Skip to content

Framework 13 - AMD Ryzen AI 5 340 #90

Open

Description

@geerlingguy

Image

Basic information

  • Board URL (official): https://frame.work/laptop13
  • Board purchased from: Provided for testing by Framework
  • Board purchase date: November 5, 2025
  • Board specs (as tested): Ryzen AI 5 340 (6-core/12-thread) / 16 GB RAM / 2.2K display / No Storage
  • Board price (as tested): $979

Linux/system information

# output of `screenfetch`
                          ./+o+-       jgeerling@framework-13
                  yyyyy- -yyyyyy+      OS: Ubuntu 25.04 plucky
               ://+//////-yyyyyyo      Kernel: x86_64 Linux 6.14.0-35-generic
           .++ .:/++++++/-.+sss/`      Uptime: 7m
         .:++o:  /++++++++/:--:/-      Packages: 1680
        o:+o+:++.`..```.-/oo+++++/     Shell: dash
       .:+o:+o/.          `+sssoo+/    Disk: 12G / 1.9T (1%)
  .++/+:+oo+o:`             /sssooo.   CPU: AMD Ryzen AI 5 340 w/ Radeon 840M @ 12x 4.9GHz
 /+++//+:`oo+o               /::--:.   GPU: 
 \+/+o+++`o++o               ++////.   RAM: 1652MiB / 14760MiB
  .++.o+++oo+:`             /dddhhh.  
       .+.o+oo:.          `oddhhhh+   
        \+.++o+o``-````.:ohdhhhhh+    
         `:o+++ `ohhhhhhhhyo++os:     
           .o:`.syhhhhhhh/.oo++o`     
               /osyyyyyyo++ooo+++/    
                   ````` +oo+++o\:    
                          `oo++.      

# output of `uname -a`
Linux framework-13 6.14.0-35-generic #35-Ubuntu SMP PREEMPT_DYNAMIC Sat Oct 11 10:06:31 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux

Benchmark results

CPU

Power

  • Idle power draw (at wall): 2.7 W (5W with screen on at 50% brightness)
  • Maximum simulated power draw (stress-ng --matrix 0): 46.7 W
  • During Geekbench multicore benchmark: 48 W
  • During top500 HPL benchmark: 41.6 W

Disk

KXG80ZNV2T04 KIOXIA

Benchmark Result
iozone 4K random read 71.25 MB/s
iozone 4K random write 343.17 MB/s
iozone 1M random read 4065.44 MB/s
iozone 1M random write 4676.18 MB/s
iozone 1M sequential read 4593.00 MB/s
iozone 1M sequential write 5209.89 MB/s

Network

iperf3 results:

Intel Corporation Wi-Fi 6 AX200

  • iperf3 -c $SERVER_IP: 6.40 Mbps
  • iperf3 -c $SERVER_IP --reverse: 114 Mbps
  • iperf3 -c $SERVER_IP --bidir: 5.14 Mbps up, 13.8 Mbps down

(Not sure why speeds were so low. I may need to pull the antennas out and see if there's an issue there, maybe swapping in an external antenna.)

GPU

glmark2

glmark2-es2 / glmark2-es2-wayland results:

