How The Radeon RX 5700 XT Navi Linux Performance Has Evolved Since Launch

[Mechanix]

Thanks, but you are doing what people always do when I complain about some sorry state of affairs: try to fix my particular issue. It comes off as "If you get fries for free, will you stop complaining about the spoiled meat?". I bet you mean well though.

It is not EAC - when I put the old GTX970 back in, the framerate returned. And I am not looking for solutions. I am stating that there is a wide and undeserved praise for Radeons around here and on many other Linux forums - for apparent political reasons as everyone carries the open-source badge with pride - and I feel lied to. The performance is mostly great, but there seems to be a number of issues that are well known but swept under the rug.

I am not a fanboy, I have spent as much money on AMD as on Intel chips for the last 25 years, and I was just about to drop my GeForce for a Radeon because of the glowing reviews. I have had issues with ATI/AMD Linux drivers many years before, but figured that things had straightened out what with the fierce advocacy I have seen lately. I think I was wrong.

[moilami]

Something is not supported in War Thunder + RX5700, hence abysmal performance. It cannot merely be a matter of code quality if you get one fifth the expected performance, it must be a matter of something being missing altogether.

I have almost mostly used nVidia cards since, well, Riva TNT. And I remember a time when I had to manually enter resolution scan frequencies for CRTs in the X settings or whatever. And then somewhere early along the timeline that disappeared and graphics was not an issue anymore. Install the nVidia drivers and off you went.

Roll time forward to 2019, and Navi appears and looks good. And people like y'all speak at length about how fantastic AMD drivers are these days. Forums such as this are replete with posts how things "got better recently", usually amended with some genuflection about how open source is the only way, just like you just did there.

When you reveal a political bias like that, I can't take your arguments at face value. You are apparently partisan and would lie to me. Because AMD drivers are still crap from all I can tell after having taken a deep dive into the issue.

If you scrutinize benchmark results, you quite often find entries missing - either without explanation or at best with some footnote mumbling about problems, often with a "we have reached out to AMD about this issue".

Sometimes the notes that are post-fix. There was a problem, it was patched up during the review, there is no problem now, swear to god.

Sometimes there are subsequent reviews that pitch former ones in new light. Card X is released. It gets glowing praise. Halfyear passes. New article appears basically saying "Remember how X had issues? It has been improved recently!" Halfyear passes. New article: "X2 released, replacing shitty X". That kind of journalism betrays an intense desire to lift X by the armpits, or in other words: dishonesty.

It is possible that the open source drivers of AMD may become better than nVidias closed source drivers. As it stands now, this does not seem to be the case by far. And I feel I have been lied to by fanboys.

I am not a fanboy. I used to recommend nvidia for people who do not care of free software and who do demanding gaming. That was years ago because drivers for AMD has developed to the point that I see no reason to recommend nvidia to anyone anymore. The only reason to recommend nvidia now would be if someone want to play War Thunder and must get this week more than 30 FPS, but even before that I would test the performance myself first to make sure it really is less than 30 FPS. It is a matter of time, by the way, when the performance will be better, if it really is so low now.

I do, however, greatly appreciate the work AMD community has done, and even more so because I have been observing it for years. I get good feelings of illuminated AMD text on Wraith CPU cooler. Each time I see it reminds me of excellent underdog AMD community, which made great achievements and became the best.

[theboomboomcars]

It looks like OpenCL is working on Navi now.

clinfo
Number of platforms                               2
  Platform Name                                   Clover
  Platform Vendor                                 Mesa
  Platform Version                                OpenCL 1.1 Mesa 20.0.0-devel (git-680d806950)
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd
  Platform Extensions function suffix             MESA

  Platform Name                                   AMD Accelerated Parallel Processing
  Platform Vendor                                 Advanced Micro Devices, Inc.
  Platform Version                                OpenCL 2.1 AMD-APP (2906.7)
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd cl_amd_event_callback cl_amd_offline_devices
  Platform Host timer resolution                  1ns
  Platform Extensions function suffix             AMD

