7...Ne4 leads to more lively play, as White gets nothing by trading queens in this line. I have yet to see a refutation of Black's opening scheme after 8.Qe2 Bf5 9.Rd1 Qc8.
8.Qxd8+ Kxd8 9.Nc3
9...h6
This old move came as a bit of a surprise, since most theory books seem to suggest it's inferior. Kramnik played 9...Bd7 (with the idea of walking the king to the queenside) in games 1 and 3.
10.Rd1+ Ke8
10...Bd7? goes down like a ton of bricks after 11.g4 Ne7 12.e6 fxe6 13.Ne5.
11.h3 a5 12.Bf4 Be6
13.g4!?
Very double-edged, as this bold advance might turn out to be a serious weakening of White's kingside.
13...Ne7 14.Nd4 Nd5 15.Nce2 Bc5
If you compare this position with games 1 and 3, you'll find that Black's pieces are much more active here.
16.Nxe6 fxe6 17.c4 Nb6
Not exactly the ideal square for Mr. Ed, but there's always the option of recycling it to c5 at a later stage.
18.b3
Apparently the first new move of the game. In Vuchkovic-Sakaev earlier this year, White tried 18.Bg3. Kasparov probably felt that the bishop would by offside here.
18...a4 19.Bd2 Kf7
Connecting the rooks, thus completing Black's development. White's kingside majority is not enough for an advantage here. Note that 19...Rf8 20.Nf4 Bxf2+?! (20...Bd4? 21.Nxe6) 21.Kxf2 g5 22.Kg3 gxf4+ 23.Bxf4 was clearly better for White in view of Black's weak h-pawn.
White has some targets on the queenside, but Black's counterplay turns out to be sufficient.
26...Be7 27.Ra7 Nc5 28.f3
28.b4 Ne4 29.Be1 Rd1 and Black has at least a draw.
28...Nxb3 29.Rxb7
29...Nc1 30.Nxc1 Rxc3 draw agreed.
Black gets a passed c-pawn after 31.Ne2 Rxc4, but White holds with ease after 32.Rxc7 c5 33.Ng3 Rc2+ 34.Kf1 c4 35.Ne4. For example 35...Ke8 36.Nf6+ Kf8 37.Rc8+ Kf7 38.Ne4 Bb4 39.Nd6+ Bxd6 40.exd6 Rd2 41.Rxc4 Rxd6 and the rook ending is a dead draw.
1/2-1/2
Related articles
The Kasparov - Kramnik index.
The Kasparov - Kramnik FAQ.