=======================================================
    glmark2 2023.01
=======================================================
    OpenGL Information
    GL_VENDOR:      AMD
    GL_RENDERER:    AMD Radeon Graphics (radeonsi, gfx1152, LLVM 20.1.2, DRM 3.61, 6.14.0-35-generic)
    GL_VERSION:     OpenGL ES 3.2 Mesa 25.0.7-0ubuntu0.25.04.2
    Surface Config: buf=32 r=8 g=8 b=8 a=8 depth=24 stencil=0 samples=0
    Surface Size:   800x600 windowed
=======================================================
[build] use-vbo=false: FPS: 18606 FrameTime: 0.054 ms
[build] use-vbo=true: FPS: 29730 FrameTime: 0.034 ms
[texture] texture-filter=nearest: FPS: 22656 FrameTime: 0.044 ms
[texture] texture-filter=linear: FPS: 22592 FrameTime: 0.044 ms
[texture] texture-filter=mipmap: FPS: 23297 FrameTime: 0.043 ms
[shading] shading=gouraud: FPS: 23555 FrameTime: 0.042 ms
[shading] shading=blinn-phong-inf: FPS: 21084 FrameTime: 0.047 ms
[shading] shading=phong: FPS: 16872 FrameTime: 0.059 ms
[shading] shading=cel: FPS: 15514 FrameTime: 0.064 ms
[bump] bump-render=high-poly: FPS: 13117 FrameTime: 0.076 ms
[bump] bump-render=normals: FPS: 25773 FrameTime: 0.039 ms
[bump] bump-render=height: FPS: 22780 FrameTime: 0.044 ms
[effect2d] kernel=0,1,0;1,-4,1;0,1,0;: FPS: 12020 FrameTime: 0.083 ms
[effect2d] kernel=1,1,1,1,1;1,1,1,1,1;1,1,1,1,1;: FPS: 8243 FrameTime: 0.121 ms
[pulsar] light=false:quads=5:texture=false: FPS: 17688 FrameTime: 0.057 ms
[desktop] blur-radius=5:effect=blur:passes=1:separable=true:windows=4: FPS: 6112 FrameTime: 0.164 ms
[desktop] effect=shadow:windows=4: FPS: 8572 FrameTime: 0.117 ms
[buffer] columns=200:interleave=false:update-dispersion=0.9:update-fraction=0.5:update-method=map: FPS: 2441 FrameTime: 0.410 ms
[buffer] columns=200:interleave=false:update-dispersion=0.9:update-fraction=0.5:update-method=subdata: FPS: 4313 FrameTime: 0.232 ms
[buffer] columns=200:interleave=true:update-dispersion=0.9:update-fraction=0.5:update-method=map: FPS: 2632 FrameTime: 0.380 ms
[ideas] speed=duration: FPS: 12130 FrameTime: 0.082 ms
[jellyfish] <default>: FPS: 11614 FrameTime: 0.086 ms
[terrain] <default>: FPS: 944 FrameTime: 1.060 ms
[shadow] <default>: FPS: 12543 FrameTime: 0.080 ms
[refract] <default>: FPS: 1464 FrameTime: 0.683 ms
[conditionals] fragment-steps=0:vertex-steps=0: FPS: 22047 FrameTime: 0.045 ms
[conditionals] fragment-steps=5:vertex-steps=0: FPS: 18565 FrameTime: 0.054 ms
[conditionals] fragment-steps=0:vertex-steps=5: FPS: 21956 FrameTime: 0.046 ms
[function] fragment-complexity=low:fragment-steps=5: FPS: 21136 FrameTime: 0.047 ms
[function] fragment-complexity=medium:fragment-steps=5: FPS: 18739 FrameTime: 0.053 ms
[loop] fragment-loop=false:fragment-steps=5:vertex-steps=5: FPS: 19952 FrameTime: 0.050 ms
[loop] fragment-steps=5:fragment-uniform=false:vertex-steps=5: FPS: 21601 FrameTime: 0.046 ms
[loop] fragment-steps=5:fragment-uniform=true:vertex-steps=5: FPS: 21644 FrameTime: 0.046 ms
=======================================================
                                  glmark2 Score: 15815 
=======================================================

vkmark

vkmark results:

=======================================================
    vkmark 2025.01
=======================================================
    Vendor ID:      0x1002
    Device ID:      0x1114
    Device Name:    AMD Radeon Graphics (RADV GFX1152)
    Driver Version: 104857607
    Device UUID:    695bf57dbae8b103272568cd8e950eca
=======================================================
[vertex] device-local=true: FPS: 42875 FrameTime: 0.023 ms
[vertex] device-local=false: FPS: 39277 FrameTime: 0.025 ms
[texture] anisotropy=0: FPS: 25757 FrameTime: 0.039 ms
[texture] anisotropy=16: FPS: 25110 FrameTime: 0.040 ms
[shading] shading=gouraud: FPS: 24568 FrameTime: 0.041 ms
[shading] shading=blinn-phong-inf: FPS: 22000 FrameTime: 0.045 ms
[shading] shading=phong: FPS: 20788 FrameTime: 0.048 ms
[shading] shading=cel: FPS: 19903 FrameTime: 0.050 ms
[effect2d] kernel=edge: FPS: 19822 FrameTime: 0.050 ms
[effect2d] kernel=blur: FPS: 9088 FrameTime: 0.110 ms
[desktop] <default>: FPS: 9847 FrameTime: 0.102 ms
[cube] <default>: FPS: 31843 FrameTime: 0.031 ms
[clear] <default>: FPS: 26462 FrameTime: 0.038 ms
=======================================================
                                   vkmark Score: 24410
=======================================================

