Alguém tem alguma ideia de por que a taxa de erros diminui a cada loop?
canal duplo Geil EVO Spear DDR4-2133
sudo memtester 6000 5
memtester version 4.5.1 (64-bit)
Copyright (C) 2001-2020 Charles Cazabon.
Licensed under the GNU General Public License version 2 (only).
pagesize is 4096
pagesizemask is 0xfffffffffffff000
want 6000MB (6291456000 bytes)
got 6000MB (6291456000 bytes), trying mlock ...locked.
Loop 1/5:
Stuck Address : testing 2FAILURE: possible bad address line at offset 0x000c04d0.
Skipping to next test...
Random Value : FAILURE: 0x3ed5db7679ece79f != 0x3ed7db7679ece79f at offset 0x2027fce0.
FAILURE: 0xf7fc8f5ffe7e74ef != 0xf7fe8f5ffe7e74ef at offset 0x35fa5ce0.
FAILURE: 0x10400b8d82107fb3 != 0x10420b8d82107fb3 at offset 0x35fa5ce0.
Compare XOR : FAILURE: 0x60a20d1c0800e073 != 0x60a00d1c0800e073 at offset 0x61914de0.
Compare SUB : Compare MUL : ok
FAILURE: 0x2000000000004 != 0x00000004 at offset 0x319f1ce0.
FAILURE: 0x2000000000004 != 0x00000004 at offset 0x659e9ce0.
Compare DIV : FAILURE: 0xfffdffffffffffff != 0xffffffffffffffff at offset 0x35fa5ce0.
Compare OR : FAILURE: 0xc7f4e347d8fafda != 0xc7e4e347d8fafda at offset 0x0a5fb5d0.
FAILURE: 0xc7f4e347d8fafda != 0xc7e4e347d8fafda at offset 0x3c3094d0.
Compare AND : FAILURE: 0xffec15854f70c19d != 0xffee15854f70c19d at offset 0x088b5ce0.
FAILURE: 0xffec1585563811bd != 0xffee1585563811bd at offset 0x3ec5dde0.
Sequential Increment: Solid Bits : testing 2FAILURE: 0xfffeffffffffffff != 0xffffffffffffffff at offset 0x09c774d0.
Block Sequential : testing 0FAILURE: 0x2000000000000 != 0x00000000 at offset 0x659e9ce0.
FAILURE: 0x1000000000000 != 0x00000000 at offset 0x692f15d0.
FAILURE: 0x2000000000000 != 0x00000000 at offset 0x7a8adce0.
Checkerboard : testing 1FAILURE: 0xaaa8aaaaaaaaaaaa != 0xaaaaaaaaaaaaaaaa at offset 0x35fa5ce0.
Bit Spread : testing 0FAILURE: 0x1000000000005 != 0x00000005 at offset 0x3abe94d0.
Bit Flip : testing 3FAILURE: 0x2000000000001 != 0x00000001 at offset 0x3abe5de0.
Walking Ones : ok
Walking Zeroes : ok
8-bit Writes : ok
16-bit Writes : ok
Loop 2/5:
Stuck Address : ok
Random Value : ok
Compare XOR : ok
Compare SUB : ok
Compare MUL : ok
FAILURE: 0x2000000000001 != 0x00000001 at offset 0x53351ce0.
FAILURE: 0x2000000000001 != 0x00000001 at offset 0x65ae9ce0.
Compare DIV : Compare OR : ok
Compare AND : ok
Sequential Increment: ok
Solid Bits : testing 3FAILURE: 0x2000000000000 != 0x00000000 at offset 0x659e9ce0.
Block Sequential : testing 0FAILURE: 0x2000000000000 != 0x00000000 at offset 0x319f1ce0.
Checkerboard : testing 11FAILURE: 0xaaa8aaaaaaaaaaaa != 0xaaaaaaaaaaaaaaaa at offset 0x35fa5ce0.
Bit Spread : ok
Bit Flip : testing 77FAILURE: 0x1000000000200 != 0x00000200 at offset 0x692f15d0.
Walking Ones : ok
Walking Zeroes : ok
8-bit Writes : ok
16-bit Writes : ok
Loop 3/5:
Stuck Address : testing 5FAILURE: possible bad address line at offset 0x3e2814d0.
Skipping to next test...
Random Value : ok
Compare XOR : ok
Compare SUB : ok
Compare MUL : ok
Compare DIV : ok
Compare OR : ok
Compare AND : ok
Sequential Increment: ok
Solid Bits : testing 15FAILURE: 0x1000000000000 != 0x00000000 at offset 0x692f15d0.
Block Sequential : testing 152FAILURE: 0x989a989898989898 != 0x9898989898989898 at offset 0x22c61de0.
Checkerboard : ok
Bit Spread : ok
Bit Flip : ok
Walking Ones : ok
Walking Zeroes : ok
8-bit Writes : ok
16-bit Writes : ok
Loop 4/5:
Stuck Address : ok
Random Value : ok
Compare XOR : ok
Compare SUB : ok
Compare MUL : ok
Compare DIV : ok
Compare OR : ok
Compare AND : ok
Sequential Increment: ok
Solid Bits : ok
Block Sequential : ok
Checkerboard : ok
Bit Spread : ok
Bit Flip : ok
Walking Ones : ok
Walking Zeroes : ok
8-bit Writes : ok
16-bit Writes : ok
Loop 5/5:
Stuck Address : ok
Random Value : ok
Compare XOR : ok
Compare SUB : ok
Compare MUL : ok
Compare DIV : ok
Compare OR : ok
Compare AND : ok
Sequential Increment: ok
Solid Bits : ok
Block Sequential : ok
Checkerboard : ok
Bit Spread : ok
Bit Flip : ok
Walking Ones : ok
Walking Zeroes : ok
8-bit Writes : ok
16-bit Writes : ok
Done.
Substitua os módulos de memória. Se os problemas persistirem, peça ajuda ao suporte de hardware para substituir outros componentes que podem falhar, como ventoinhas, fonte de alimentação, CPU ou placas de sistema.
Verifique o hardware de detecção de erros, a memória ECC e outros recursos RAS para saber exatamente o que falhou.
O memtester não informará onde ou por que o hardware falhou, apenas que os valores armazenados das funções matemáticas não eram como esperado.
Levaria uma quantidade enorme de tempo para isolar fatores ambientais como temperatura, voltagem e radiação, descobrir quais módulos físicos podem ter um problema e, mais difícil, por que é um problema intermitente. Não vale a pena para componentes consumíveis, tire-os da produção. Talvez mantê-los em um ambiente de teste com o qual você não se importa, para observar se esse tipo de falha de hardware é um problema para sua carga de trabalho.