11][12][c] 4.6 (on Linux: 4.6 (Mesa 3D 20.0))
Vulkan — 1.0
(Win 7+ or Mesa 17+) 		1.2 (Adrenalin 20.1.2, Linux Mesa 3D 20.0)
1.3 (GCN 4 and above (with Adrenalin 22.1.2, Mesa 22.0)) 		1.3
OpenCL — Close to Metal 1.1 (no Mesa 3D support) 1.2+ (on Linux: 1.1+ (no Image support on clover, with by rustiCL) with Mesa 3D, 1.2+ on GCN 1.Gen) 2.0+ (Adrenalin driver on Win7+)
(on Linux ROCM, Linux Mesa 3D 1.2+ (no Image support in clover, but in rustiCL with Mesa 3D, 2.0+ and 3.0 with AMD drivers or AMD ROCm), 5th gen: 2.2 win 10+ and Linux RocM 5.0+ 		2.2+ and 3.0 windows 8.1+ and Linux ROCM 5.0+ (Mesa 3D rustiCL 1.2+ and 3.0 (2.1+ and 2.2+ wip))[13][14][15]
HSA / ROCm —   ?
Video decoding ASIC — Avivo/UVD UVD+ UVD 2 UVD 2.2 UVD 3 UVD 4 UVD 4.2 UVD 5.0 or 6.0 UVD 6.3 UVD 7 [16][d] VCN 2.0 [16][d] VCN 3.0 [17] VCN 4.0
Video encoding ASIC — VCE 1.0 VCE 2.0 VCE 3.0 or 3.1 VCE 3.4 VCE 4.0 [16][d]
Fluid Motion [e]       ?
Power saving ? PowerPlay PowerTune PowerTune & ZeroCore Power ?
TrueAudio — Via dedicated DSP Via shaders
FreeSync — 1
2
HDCP[f] ? 1.4 2.2 2.3 [18]
PlayReady[f] — 3.0   3.0
Supported displays[g] 1–2 2 2–6 ?
Max. resolution ? 2–6 ×
2560×1600 		2–6 ×
4096×2160 @ 30 Hz 		2–6 ×
5120×2880 @ 60 Hz 		3 ×
7680×4320 @ 60 Hz [19]
7680×4320 @ 60 Hz PowerColor 		7680x4320

@165 HZ

/drm/radeon[h]   —
/drm/amdgpu[h] — Experimental [20] Optional [21]  
  1. ^ The Radeon 100 Series has programmable pixel shaders, but do not fully comply with DirectX 8 or Pixel Shader 1.0. See article on R100's pixel shaders.
  2. ^ R300, R400 and R500 based cards do not fully comply with OpenGL 2+ as the hardware does not support all types of non-power of two (NPOT) textures.
  3. ^ OpenGL 4+ compliance requires supporting FP64 shaders and these are emulated on some TeraScale chips using 32-bit hardware.
  4. ^ a b c The UVD and VCE were replaced by the Video Core Next (VCN) ASIC in the Raven Ridge APU implementation of Vega.
  5. ^ Video processing for video frame rate interpolation technique. In Windows it works as a DirectShow filter in your player. In Linux, there is no support on the part of drivers and / or community.
  6. ^ a b To play protected video content, it also requires card, operating system, driver, and application support. A compatible HDCP display is also needed for this. HDCP is mandatory for the output of certain audio formats, placing additional constraints on the multimedia setup.
  7. ^ More displays may be supported with native DisplayPort connections, or splitting the maximum resolution between multiple monitors with active converters.
  8. ^ a b DRM (Direct Rendering Manager) is a component of the Linux kernel. AMDgpu is the Linux kernel module. Support in this table refers to the most current version.

The following table shows features of AMD's processors with 3D graphics, including APUs (see also: List of AMD processors with 3D graphics).

