import numpy as np
N, P = map(int, input().split())
A = list(map(int, input().split()))
A.sort()

dp = np.zeros(P+1, dtype=np.int64)
ndp = np.zeros_like(dp)
for i in range(1, N):
    get = A[i]
    pena = A[i] - A[i-1]

    newDP = np.copy(dp)
    if pena:
        newDP[pena:] = np.maximum(newDP[pena:], dp[:-pena] + get)
    else:
        newDP += get

    np.maximum(dp, ndp, out=ndp)
    dp, ndp = ndp, newDP

np.maximum(dp, ndp, out=ndp)
print(ndp.max())