Benchmark Sipeed Lichee Console 4A

[geerlingguy]

I would like to see how the Lichee Console 4A runs; it uses the 4-core TH1520 SoC, underclocked to 1.5 GHz. See related: geerlingguy/sbc-reviews#39

[geerlingguy]

If trying to install Ansible locally, to run the playbook on the device itself, I run into:

            building '_cffi_backend' extension
            creating build/temp.linux-riscv64-cpython-311
            creating build/temp.linux-riscv64-cpython-311/src
            creating build/temp.linux-riscv64-cpython-311/src/c
            riscv64-linux-gnu-gcc -Wsign-compare -DNDEBUG -g -fwrapv -O2 -Wall -g -fstack-protector-strong -Wformat -Werror=format-security -fPIC -DFFI_BUILDING=1 -DUSE__THREAD -DHAVE_SYNC_SYNCHRONIZE -I/usr/include/ffi -I/usr/include/libffi -I/usr/include/python3.11 -c src/c/_cffi_backend.c -o build/temp.linux-riscv64-cpython-311/src/c/_cffi_backend.o
            src/c/_cffi_backend.c:15:10: fatal error: ffi.h: No such file or directory
               15 | #include <ffi.h>
                  |          ^~~~~~~
            compilation terminated.
            error: command '/usr/bin/riscv64-linux-gnu-gcc' failed with exit code 1
            [end of output]
      
        note: This error originates from a subprocess, and is likely not a problem with pip.
        ERROR: Failed building wheel for cffi
      Failed to build cffi
      ERROR: Could not build wheels for cffi, which is required to install pyproject.toml-based projects
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
error: subprocess-exited-with-error

× pip subprocess to install build dependencies did not run successfully.
│ exit code: 1
╰─> See above for output.

So I'll have to run it from outside.

[geerlingguy]

================================================================================
HPLinpack 2.3  --  High-Performance Linpack benchmark  --   December 2, 2018
Written by A. Petitet and R. Clint Whaley,  Innovative Computing Laboratory, UTK
Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
Modified by Julien Langou, University of Colorado Denver
================================================================================

An explanation of the input/output parameters follows:
T/V    : Wall time / encoded variant.
N      : The order of the coefficient matrix A.
NB     : The partitioning blocking factor.
P      : The number of process rows.
Q      : The number of process columns.
Time   : Time in seconds to solve the linear system.
Gflops : Rate of execution for solving the linear system.

The following parameter values will be used:

N      :   34581
NB     :     256
PMAP   : Row-major process mapping
P      :       1
Q      :       4
PFACT  :   Right
NBMIN  :       4
NDIV   :       2
RFACT  :   Crout
BCAST  :  1ringM
DEPTH  :       1
SWAP   : Mix (threshold = 64)
L1     : transposed form
U      : transposed form
EQUIL  : yes
ALIGN  : 8 double precision words

--------------------------------------------------------------------------------

- The matrix A is randomly generated for each test.
- The following scaled residual check will be computed:
      ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
- The relative machine precision (eps) is taken to be               1.110223e-16
- Computational tests pass if scaled residuals are less than                16.0

================================================================================
T/V                N    NB     P     Q               Time                 Gflops
--------------------------------------------------------------------------------
WR11C2R4       34581   256     1     4            6225.22             4.4289e+00
HPL_pdgesv() start time Sat Mar 23 19:39:43 2024

HPL_pdgesv() end time   Sat Mar 23 21:23:28 2024

--------------------------------------------------------------------------------
||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)=   3.03925351e-03 ...... PASSED
================================================================================

Finished      1 tests with the following results:
              1 tests completed and passed residual checks,
              0 tests completed and failed residual checks,
              0 tests skipped because of illegal input values.
--------------------------------------------------------------------------------

End of Tests.
================================================================================

I don't have power metrics. I'll run it again on the test bench later and get that.

[geerlingguy]

To get a power reading, I plugged in the wall charger and charged the device for a few hours until the power reading was under 1W (meaning the battery was fully charged):

Then I re-ran the benchmark a couple times, measuring power consumption:

Here is the final result:

================================================================================
HPLinpack 2.3  --  High-Performance Linpack benchmark  --   December 2, 2018
Written by A. Petitet and R. Clint Whaley,  Innovative Computing Laboratory, UTK
Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
Modified by Julien Langou, University of Colorado Denver
================================================================================

An explanation of the input/output parameters follows:
T/V    : Wall time / encoded variant.
N      : The order of the coefficient matrix A.
NB     : The partitioning blocking factor.
P      : The number of process rows.
Q      : The number of process columns.
Time   : Time in seconds to solve the linear system.
Gflops : Rate of execution for solving the linear system.

The following parameter values will be used:

N      :   34581
NB     :     256
PMAP   : Row-major process mapping
P      :       1
Q      :       4
PFACT  :   Right
NBMIN  :       4
NDIV   :       2
RFACT  :   Crout
BCAST  :  1ringM
DEPTH  :       1
SWAP   : Mix (threshold = 64)
L1     : transposed form
U      : transposed form
EQUIL  : yes
ALIGN  : 8 double precision words

--------------------------------------------------------------------------------

- The matrix A is randomly generated for each test.
- The following scaled residual check will be computed:
      ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
- The relative machine precision (eps) is taken to be               1.110223e-16
- Computational tests pass if scaled residuals are less than                16.0

================================================================================
T/V                N    NB     P     Q               Time                 Gflops
--------------------------------------------------------------------------------
WR11C2R4       34581   256     1     4            6184.07             4.4584e+00
HPL_pdgesv() start time Mon Mar 25 16:41:12 2024

HPL_pdgesv() end time   Mon Mar 25 18:24:16 2024

--------------------------------------------------------------------------------
||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)=   3.03925351e-03 ...... PASSED
================================================================================

Finished      1 tests with the following results:
              1 tests completed and passed residual checks,
              0 tests completed and failed residual checks,
              0 tests skipped because of illegal input values.
--------------------------------------------------------------------------------

End of Tests.
================================================================================

So we have a total of 4.4584 Gflops, at 8.5 W, giving 0.52 Gflops/W.