[ VisualEditor ]
Platform High, standard and low power Low and ultra-low power
Codename Server Basic Toronto
Micro Kyoto
Desktop Performance Raphael Phoenix
Mainstream Llano Trinity Richland Kaveri Kaveri Refresh (Godavari) Carrizo Bristol Ridge Raven Ridge Picasso Renoir Cezanne
Entry
Basic Kabini Dalí
Mobile Performance Renoir Cezanne Rembrandt Dragon Range
Mainstream Llano Trinity Richland Kaveri Carrizo Bristol Ridge Raven Ridge Picasso Renoir
Lucienne 		Cezanne
Barceló 		Phoenix
Entry Dalí Mendocino
Basic Desna, Ontario, Zacate Kabini, Temash Beema, Mullins Carrizo-L Stoney Ridge Pollock
Embedded Trinity Bald Eagle Merlin Falcon,
Brown Falcon 		Great Horned Owl Grey Hawk Ontario, Zacate Kabini Steppe Eagle, Crowned Eagle,
LX-Family 		Prairie Falcon Banded Kestrel River Hawk
Released Aug 2011 Oct 2012 Jun 2013 Jan 2014 2015 Jun 2015 Jun 2016 Oct 2017 Jan 2019 Mar 2020 Jan 2021 Jan 2022 Sep 2022 Jan 2023 Jan 2011 May 2013 Apr 2014 May 2015 Feb 2016 Apr 2019 Jul 2020 Jun 2022 Nov 2022
CPU microarchitecture K10 Piledriver Steamroller Excavator "Excavator+"[22] Zen Zen+ Zen 2 Zen 3 Zen 3+ Zen 4 Bobcat Jaguar Puma Puma+[23] "Excavator+" Zen Zen+ "Zen 2+"
ISA x86-64 v1 x86-64 v2 x86-64 v3 x86-64 v4 x86-64 v1 x86-64 v2 x86-64 v3
Socket Desktop Performance — AM5 — —
Mainstream — AM4 — —
Entry FM1 FM2 FM2+ FM2+[a], AM4 AM4 —
Basic — — AM1 — FP5 —
Other FS1 FS1+, FP2 FP3 FP4 FP5 FP6 FP7 FL1 FP7
FP7r2
FP8 		? FT1 FT3 FT3b FP4 FP5 FT5 FP5 FT6
PCI Express version 2.0 3.0 4.0 5.0 4.0 2.0 3.0
CXL — —
Fab. (nm) GF 32SHP
(HKMG SOI) 		GF 28SHP
(HKMG bulk) 		GF 14LPP
(FinFET bulk) 		GF 12LP
(FinFET bulk) 		TSMC N7
(FinFET bulk) 		TSMC N6
(FinFET bulk) 		CCD: TSMC N5
(FinFET bulk)
cIOD: TSMC N6
(FinFET bulk) 		TSMC 4nm
(FinFET bulk) 		TSMC N40
(bulk) 		TSMC N28
(HKMG bulk) 		GF 28SHP
(HKMG bulk) 		GF 14LPP
(FinFET bulk) 		GF 12LP
(FinFET bulk) 		TSMC N6
(FinFET bulk)
Die area (mm2) 228 246 245 245 250 210[24] 156 180 210 CCD: (2x) 70
cIOD: 122 		178 75 (+ 28 FCH) 107 ? 125 149 ~100
Min TDP (W) 35 17 12 10 15 105 35 4.5 4 3.95 10 6 12 8
Max APU TDP (W) 100 95 65 45 170 54 18 25 6 54 15
Max stock APU base clock (GHz) 3 3.8 4.1 4.1 3.7 3.8 3.6 3.7 3.8 4.0 3.3 4.7 4.3 1.75 2.2 2 2.2 3.2 2.6 1.2 3.35 2.8
Max APUs per node[b] 1 1
Max core dies per CPU 1 2 1 1
Max CCX per core die 1 2 1 1
Max cores per CCX 4 8 2 4 2 4
Max CPU[c] cores per APU 4 8 16 8 2 4 2 4
Max threads per CPU core 1 2 1 2
Integer pipeline structure 3+3 2+2 4+2 4+2+1 1+3+3+1+2 1+1+1+1 2+2 4+2 4+2+1
i386, i486, i586, CMOV, NOPL, i686, PAE, NX bit, CMPXCHG16B, AMD-V, RVI, ABM, and 64-bit LAHF/SAHF    
