EBC encryption: 5.833 GB/s
EBC encryption blocked: 802.582 MB/s
EBC decryption: 5.74 GB/s
EBC decryption blocked: 802.291 MB/s

public void EncryptEcb(Span<byte> data)
{
    Vector128<byte>[] keys = RoundKeys;
    Span<Vector128<byte>> blocks = MemoryMarshal.Cast<byte, Vector128<byte>>(data);

    // Makes the JIT remove all the other range checks on keys
    Vector128<byte> key10 = keys[10];

    for (int i = 0; i < blocks.Length; i++)
    {
        Vector128<byte> b = blocks[i];

        b = Sse2.Xor(b, keys[0]);
        b = Aes.Encrypt(b, keys[1]);
        b = Aes.Encrypt(b, keys[2]);
        b = Aes.Encrypt(b, keys[3]);
        b = Aes.Encrypt(b, keys[4]);
        b = Aes.Encrypt(b, keys[5]);
        b = Aes.Encrypt(b, keys[6]);
        b = Aes.Encrypt(b, keys[7]);
        b = Aes.Encrypt(b, keys[8]);
        b = Aes.Encrypt(b, keys[9]);
        b = Aes.EncryptLast(b, keys[10]);

        blocks[i] = b;
    }
}


; Assembly listing for method AesContext:EncryptEcb(struct):this
; Emitting BLENDED_CODE for X64 CPU with AVX - Windows
; optimized code
; rsp based frame
; fully interruptible
; Final local variable assignments
;
;  V00 this         [V00,T06] (  3,  3   )     ref  ->  rcx         this class-hnd
;  V01 arg1         [V01,T04] (  4,  8   )   byref  ->  rdx
;  V02 loc0         [V02,T01] ( 13, 46   )     ref  ->  rax         class-hnd
;* V03 loc1         [V03    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op
;  V04 loc2         [V04,T02] (  5, 17   )     int  ->   r8
;  V05 loc3         [V05,T00] ( 24, 96   )  simd16  ->  mm0
;  V06 OutArgs      [V06    ] (  1,  1   )  lclBlk (32) [rsp+0x00]   "OutgoingArgSpace"
;* V07 tmp1         [V07,T16] (  0,  0   )     int  ->  zero-ref    "Inline stloc first use temp"
;* V08 tmp2         [V08,T17] (  0,  0   )     int  ->  zero-ref    "Inline stloc first use temp"
;* V09 tmp3         [V09    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "Inlining Arg"
;  V10 tmp4         [V10,T11] (  2,  2   )     int  ->  rdx         "Inline stloc first use temp"
;* V11 tmp5         [V11    ] (  0,  0   )    long  ->  zero-ref    "Inline stloc first use temp"
;  V12 tmp6         [V12,T12] (  2,  2   )     int  ->  rdx         "Inline stloc first use temp"
;* V13 tmp7         [V13    ] (  0,  0   )  struct (16) zero-ref    "NewObj constructor temp"
;* V14 tmp8         [V14    ] (  0,  0   )   byref  ->  zero-ref    "Inlining Arg"
;* V15 tmp9         [V15    ] (  0,  0   )  struct ( 8) zero-ref    "NewObj constructor temp"
;* V16 tmp10        [V16    ] (  0,  0   )   byref  ->  zero-ref    V25._pointer(offs=0x00) P-INDEP "field V01._pointer (fldOffset=0x0)"
;* V17 tmp11        [V17    ] (  0,  0   )     int  ->  zero-ref    V25._length(offs=0x08) P-INDEP "field V01._length (fldOffset=0x8)"
;  V18 tmp12        [V18,T07] (  2,  5   )   byref  ->  rcx         V03._pointer(offs=0x00) P-INDEP "field V03._pointer (fldOffset=0x0)"
;  V19 tmp13        [V19,T05] (  3,  6   )     int  ->  rdx         V03._length(offs=0x08) P-INDEP "field V03._length (fldOffset=0x8)"
;  V20 tmp14        [V20,T08] (  2,  2   )   byref  ->  rcx         V09._pointer(offs=0x00) P-INDEP "field V09._pointer (fldOffset=0x0)"
;  V21 tmp15        [V21,T13] (  2,  2   )     int  ->  rdx         V09._length(offs=0x08) P-INDEP "field V09._length (fldOffset=0x8)"
;  V22 tmp16        [V22,T09] (  2,  2   )   byref  ->  rcx         V13._pointer(offs=0x00) P-INDEP "field V13._pointer (fldOffset=0x0)"
;  V23 tmp17        [V23,T14] (  2,  2   )     int  ->  rdx         V13._length(offs=0x08) P-INDEP "field V13._length (fldOffset=0x8)"
;  V24 tmp18        [V24,T10] (  2,  2   )   byref  ->  rcx         V15._value(offs=0x00) P-INDEP "field V15._value (fldOffset=0x0)"
;* V25 tmp19        [V25    ] (  0,  0   )  struct (16) zero-ref    "Promoted implicit byref"
;  V26 cse0         [V26,T03] (  3, 12   )   byref  ->   r9         "ValNumCSE"
;* V27 cse1         [V27    ] (  0,  0   )    long  ->  zero-ref    "ValNumCSE"
;  V28 cse2         [V28,T15] (  2,  2   )     int  ->   r8         "ValNumCSE"
;
; Lcl frame size = 40

G_M18342_IG02:
       mov      rax, gword ptr [rcx+8]
       mov      rcx, bword ptr [rdx]
       mov      edx, dword ptr [rdx+8]
       shr      edx, 4
       mov      r8d, dword ptr [rax+8]
       cmp      r8d, 10
       jbe      G_M18342_IG06
       xor      r8d, r8d
       test     edx, edx
       jle      SHORT G_M18342_IG04

G_M18342_IG03:
       movsxd   r9, r8d
       shl      r9, 4
       add      r9, rcx
       vmovupd  xmm0, xmmword ptr [r9]
       vpxor    xmm0, xmm0, xmmword ptr [rax+16]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+32]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+48]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+64]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+80]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+96]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+112]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+128]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+144]
       vaesenc  xmm0, xmm0, xmmword ptr [rax+160]
       vaesenclast xmm0, xmm0, xmmword ptr [rax+176]
       vmovupd  xmmword ptr [r9], xmm0
       inc      r8d
       cmp      r8d, edx
       jl       SHORT G_M18342_IG03

G_M18342_IG04:
       add      rsp, 40
       ret

G_M18342_IG05:
       call     CORINFO_HELP_OVERFLOW
       int3

; Total bytes of code 163, prolog size 7 for method AesContext:EncryptEcb(struct):this
; ============================================================