Note: vkmark needs to be compiled from source on Debian 12 and earlier.

GravityMark

GravityMark results: https://gravitymark.tellusim.com/report/?id=ff141d63a99826c825c259741f32ddc0b90599fd

Image

AI / LLM Inference

Basic ollama LLM model inference results: geerlingguy/ai-benchmarks#33

System CPU/GPU Model Eval Rate Power (Peak)
Framework 13 (Ryzen AI 5 340) CPU llama3.2:3b 23.81 Tokens/s 51.1 W
Framework 13 (Ryzen AI 5 340) CPU llama3.1:8b 10.46 Tokens/s 51.7 W
Framework 13 (Ryzen AI 5 340) CPU deepseek-r1:14b 5.83 Tokens/s 50.4 W

Memory

tinymembench results:

Click to expand memory benchmark result
tinymembench v0.4.10 (simple benchmark for memory throughput and latency)

==========================================================================
== Memory bandwidth tests                                               ==
==                                                                      ==
== Note 1: 1MB = 1000000 bytes                                          ==
== Note 2: Results for 'copy' tests show how many bytes can be          ==
==         copied per second (adding together read and writen           ==
==         bytes would have provided twice higher numbers)              ==
== Note 3: 2-pass copy means that we are using a small temporary buffer ==
==         to first fetch data into it, and only then write it to the   ==
==         destination (source -> L1 cache, L1 cache -> destination)    ==
== Note 4: If sample standard deviation exceeds 0.1%, it is shown in    ==
==         brackets                                                     ==
==========================================================================

 C copy backwards                                     :  19828.3 MB/s (0.6%)
 C copy backwards (32 byte blocks)                    :  19836.1 MB/s
 C copy backwards (64 byte blocks)                    :  19826.2 MB/s
 C copy                                               :  20017.6 MB/s
 C copy prefetched (32 bytes step)                    :  22003.1 MB/s
 C copy prefetched (64 bytes step)                    :  21885.2 MB/s
 C 2-pass copy                                        :  19881.1 MB/s (0.1%)
 C 2-pass copy prefetched (32 bytes step)             :  19780.0 MB/s
 C 2-pass copy prefetched (64 bytes step)             :  21066.9 MB/s (2.5%)
 C fill                                               :  34035.5 MB/s
 C fill (shuffle within 16 byte blocks)               :  34088.3 MB/s (0.2%)
 C fill (shuffle within 32 byte blocks)               :  33982.5 MB/s
 C fill (shuffle within 64 byte blocks)               :  34060.1 MB/s
 ---
 standard memcpy                                      :  26617.6 MB/s
 standard memset                                      :  36943.0 MB/s
 ---
 MOVSB copy                                           :  33799.5 MB/s (0.2%)
 MOVSD copy                                           :  33620.1 MB/s (1.0%)
 SSE2 copy                                            :  20418.6 MB/s
 SSE2 nontemporal copy                                :  33373.0 MB/s
 SSE2 copy prefetched (32 bytes step)                 :  22531.8 MB/s (0.1%)
 SSE2 copy prefetched (64 bytes step)                 :  22516.4 MB/s
 SSE2 nontemporal copy prefetched (32 bytes step)     :  33708.7 MB/s
 SSE2 nontemporal copy prefetched (64 bytes step)     :  33481.8 MB/s
 SSE2 2-pass copy                                     :  20035.3 MB/s
 SSE2 2-pass copy prefetched (32 bytes step)          :  21118.4 MB/s (1.0%)
 SSE2 2-pass copy prefetched (64 bytes step)          :  21184.2 MB/s
 SSE2 2-pass nontemporal copy                         :   6698.9 MB/s
 SSE2 fill                                            :  36075.0 MB/s
 SSE2 nontemporal fill                                :  62525.8 MB/s