IOMMU[d] — v2 v1 v2
BMI1, AES-NI, CLMUL, and F16C   —  
MOVBE —  
AVIC, BMI2, RDRAND, and MWAITX/MONITORX —  
SME[e], TSME[e], ADX, SHA, RDSEED, SMAP, SMEP, XSAVEC, XSAVES, XRSTORS, CLFLUSHOPT, CLZERO, and PTE Coalescing —   —  
GMET, WBNOINVD, CLWB, QOS, PQE-BW, RDPID, RDPRU, and MCOMMIT —   —  
MPK, VAES —   —
SGX — —
FPUs per core 1 0.5 1 1 0.5 1
Pipes per FPU 2 2
FPU pipe width 128-bit 256-bit 80-bit 128-bit 256-bit
CPU instruction set SIMD level SSE4a[f] AVX AVX2 AVX-512 SSSE3 AVX AVX2
3DNow! 3DNow!+ — —
PREFETCH/PREFETCHW    
GFNI —   —
AMX —
FMA4, LWP, TBM, and XOP —   — —   —
FMA3    
AMD XDNA —   —
L1 data cache per core (KiB) 64 16 32 32
L1 data cache associativity (ways) 2 4 8 8
L1 instruction caches per core 1 0.5 1 1 0.5 1
Max APU total L1 instruction cache (KiB) 256 128 192 256 512 256 64 128 96 128
L1 instruction cache associativity (ways) 2 3 4 8 2 3 4 8
L2 caches per core 1 0.5 1 1 0.5 1
Max APU total L2 cache (MiB) 4 2 4 16 1 2 1 2
L2 cache associativity (ways) 16 8 16 8
Max on--die L3 cache per CCX (MiB) — 4 16 32 — 4
Max 3D V-Cache per CCD (MiB) — 64 — —
Max total in-CCD L3 cache per APU (MiB) 4 8 16 64 4
Max. total 3D V-Cache per APU (MiB) — 64 — —
Max. board L3 cache per APU (MiB) — —
Max total L3 cache per APU (MiB) 4 8 16 128 4
APU L3 cache associativity (ways) 16 16
L3 cache scheme Victim Victim
Max. L4 cache — —
Max stock DRAM support DDR3-1866 DDR3-2133 DDR3-2133, DDR4-2400 DDR4-2400 DDR4-2933 DDR4-3200, LPDDR4-4266 DDR5-4800, LPDDR5-6400 DDR5-5200 DDR5-5600, LPDDR5x-7500 DDR3L-1333 DDR3L-1600 DDR3L-1866 DDR3-1866, DDR4-2400 DDR4-2400 DDR4-1600 DDR4-3200 LPDDR5-5500
Max DRAM channels per APU 2 1 2 1 2
Max stock DRAM bandwidth (GB/s) per APU 29.866 34.132 38.400 46.932 68.256 102.400 83.200 120.000 10.666 12.800 14.933 19.200 38.400 12.800 51.200 88.000
GPU microarchitecture TeraScale 2 (VLIW5) TeraScale 3 (VLIW4) GCN 2nd gen GCN 3rd gen GCN 5th gen[25] RDNA 2 RDNA 3 TeraScale 2 (VLIW5) GCN 2nd gen GCN 3rd gen[25] GCN 5th gen RDNA 2
GPU instruction set TeraScale instruction set GCN instruction set RDNA instruction set TeraScale instruction set GCN instruction set RDNA instruction set
Max stock GPU base clock (MHz) 600 800 844 866 1108 1250 1400 2100 2400 400 538 600 ? 847 900 1200 600 1300 1900
Max stock GPU base GFLOPS[g] 480 614.4 648.1 886.7 1134.5 1760 1971.2 2150.4 3686.4 102.4 86 ? ? ? 345.6 460.8 230.4 1331.2 486.4
3D engine[h] Up to 400:20:8 Up to 384:24:6 Up to 512:32:8 Up to 704:44:16[26] Up to 512:32:8 768:48:8 128:8:4 80:8:4 128:8:4 Up to 192:12:8 Up to 192:12:4 192:12:4 Up to 512:?:? 128:?:?
