TalkChess.com

Looking forward to see the first Stockfish/Komodo Threadripper 1950x benchmarks.... anyway, I have placed my order to receive my own.... hopefully if data are not published by then, I will be able to do my own.

Rgds

Lion wrote:Looking forward to see the first Stockfish/Komodo Threadripper 1950x benchmarks.... anyway, I have placed my order to receive my own.... hopefully if data are not published by then, I will be able to do my own.

Rgds

I am also looking forward to it. Nice purchase.

Sweet!

Really looking forward to see some crazy benchmark scores I hope they start manufacturing in 7nm in 2018 so we can have 24 core threadrippers in late 2018/ early 2019

I found some new quotes about threadripper:

"AMD Threadripper 1950X OCed To 4.1GHz On All 16 Cores & Tested, Delivers Otherworldly Performance"

"We’re less than 48 hours away from AMD’s official Ryzen Threadripper launch."

"Let’s start out with the Cinebench R15 result first. Here, the 1950X manages to deliver a whopping 3337 points with all sixteen of its cores running at 4.0GHz. By comparison a similarly priced Intel Core i7 7900X gets 2169 points at stock and maxes out at 2464 with a hefty 4.6GHz overclock. That comes out to a 35%+ difference when both chips are overclocked. Not even the world record Cinebench R15 run of the 7900X at 5.96GHz with liquid nitrogen cooling is enough to surpass Threadripper’s score."

"AMD Takes 10.4% CPU Share From Intel In Q2 2017"

The release and bench lift-up should be today.
I dont know at which hour, but probably US time zone.... so we should know more very very soon.

rgds

First results look really promising ?

http://www.hardware.fr/articles/967-11/ ... omodo.html


They mention that under Stockfish there is a little of "throttling".... Not sure what they are meaning by that.

rgds

It is pity but meanwhile the speed of Threadripper 1950x is near two times higher then Ryzen 7 1800x, the enhancement of Elo may be not more as 40-50 Elo.

I dont look at it this way.

The way I look at it, is that I nearly takes me 2x less time to have the same result and thus, can spend more time doing something else.

rgds

If you search for a good move in a given position you should fill more than half time for it. The gain of time is about 30% instead of 50%.