  Platform Name                                   Clover
Number of devices                                 1
  Device Name                                     AMD NAVI14 (DRM 3.36.0, 5.5.0-rc5-1-mainline, LLVM 9.0.1)
  Device Vendor                                   AMD
  Device Vendor ID                                0x1002
  Device Version                                  OpenCL 1.1 Mesa 20.0.0-devel (git-680d806950)
  Driver Version                                  20.0.0-devel
  Device OpenCL C Version                         OpenCL C 1.1
  Device Type                                     GPU
  Device Profile                                  FULL_PROFILE
  Device Available                                Yes
  Compiler Available                              Yes
  Max compute units                               22
  Max clock frequency                             1825MHz
  Max work item dimensions                        3
  Max work item sizes                             256x256x256
  Max work group size                             256
=== CL_PROGRAM_BUILD_LOG ===
fatal error: cannot open file '/usr/lib/clc/gfx1012-amdgcn-mesa-mesa3d.bc': No such file or directory
  Preferred work group size multiple              <getWGsizes:1200: create kernel : error -46>
  Preferred / native vector sizes                 
    char                                                16 / 16      
    short                                                8 / 8       
    int                                                  4 / 4       
    long                                                 2 / 2       
    half                                                 8 / 8        (cl_khr_fp16)
    float                                                4 / 4       
    double                                               2 / 2        (cl_khr_fp64)
  Half-precision Floating-point support           (cl_khr_fp16)
    Denormals                                     No
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
  Single-precision Floating-point support         (core)
    Denormals                                     No
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
  Address bits                                    64, Little-Endian
  Global memory size                              8589934592 (8GiB)
  Error Correction support                        No
  Max memory allocation                           6871947673 (6.4GiB)
  Unified memory for Host and Device              No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       32768 bits (4096 bytes)
  Global Memory cache type                        None
  Image support                                   No
  Local memory type                               Local
  Local memory size                               32768 (32KiB)
  Max number of constant args                     16
  Max constant buffer size                        2147483647 (2GiB)
  Max size of kernel argument                     1024
  Queue properties                                
    Out-of-order execution                        No
    Profiling                                     Yes
  Profiling timer resolution                      0ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            No
  Device Extensions                               cl_khr_byte_addressable_store cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_fp64 cl_khr_fp16

