Can the top players beat the best computers anymore?

Or has the software gotten so good, even the best players can only hope for a draw at best?  The reason I ask is that I used to like to watch when Kasparov (or whoever) played against a top ranked computer opponent.  I haven't seen that in a long time.  I'm guessing that since the top players have no hope to get a win, there is no point.  Is that true?  Is it really now virtually impossible for anybody to get a win against the top program out there?  Are there no "anti-computer" strategies that work anymore?

from what I understand, on a timed game computers will almost always win because they attack with deadly accuracy and they do not blunder and with a limited time the human player wears down and can blunder or make a similar error. but with an unlimited amout of play, human players still win against the computer. I may be wrong but I believe this is so

Yelena Dembo might know such a strategy.

Rybka running on a $1,000 desktop computer has an estimated ELO of around 3200, so it should smoke any human competitor.  I'm sure GM's can get a win every now and then but I don't think a human can beat Rybka in a match of games.  Kind of sad, really.

When Kasparov lost the rematch against Deep Blue in 1997, the computer chess landscape was much different.  Look at the specs it took to beat Kasparov, 3.5-2.5. (from Wikipedia of course):

It was a massively parallel, RS/6000 SP Thin P2SC-based system with 30-nodes, with each node containing a 120 MHz P2SC microprocessor for a total of 30, enhanced with 480 special purpose VLSI chess chips.

The effective speed of specialized chips can not be deduced from their quantity or frequency, so the above information does not help much to compare to a regular desktop pc. The only thing of use would be how many positions are reviewed per sec. (I think that number still supercedes what today's desktops achieve).

On top of that, chess engines look at x positions and evaluate them. The rules for this evaluation differ from program to program. So even if we could say Deep Blue II was faster than an i7 cpu, the newer engines might still yield surprising results.

This is pretty much why engine ratings are computed by actually making them play each other. Even with comparable hardware, you can't predict what effect the evaluation has. If we knew what evaluation was superior in advance, we'd just write that chess program and be done with it. 

Top GMs have better technique and understanding in general, but they don't have the consistency to compete against the machine.  A correspondence match, for example, would be interesting -- but OTB even the best players would lose.

As notlesu points out, the engines that put up fights against past world champs are quite a bit outdated by today's standards.  Of course computer's rating are inflated vs each other, but they've certainly improved nonetheless.

There's no chance whatosever. Carlsen would maybe have a chance with draw odds against firebird on my laptop, but definitely not in a regular match. Rybka 2 had won against Grandmasters with pawn odds, and I think drew with draw odds, but engines are even a lot stronger now.

I still believe that human can beat the machine.

Why?

Because the Machine is made by Human!!

Doesn't work that way.

I dont think top players can beat the best computers but i dont think drawing is much of a problem.. Top players would be able to draw with top computers pretty easily but beating them wouldnt be really possible since computer are blunder-free and makes very little inaccuracies... Carlsen and maybe Anand are the only players IMO that have a decent chance of beating them..

I still believe that human can beat the machine.

Try flying faster than an airplane.

Humans can still beat the machines, we just need a hammer to do it now. 

 The people who designed Deep Blue weren't great at chess. The best player was a SM at his peak,I believe, but has stopped playing for several years. Afew people who were working on the project didn't know how to play chess at all. I suppose today's chess computers are no different.

Today's top programs have been created with the assistance of human GMs, and that is why they are superior in positional understanding to the previous generation.

My understanding  is that computers tend to have a good advantage AFTER the opening.  Their ability to calculate is enhanced in the complexity of the middle game or incredibly lengthy lines of endings.  

So, if you have a GM play a rybka or whatever but don't allow the puter to use an "opening book", the advantage may well turn to the GM.  However, this is not usually the case; the puters given extensive opening libraries to choose from. 

Anyway, that's my understanding of it...

If the top players are given draw odds and white every game, its a close match. 

That's been tried as well. Rybka 2 trashed GM Ehlvest with a 3 move opening book.

Given that humans cannot beat chess engines anymore, it is only a matter of time and circumstances (like how much money it will bring software developers and success for professional Chess960 players) for chess engines to dominate in Chess960 as well.

The game of Go is older than the game of Chess by a few thousand years, but the strongest Go engines are still weaker than master Go players.

As I see it, computers in current days will beat all human masters at timed games, as the computer analysis softwares are updated much more rapidly than any individual top grandmaster.

A human chess player may forget what he/she has played a game with success, but neglect some tactic composition opportunities. Such cases never happen on computers. Human players may be misled by their opponents for their dispositions, historical records or whatever empressions. However, a computer may always focus its evaluation and calcultion on the simple game state, irregardless of any human factors.