IOMMUv1 IOMMUv2 IOMMUv1 ? IOMMUv2
Video decoder UVD 3.0 UVD 4.2 UVD 6.0 VCN 1.0[27] VCN 2.1[28] VCN 2.2[28] VCN 3.1 ? UVD 3.0 UVD 4.0 UVD 4.2 UVD 6.0 UVD 6.3 VCN 1.0 VCN 3.1
Video encoder — VCE 1.0 VCE 2.0 VCE 3.1 — VCE 2.0 VCE 3.1
AMD Fluid Motion            
GPU power saving PowerPlay PowerTune PowerPlay PowerTune[29]
TrueAudio —  [30] ? —  
FreeSync 1
2 		1
2
HDCP[i] ? 1.4 2.2 2.3 ? 1.4 2.2 2.3
PlayReady[i] — 3.0 not yet — 3.0 not yet
Supported displays[j] 2–3 2–4 3 3 (desktop)
4 (mobile, embedded)		 4 2 3 4 4
/drm/radeon[k][32][33]   —   —
/drm/amdgpu[k][34] —  [35] —  [35]
  1. ^ For FM2+ Excavator models: A8-7680, A6-7480 & Athlon X4 845.
  2. ^ A PC would be one node.
  3. ^ An APU combines a CPU and a GPU. Both have cores.
  4. ^ Requires firmware support.
  5. ^ a b Requires firmware support.
  6. ^ No SSE4. No SSSE3.
  7. ^ Single-precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.
  8. ^ Unified shaders : texture mapping units : render output units
  9. ^ a b To play protected video content, it also requires card, operating system, driver, and application support. A compatible HDCP display is also needed for this. HDCP is mandatory for the output of certain audio formats, placing additional constraints on the multimedia setup.
  10. ^ To feed more than two displays, the additional panels must have native DisplayPort support.[31] Alternatively active DisplayPort-to-DVI/HDMI/VGA adapters can be employed.
  11. ^ a b DRM (Direct Rendering Manager) is a component of the Linux kernel. Support in this table refers to the most current version.
  1. ^ "RADV conforms to Vulkan 1.0". 2017-10-04.
  2. ^ "Alex Deucher - AMD's New Unified Open Source Driver". 2014-10-07.
  3. ^ "Initial amdgpu driver release". 2015-04-20.
  4. ^ "update on amdgpu and future". 2015-09-17.
  5. ^ Torvalds, Linus (2016-08-07). "Linux 4.8-rc1".
  6. ^ "No support for GCN 1.0 hardware, e.g. Tahiti XT". 2018-01-29.
  7. ^ "AMD Radeon HD 6900 (AMD Cayman) series graphics cards". HWlab. hw-lab.com. December 19, 2010. Archived from the original on August 23, 2022. Retrieved August 23, 2022. New VLIW4 architecture of stream processors allowed to save area of each SIMD by 10%, while performing the same compared to previous VLIW5 architecture
  8. ^ "GPU Specs Database". TechPowerUp. Retrieved August 23, 2022.
  9. ^ "NPOT Texture (OpenGL Wiki)". Khronos Group. Retrieved February 10, 2021.
  10. ^ "AMD Radeon Software Crimson Edition Beta". AMD. Retrieved 2018-04-20.
  11. ^ "Mesamatrix". mesamatrix.net. Retrieved 2018-04-22.
  12. ^ "RadeonFeature". X.Org Foundation. Retrieved 2018-04-20.
  13. ^ "AMD Radeon RX 6800 XT Specs". TechPowerUp. Retrieved January 1, 2021.