==========================================================================
== Memory latency test                                                  ==
==                                                                      ==
== Average time is measured for random memory accesses in the buffers   ==
== of different sizes. The larger is the buffer, the more significant   ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM      ==
== accesses. For extremely large buffer sizes we are expecting to see   ==
== page table walk with several requests to SDRAM for almost every      ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest).                                         ==
==                                                                      ==
== Note 1: All the numbers are representing extra time, which needs to  ==
==         be added to L1 cache latency. The cycle timings for L1 cache ==
==         latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
==         two independent memory accesses at a time. In the case if    ==
==         the memory subsystem can't handle multiple outstanding       ==
==         requests, dual random read has the same timings as two       ==
==         single reads performed one after another.                    ==
==========================================================================

block size : single random read / dual random read, [MADV_NOHUGEPAGE]
      1024 :    0.0 ns          /     0.0 ns 
      2048 :    0.0 ns          /     0.0 ns 
      4096 :    0.0 ns          /     0.0 ns 
      8192 :    0.0 ns          /     0.0 ns 
     16384 :    0.0 ns          /     0.0 ns 
     32768 :    0.0 ns          /     0.0 ns 
     65536 :    0.5 ns          /     0.9 ns 
    131072 :    1.3 ns          /     1.7 ns 
    262144 :    1.7 ns          /     2.0 ns 
    524288 :    2.3 ns          /     2.6 ns 
   1048576 :    3.0 ns          /     3.4 ns 
   2097152 :    7.0 ns          /     9.1 ns 
   4194304 :    9.0 ns          /    10.9 ns 
   8388608 :   10.0 ns          /    11.4 ns 
  16777216 :   18.7 ns          /    24.7 ns 
  33554432 :   59.9 ns          /    83.5 ns 
  67108864 :   86.9 ns          /   105.2 ns 

block size : single random read / dual random read, [MADV_HUGEPAGE]
      1024 :    0.0 ns          /     0.0 ns 
      2048 :    0.0 ns          /     0.0 ns 
      4096 :    0.0 ns          /     0.0 ns 
      8192 :    0.0 ns          /     0.0 ns 
     16384 :    0.0 ns          /     0.0 ns 
     32768 :    0.0 ns          /     0.0 ns 
     65536 :    0.5 ns          /     0.9 ns 
    131072 :    1.3 ns          /     1.7 ns 
    262144 :    1.7 ns          /     2.0 ns 
    524288 :    1.9 ns          /     2.0 ns 
   1048576 :    2.3 ns          /     2.2 ns 
   2097152 :    5.8 ns          /     7.9 ns 
   4194304 :    8.0 ns          /     9.5 ns 
   8388608 :    8.2 ns          /     9.4 ns 
  16777216 :    9.7 ns          /    11.3 ns 
  33554432 :   53.9 ns          /    77.5 ns 
  67108864 :   78.0 ns          /    96.4 ns 

Core to Core Memory Latency

Image

Phoronix Test Suite

Results from pi-general-benchmark.sh:

  • pts/encode-mp3: 4.909 sec
  • pts/x264 1080p: 118.38 fps
  • pts/x264 4K: 26.20 fps
  • pts/phpbench: 1371865
  • pts/build-linux-kernel (defconfig): 145.648 sec

Activity

  1. geerlingguy commented on Nov 12, 2025

    @geerlingguy
    OwnerAuthor

    For idle power measurement with battery at 100% and plugged into wall power:

    • With screen on, at 50% brightness, about 5W
    • With screen off, about 2.8W

    I used the command:

    busctl --user set-property org.gnome.Mutter.DisplayConfig /org/gnome/Mutter/DisplayConfig org.gnome.Mutter.DisplayConfig PowerSaveMode i 1
    

    To blank the screen.

  2. added a commit that references this issue on Dec 1, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions