Iphone 4

[Swaminathan]

Yes, I know that I shouldn't assume the engine to perform at 2800's. But even 2500's or just around that rating would be Ok for me as I had bought the I-phone not caring about chess programs at all. I bought it mainly for phone/internet/music/camera etc. So, the knowledge of 2500 and beyond is just good enough for me to want to install the application. It's just that I'm amazed that I could have a program about IM or even GM's strength in iphone.

Best wishes,
Swami

[Tord]

Yes, I know that I shouldn't assume the engine to perform at 2800's. But even 2500's or just around that rating would be Ok for me as I had bought the I-phone not caring about chess programs at all. I bought it mainly for phone/internet/music/camera etc. So, the knowledge of 2500 and beyond is just good enough for me to want to install the application. It's just that I'm amazed that I could have a program about IM or even GM's strength in iphone.

I can already promise you more than 2500 on the CCRL scale on the iPhone 4. Although my two matches are not even halfway yet, it is already clear that Stockfish 1.8 with a 15x time handicap is much stronger than Kiwi 0.6d (CCRL 2526) and weaker than Glaurung 1.2.1 (2771). It should also be kept in mind that my matches run on a 2.8 GHz i7, which is probably much faster than CCRL's Athlon X2 2.4 GHz. If someone has an X2 and can benchmark SF 1.8 there, it would be very useful.

And of course, if you use the remote engine feature, Glaurung on the iPhone will be almost as strong as Stockfish on your home computer (the "almost" part is because the remote engine does not yet support pondering, and because there is often some lag in the communication with the remote engine), i.e. way above 3000 on the CCRL scale.

[Tord]

In order to get a first rough estimate, I've now started two time handicap matches on my iMac, one with SF 1.8 vs Kiwi 0.6d (2526 on the CCRL list), the other with SF 1.8 vs Glaurung 1.2.1 (2771 on the CCRL list). In both matches, SF has 20 seconds per game, and the opponent 5 minutes per game. All engines are running with a single thread. Because 20 seconds equals 5 minutes / 15 and one CPU core of the iMac is 15 times faster than the iPhone, SF in these matches should be approximately as strong as Glaurung on the iPhone 4.

Results tomorrow.

Tomorrow is here, and the results are ready:

Stockfish - Kiwi 0.6d: 88-12 (+83,-7,=10)
Stockfish - Glaurung 1.2.1: 39-61 (+25,-47,=28)

These results put SF 1.8 on the iPhone 346 Elo points ahead of Kiwi 0.6d and 78 Elo points behind Glaurung 1.2.1. Obviously, these results don't add up, as the Elo difference between Kiwi 0.6d and Glaurung 1.2.1 on the CCRL list is only 245. If we disregard the suspiciously good result against Kiwi, we arrive at a crude estimate of about 2700 on the CCRL scale for the current iPhone Glaurung. The actual strength may be a bit higher, because my 2.8 GHz i7 is probably faster than the CCRL's 2.4 GHz Athlon X2, and the 15x time handicap I used in my matches is therefore a little too big.

I'll try to find some opponent slightly weaker than Glaurung 1.2.1 and start another test match tonight.

[Swaminathan]

In order to get a first rough estimate, I've now started two time handicap matches on my iMac, one with SF 1.8 vs Kiwi 0.6d (2526 on the CCRL list), the other with SF 1.8 vs Glaurung 1.2.1 (2771 on the CCRL list). In both matches, SF has 20 seconds per game, and the opponent 5 minutes per game. All engines are running with a single thread. Because 20 seconds equals 5 minutes / 15 and one CPU core of the iMac is 15 times faster than the iPhone, SF in these matches should be approximately as strong as Glaurung on the iPhone 4.

Results tomorrow.

Tomorrow is here, and the results are ready:

Stockfish - Kiwi 0.6d: 88-12 (+83,-7,=10)
Stockfish - Glaurung 1.2.1: 39-61 (+25,-47,=28)

These results put SF 1.8 on the iPhone 346 Elo points ahead of Kiwi 0.6d and 78 Elo points behind Glaurung 1.2.1. Obviously, these results don't add up, as the Elo difference between Kiwi 0.6d and Glaurung 1.2.1 on the CCRL list is only 245. If we disregard the suspiciously good result against Kiwi, we arrive at a crude estimate of about 2700 on the CCRL scale for the current iPhone Glaurung. The actual strength may be a bit higher, because my 2.8 GHz i7 is probably faster than the CCRL's 2.4 GHz Athlon X2, and the 15x time handicap I used in my matches is therefore a little too big.

I'll try to find some opponent slightly weaker than Glaurung 1.2.1 and start another test match tonight.

Thanks a lot, Tord. This is good enough estimate. I believe that would make Stockfish @ Iphone4 perform at the level of Slowchess, Gandalf, Aristarch and Hamsters. Those are indeed incredibly strong engines for human players.