Does it make a difference if you use this instead:

restart;

g_list:=[sin(t)*diff(x(t),t$2)+cos(t)*diff(x(t),t)+2*x(t)=0,
         diff(y(x),x)=lambda*sin(lambda*x)*y(x)^2+a*cos(lambda*x)^n*y(x)-a*cos(lambda*x)^(n-1)]:

work_func:=proc(i::posint)  
  :-dsolve(g_list[i]):
end proc:

Threads:-Seq( Threads:-Create( work_func(i)), i=1..2) ;

@C_R The error you describe in Maple 2020 is present in all previous versions I presume. This code illustrates what you are doing:

eqs := [a = b, c = d, e = f]:
map(map@`^`, eqs, 2);
map(Map@`^`, eqs, 2);

@C_R I let Maple create the matrix by submitting system of linear equations.

@nm I had this very idea, but I did not anticipate it would be faster. The code is MUCH faster in cmaple the wmaple plus theere is no memory issue. I now wonder if my wmaple is badly configured.

@dharr Normalizing an expression in a field carries the overhead of identifying and applying the field. Is this cost applies once per matrix or once per matrix entry on each operation? I have a faint recollection of there being a long ago discussion about timing/optimizing operations on matrices with entries mod a polynomial.

@mmcdara Is there a way to see a list of all variables saved in a .m file?

@Carl Love Once, at console. Does saving to .m file overwrite the file or change it?... After testing it, it does overwrite.

@Carl Love An oversight on my part. I reacted too quickly.

@Carl Love The result in your reply is the correct one, that is the one you requested. Compare:
 

sys:= [p^2-4*q = 0, 2*p^5 - 15*p^3*q + 27*p*q^2 < 0]:
solve(sys, {p});
solve(sys, {q});
solve(sys, {p, q});

.               

@C_R That helped, thank You. Any idea why the search engine would be targeted for login credentials?

You should inspect eq1,eq2,eq3. The fourth equation is redundant:

normal(eq1+eq2+eq3);

Your RootOf is in poor form and it is reducible. Reducible RootOfs are a no-no and shoud be subcased. Mixing RootOfs and radicals is no-no as well, thought I am unceratin of this one. You have much work to do here.

@acer ...odd. Somehow I ended up on EigenConditionNumbers page.

Thanks.

Are you going to post the problem itself?