  14. ^ "AMD Launches The Radeon PRO W7500/W7600 RDNA3 GPUs". Phoronix. 3 August 2023. Retrieved 4 September 2023.
  15. ^ "AMD Radeon Pro 5600M Grafikkarte". TopCPU.net (in German). Retrieved 4 September 2023.
  16. ^ a b c Killian, Zak (March 22, 2017). "AMD publishes patches for Vega support on Linux". Tech Report. Retrieved March 23, 2017.
  17. ^ Larabel, Michael (September 15, 2020). "AMD Radeon Navi 2 / VCN 3.0 Supports AV1 Video Decoding". Phoronix. Retrieved January 1, 2021.
  18. ^ Edmonds, Rich (February 4, 2022). "ASUS Dual RX 6600 GPU review: Rock-solid 1080p gaming with impressive thermals". Windows Central. Retrieved November 1, 2022.
  19. ^ "Radeon's next-generation Vega architecture" (PDF). Radeon Technologies Group (AMD). Archived from the original (PDF) on September 6, 2018. Retrieved June 13, 2017.
  20. ^ Larabel, Michael (December 7, 2016). "The Best Features of the Linux 4.9 Kernel". Phoronix. Retrieved December 7, 2016.
  21. ^ "AMDGPU". Retrieved December 29, 2023.
  22. ^ "AMD Announces the 7th Generation APU: Excavator mk2 in Bristol Ridge and Stoney Ridge for Notebooks". 31 May 2016. Retrieved 3 January 2020.
  23. ^ "AMD Mobile "Carrizo" Family of APUs Designed to Deliver Significant Leap in Performance, Energy Efficiency in 2015" (Press release). 20 November 2014. Retrieved 16 February 2015.
  24. ^ "The Mobile CPU Comparison Guide Rev. 13.0 Page 5 : AMD Mobile CPU Full List". TechARP.com. Retrieved 13 December 2017.
  25. ^ a b "AMD VEGA10 and VEGA11 GPUs spotted in OpenCL driver". VideoCardz.com. Retrieved 6 June 2017.
  26. ^ Cutress, Ian (1 February 2018). "Zen Cores and Vega: Ryzen APUs for AM4 – AMD Tech Day at CES: 2018 Roadmap Revealed, with Ryzen APUs, Zen+ on 12nm, Vega on 7nm". Anandtech. Retrieved 7 February 2018.
  27. ^ Larabel, Michael (17 November 2017). "Radeon VCN Encode Support Lands in Mesa 17.4 Git". Phoronix. Retrieved 20 November 2017.
  28. ^ a b "AMD Ryzen 5000G 'Cezanne' APU Gets First High-Res Die Shots, 10.7 Billion Transistors In A 180mm2 Package". wccftech. Aug 12, 2021. Retrieved August 25, 2021.
  29. ^ Tony Chen; Jason Greaves, "AMD's Graphics Core Next (GCN) Architecture" (PDF), AMD, retrieved 13 August 2016
  30. ^ "A technical look at AMD's Kaveri architecture". Semi Accurate. Retrieved 6 July 2014.
  31. ^ "How do I connect three or More Monitors to an AMD Radeon™ HD 5000, HD 6000, and HD 7000 Series Graphics Card?". AMD. Retrieved 8 December 2014.
  32. ^ Airlie, David (26 November 2009). "DisplayPort supported by KMS driver mainlined into Linux kernel 2.6.33". Retrieved 16 January 2016.
  33. ^ "Radeon feature matrix". freedesktop.org. Retrieved 10 January 2016.
  34. ^ Deucher, Alexander (16 September 2015). "XDC2015: AMDGPU" (PDF). Retrieved 16 January 2016.
  35. ^ a b Michel Dänzer (17 November 2016). "[ANNOUNCE] xf86-video-amdgpu 1.2.0". lists.x.org.
Retrieved from "https://en.wikipedia.org/w/index.php?title=User:ScotXW/amdgpu&oldid=905932007"