No: comparison with the 1.16m males aged 35 at the January 1959 census shows something like a third surviving until then - bad enough, but mercifully a good deal better. There could be some age-heaping (common at reported years ending in 0 or 5), but it seems very modest and nothing like as severe as clustering at 30 and 40 (the big jump from 0.96m aged 36 seems instead mostly a genuine reflection of the comparable rise in births in 1923 after the ordeals of 1915-22).

The 1923 comparison is necessarily a bit uncertain because of boundary changes during and immediately after the war and because birth registration wasn't yet universally established throughout the then USSR, but Jean-Noel Biraben's 3.43m for the Soviet territory of 1959 (Naissances et repartition par age dans l'Empire russe et l'Union sovietique, Population 32:2, 1976) looks about right, and is consistent with the 2.28m three-year-old boys recorded in the smaller interwar USSR at the December 1926 census (Lorimer's life table indicates that 28.6% of boys born in 1923 would have died by then, implying an original cohort of 3.19m to which we can add 0.25m or so for the territories annexed subsequently where births were generally lower). A cruder shortcut assuming that male births were 2.3% of a population of 139m (consistent with an estimated crude birth rate of 45 per thousand) incidentally gives the same 3.19m births in the old area.

How many of those 3.4m or so survived in 1945? Scaling up the 1959 survivors by a tenth probably gives a reasonable picture, perhaps 1.27m or so or 37%. So a Soviet boy born in 1923 had around a 63% chance of dying by 1945. But crucially, in the prewar USSR 41½% were already dead by January 1939 owing mainly to the high infant and child mortality of the 1920s (the former alone claiming a fifth). So the war period claimed perhaps a fifth of the original cohort, or 35% of the young men of this age alive at its start.

It might of course be preferable to split the 1959 survivors between the interwar and newly-incorporated terrirories to look at only the Soviet-born 1923 cohort, but as the added areas contained about 11% of the augmented population and were also the scene of bitter conflict and postwar privation I wouldn't expect the general picture to be much altered: if survivorship differed there by as much as ten percentage points it would only affect the overall result by 1%.