Could Carlsen win against Houdini in a Correspondence Chess Match?

I was wondering if given a week per move for Carlsen and Houdini who would win? I thought computers were stronger than humans but I heard somewhere that they still don't have the positional understanding necessary to win against the strongest human players in CC. 

For arguments sake, say that if the game is drawn they play another reversing colors until one of them wins.  Also say that magnus devotes his full attension to the game and is allowed the same resources as a player here on chess.com (DB's, books, analysis board, etc. but no engines, tablebases, or other people).

Houdini.

B.Ivanov won 10-0  to Houdini.     why n0t carlsen?

FWIW I don't think he'd have much trouble at least drawing the games.  Those "lowly" 2600 GMs didn't do so bad against Ivanov (i.e. houdini) in OTB chess in positional lines (at least 1 drawn game) and they didn't even know they were playing a computer.  So CC and a 2800 player shouldn't be so bad I'd think.

that's actually a really interesting point. i remember the game saric played against borislav was quite a fighting draw.

Does Houdini also get to calculate for a week straight? While Magnus has to take breaks, a computer does not.

No.

Tournament chess averages a few minutes per move, so I was thinking between 1 and 10.  It also depends on the hardware.

what would happen if Houdini was wrapped up in a box with its opening book and tablebases, then let loose onto the super-GM tournament circuit without any updating?

I'm sure it would start out at around 3200 or whatever, but would these smart GMs find it out, and it would drop and drop in rating? after all, it can't learn, and people can.

sorry for changing the quesion, ignore me. Carlsen thinking for a long time is a frightening idea. but Houdini on decent hardware is well ahead, I'd say Houdini. If Carlsen can't find the best move in 4 hours, is 12 hours going to be any better? people get tired.  unless he has a team.

It's true depending on their chess knoweldge different players can use engines to different effects.  A GM will know when to believe an engine and when to keep looking for example.  But this is in analysis.  In a man vs machine match it really defeats the purpose for a human to aid the engine.

As for suggesting moves, the engines come up with moves on their own.  Not sure what you mean by suggesting moves.  Again that's more along the lines of analysis, where a strong player forces the engine to evaluate certain lines of play when they don't trust the engine's evaluation.

I was thinking Houdini would think as much as it would when it is playing otb.

Since they prepare with engines all the time, they'd at least draw it all the time.  In any case FIDE ratings aren't CCRL ratings.  I doubt it would even be close to 3200.

Except, like I said, even 2600 GMs were getting drawn positions vs houdini in tournament play (no analysis, limited time controls, etc).  One of the games even ended in a draw (some Ivanov screwed up and lost hah).

I was imagining Houdini playing unaided by a human and running on the type of computer used at the computer chess world championship.

I don't think Ivanov counts. we don't know what he was doing or on what hardware. 

That's true, it probably wasn't great hardware, and he moved really fast.

User interface... it's this crazy thing that they came up with, well, since the invention of the computer actually.

is Carlsen allowed to walk over and pull the plug?

Could Carlsen win against Houdini in a Correspondence Chess Match?

(No)

i think carleson wode win