  Platform Name                                   AMD Accelerated Parallel Processing
Number of devices                                 1
  Device Name                                     gfx1012
  Device Vendor                                   Advanced Micro Devices, Inc.
  Device Vendor ID                                0x1002
  Device Version                                  OpenCL 2.0 AMD-APP (2906.7)
  Driver Version                                  2906.7 (PAL,LC)
  Device OpenCL C Version                         OpenCL C 2.0
  Device Type                                     GPU
  Device Board Name (AMD)                         Unknown AMD GPU
  Device Topology (AMD)                           PCI-E, 0b:00.0
  Device Profile                                  FULL_PROFILE
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Max compute units                               11
  SIMD per compute unit (AMD)                     2
  SIMD width (AMD)                                32
  SIMD instruction width (AMD)                    1
  Max clock frequency                             1825MHz
  Graphics IP (AMD)                               10.12
  Device Partition                                (core)
    Max number of sub-devices                     11
    Supported partition types                     None
    Supported affinity domains                    (n/a)
  Max work item dimensions                        3
  Max work item sizes                             1024x1024x1024
  Max work group size                             256
  Preferred work group size (AMD)                 256
  Max work group size (AMD)                       1024
  Preferred work group size multiple              32
  Wavefront width (AMD)                           32
  Preferred / native vector sizes                 
    char                                                 4 / 4       
    short                                                2 / 2       
    int                                                  1 / 1       
    long                                                 1 / 1       
    half                                                 1 / 1        (cl_khr_fp16)
    float                                                1 / 1       
    double                                               1 / 1        (cl_khr_fp64)
  Half-precision Floating-point support           (cl_khr_fp16)
    Denormals                                     No
    Infinity and NANs                             No
    Round to nearest                              No
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
  Single-precision Floating-point support         (core)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  Yes
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
  Address bits                                    64, Little-Endian
  Global memory size                              8573157376 (7.984GiB)
  Global free memory (AMD)                        8306688 (7.922GiB)
  Global memory channels (AMD)                    4
  Global memory banks per channel (AMD)           4
  Global memory bank width (AMD)                  256 bytes
  Error Correction support                        No
  Max memory allocation                           4244635648 (3.953GiB)
  Unified memory for Host and Device              No
  Shared Virtual Memory (SVM) capabilities        (core)
    Coarse-grained buffer sharing                 Yes
    Fine-grained buffer sharing                   Yes
    Fine-grained system sharing                   No
    Atomics                                       No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       2048 bits (256 bytes)
  Preferred alignment for atomics                 
    SVM                                           0 bytes
    Global                                        0 bytes
    Local                                         0 bytes
  Max size for global variable                    3820172032 (3.558GiB)
  Preferred total size of global vars             8573157376 (7.984GiB)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        16384 (16KiB)
  Global Memory cache line size                   64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             16
    Max size for 1D images from buffer            134217728 pixels
    Max 1D or 2D image array size                 2048 images
    Base address alignment for 2D image buffers   256 bytes
    Pitch alignment for 2D image buffers          256 pixels
    Max 2D image size                             16384x16384 pixels
    Max 3D image size                             2048x2048x2048 pixels
    Max number of read image args                 128
    Max number of write image args                64
    Max number of read/write image args           64
  Max number of pipe args                         16
  Max active pipe reservations                    16
  Max pipe packet size                            4244635648 (3.953GiB)
  Local memory type                               Local
  Local memory size                               65536 (64KiB)
  Local memory syze per CU (AMD)                  65536 (64KiB)
  Local memory banks (AMD)                        32
  Max number of constant args                     8
  Max constant buffer size                        4244635648 (3.953GiB)
  Preferred constant buffer size (AMD)            16384 (16KiB)
  Max size of kernel argument                     1024
  Queue properties (on host)                      
    Out-of-order execution                        No
    Profiling                                     Yes
  Queue properties (on device)                    
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Preferred size                                262144 (256KiB)
    Max size                                      8388608 (8MiB)
  Max queues on device                            1
  Max events on device                            1024
  Prefer user sync for interop                    Yes
  Number of P2P devices (AMD)                     0
  P2P devices (AMD)                               (n/a)
  Profiling timer resolution                      1ns
  Profiling timer offset since Epoch (AMD)        1578332205909545451ns (Mon Jan  6 10:36:45 2020)
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            No
    Thread trace supported (AMD)                  Yes
    Number of async queues (AMD)                  4
    Max real-time compute queues (AMD)            0
    Max real-time compute units (AMD)             0
  printf() buffer size                            4194304 (4MiB)
  Built-in kernels                                (n/a)
  Device Extensions                               cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_khr_gl_depth_images cl_amd_device_attribute_query cl_amd_media_ops cl_amd_media_ops2 cl_khr_image2d_from_buffer cl_khr_subgroups cl_khr_gl_event cl_khr_depth_images cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_amd_copy_buffer_p2p


NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  Clover
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   Success [MESA]
  clCreateContext(NULL, ...) [default]            Success [MESA]
  clCreateContext(NULL, ...) [other]              Success [AMD]
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT)  Success (1)
    Platform Name                                 Clover
    Device Name                                   AMD NAVI14 (DRM 3.36.0, 5.5.0-rc5-1-mainline, LLVM 9.0.1)
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  Success (1)
    Platform Name                                 Clover
    Device Name                                   AMD NAVI14 (DRM 3.36.0, 5.5.0-rc5-1-mainline, LLVM 9.0.1)
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  Success (1)
    Platform Name                                 Clover
    Device Name                                   AMD NAVI14 (DRM 3.36.0, 5.5.0-rc5-1-mainline, LLVM 9.0.1)

ICD loader properties
  ICD loader Name                                 OpenCL ICD Loader
  ICD loader Vendor                               OCL Icd free software
  ICD loader Version                              2.2.12
  ICD loader Profile                              OpenCL 2.2

[tenzero]

phoronix Hi Michael, I really appreciate the volume and depth of your Articles this last year. You have been outstanding.

Is there a reason for dropping the benchmarking of F1 2017/2018? I am curious as to how the 5700 XT performs at 1440p and 4k. (I'm thinking of an upgrade)
I've been going back through all of your GPU articles but can't find anything on F1 with the arrival of the 5700.

Cheers.
d.