Rnorm drawing issue - draw sequentially or all at once

I am working on a Monte Carlo project, and in this matter, I have to draw a lot of random numbers. I have made two implementations, one where I draw the random numbers all at once and one here I draw them as part of a loop. My question appeared when I compared the outcome and became aware that they weren't the same. I immediately suspected the drawing of the random numbers to be the cause of the differences, and it appears to be the case from the study below.

x_2 <- matrix(0,n,t)
set.seed(123)
for(s in 1:t) {
  x_2[, s] <-rnorm(5000,0,3)
  if (s<=10){ print(rnorm(5000,0,3)[1:10])}
}  
set.seed(123)
e_2 <- matrix(rnorm(5000 * 5000, 0, 3), 5000, 5000)
print(e_2[1:10,1:10])

The outcome of the above can be seen below.

[1] -1.482522  3.382780 -3.440849  4.443056  2.748574  1.005393  1.724026  0.610859 -1.341124 -1.030578
[1] -4.061547 -1.738132 -2.583133  2.918035  1.857437  4.156337 -4.461811  1.917578  1.124205  1.110564
[1] -2.3856225 -3.4053345  1.7403037  1.5541277  0.6230563 -1.4502206 -1.1568202 -1.3942097  3.6861666 -0.2023702
[1] -2.6641585  1.0807931  1.2662698 -2.2296928  0.5406639  2.8972701  1.0827109 -1.4748768  0.6064607 -0.7711647
[1]  2.9164836 -0.6704332  0.4415908 -0.6774140  1.7499817  2.9110134  0.1417097 -1.2382027 -1.2465868  1.5946641
[1]  1.24145410  4.74917343 -0.04523692  3.16943742  2.37547831  1.40694830  0.24448225 -1.65932819  1.13039946 -0.55707338
[1]  1.3946592 -1.0632645 -2.7107054  5.1341154  4.5597899 -2.9279816 -2.5342361 -2.8255961 -0.8627209  0.2887078
[1]  0.27116861  1.31304754  0.65551315 -0.04878318 -2.75273200  1.59058082  4.08446872  0.58036680  0.41595295 -4.32112988
[1]  4.342664  2.175384  7.098014 -4.497943 -3.546591 -1.108493  1.748281  2.921789  1.493995 -3.192363
[1] -2.1023524  1.0911300  0.4935613 -0.1076302 -3.9271420  1.4016313 -3.7191495 -1.2465567 -0.1952575  3.8618488

       [,1]      [,2]       [,3]      [,4]       [,5]       [,6]        [,7]       [,8]       [,9]      [,10]
[1,] -1.6814269 -1.482522  7.1121757 -4.061547 -2.5088902 -2.3856225 -0.58081924 -2.6641585  1.4475821  2.9164836
[2,] -0.6905325  3.382780 -0.5004359 -1.738132 -0.6617190 -3.4053345  0.77444184  1.0807931  2.1640940 -0.6704332
[3,]  4.6761249 -3.440849  2.7808842 -2.583133 -6.3105443  1.7403037 -1.61493772  1.2662698 -1.5233769  0.4415908
[4,]  0.2115252  4.443056 -1.7044552  2.918035 -5.0034226  1.5541277 -3.53718843 -2.2296928 -0.1941262 -0.6774140
[5,]  0.3878632  2.748574  0.6752702  1.857437 -3.2938886  0.6230563  2.70194227  0.5406639  3.9063635  1.7499817
[6,]  5.1451950  1.005393  3.3959577  4.156337 -4.9968636 -1.4502206 -0.04874864  2.8972701 -0.6559106  2.9110134
[7,]  1.3827486  1.724026  4.1414446 -4.461811 -0.1485019 -1.1568202  0.53589147  1.0827109  3.6963324  0.1417097
[8,] -3.7951837  0.610859 -0.6984102  1.917578  4.6779088 -1.3942097  1.80572706 -1.4748768  4.4850740 -1.2382027
[9,] -2.0605586 -1.341124 -4.8019300  1.124205 -1.2138718  3.6861666  0.78340314  0.6064607 -0.4918906 -1.2465868
[10,] -1.3369859 -1.030578 -0.8951075  1.110564  2.3580783 -0.2023702  0.91219698 -0.7711647 -4.1529154  1.5946641

What I noted from the outcome above, is that each print of the loop corresponds to every second column in the matrix, which contains all draws. I can't figure out why this happens, and I really hope that someone can point out my mistake.