Author: Ed Schröder
Date: 23:33:33 11/11/98
Go up one level in this thread
>>He lost his temper. He was not a good loser. All true. What remains in the >>memory of people is a) Kasparov lost b) Deep Blue won c) Kasparov is >>a bad loser. So in the end he didn't do himself a favor. >>There is no damage to IBM. There is no damage to Hsu. There is only >>damage for Kasparov. >Ok, he lost his temper. But he not only said immediately after the >match, he repeated it (actually, he even enforced his statement) >several times *after*. When people from IBM tried to answer him >"We have no way to know what *exactly* happened in multi-CPU >environment under a real-time conditions", he answered "I don't >beleive you. If you'd want to do so, I'm absolutely sure you'll >find a way to do that". He could consult *any* specialist in a >concurrent programming (IMHO any CS student will be enough) before >doing that statement. >And that changes the situation, doesn't it? I tried to find the "Kasparov-Oxford" interview in my mailboxes but to no avail. As far as I remember it was on the Chessbase site. Does somebody have a copy? I like to have it again. Besides all Kasparov's allegations there was also lots of new information in the article. >Also, please note that he said "Money for new match should not come >from IBM". Who instead of IBM would arrange the match, if the only >real winner will be IBM? >IMHO Kasparov is scared to death... It's very hard to beat the beast >in his favorite tactical style, and he spends too much energy playing >strictly positional chess, so he'll be exhausted after several games. >So, current situation - IBM won, but there are a lot of opened >questions - is benefitical to both sides. >Please note that I personally respected Kasparov before the match... >I have his books at home, and he struggled - and won - against >communist system in the beginning of 80's. > >Eugene About respect.... I also don't like the way Kasparov behaved. I also think in the end besides losing the match he lost a second time because of his behavior in public. With respect I meant respect for the "chess player" Kasparov. He is the best. For that purpose IBM wanted a match against him. Not because of his character. IBM knew (or should have known) who they invited. The match was NOT against Kasparov but against the strongest chess player of that moment. That was the goal, beat the strongest chess player of the world. In that respect Kasparov has full right on a revanche. IBM by not giving Kasparov this opportunity lacks all rules of sportsmanship. If you want compete in the area of the "creme de la creme" then realize there is an etiquette to stick to. Kasparov said, "I want a re-match". IBM does not have given him that. The only thing in defence to IBM can be that Kasparov didn't really meant what he said for reasons he only knows. But this is unlikely. Below is the summarize of an interview Keith Ian Price had with Hsu. Special attention to the following points (quotes) about the chances of a re-match. Subject: a re-match. [ begin quote ] #1. He stated that no one on the team wanted to put that amount of effort into it, since the history-making goal had already been achieved. [ end quote ] Not much chances for Kasparov I would say. [ begin quote ] #3. DB was dismantled right after the match because the SP2 processors used during the match were part of an order that was shipped right after the match [ end quote ] It looks like IBM never considered to possibility to offer Kasparov a revanche. Although not related to this subject about etiquette another quote: [ begin quote ] #9. I asked him if it were possible to get scores of these games. He said he did not want to release them, as he did not want to give out any help to future competitors. [ end quote ] It's about the 10-0 against the micro's. Here is the full article.... - Ed - ------------------------------------------------------------------ Posted by Keith Ian Price on May 01, 1998 at 04:14:54: I just got back from a lecture on Deep Blue given at Oregon Graduate Institute by Feng Hsiung Hsu, and was able to ask him many questions after the presentation. Since a lot of unanswered questions have been debated on rgcc and CCC, I thought I would ask him about several of these subjects. It's late so I will mention a couple and add the rest tomorrow. 1. As to whether there will ever be a rematch or a match with someone else such as Vishy Anand or Karpov, Hsu stated that the chances are "slim to none". In his view the program is only marginally better than Kasparov at present, and in order to have a rematch, they would have to make it much better than Kasparov, so that they would be able to guarantee a win. He stated that no one on the team wanted to put that amount of effort into it, since the history-making goal had already been achieved. 2. The draw in Game two never showed up on DB's screen as a possibility. After the game, Hsu went to bed thinking that DB had played the most brilliant game ever played by a computer. When he woke up the next morning and checked the Internet, he saw "Kasparov misses draw in game 2." At first he thought, "they just won't believe that DB could win, and it's just sour grapes." But when he ran it on DB, Jr. for quite a while, the possibility showed up, but he still thought there was a way to avoid the perpetual, although to do it, would mean losing enough advantage to make it a draw anyway. He never ran the position on DB, since between-rounds maintenance was being done. 3. DB was dismantled right after the match because the SP2 processors used during the match were part of an order that was shipped right after the match. The SP2 was a brand new unit, and the upgrade orders had left none for DB's use, so the SP2s used in the match were from a lot donated by IBM to the NIH, and were shipped to them immediately following the match. The frames were kept for historical reasons. The Smithsonian has expressed interest in having Deep Blue on display some day, and that is where the frames may show up, with new SP2s installed. 4. There are no plans for a "commercial" version of Deep Blue from IBM, although Hsu is trying to get the rights to the chess processors. If he does, he may consider doing a commercially available product that would run on a PC and have a chess processor on a card to plug in. He considers it risky, though. I asked how much each unit would sell for if he were to sell 1 million. He said that with that amount of sales, he could sell the program and board for under $200. I also asked about if only 10,000 were sold. He said then it would be around $400. I told him I would buy 3 at that price, and that I suspect there are others who would also. The real problem is that he could not call it DB or even Baby Blue. IBM would not want to be associated with it if they gave him the rights. So a lot of quick advertising would not be available, which could generate the mass sales needed to get the under $200 price. Ok, it's late, and although I have a lot more, it will have to wait till tomorrow :-(. kp Here is part two of the report on Deep Blue. 5. As to the fairness of the matches between Deep Blue and Kasparov, many have brought up the point that IBM was quite secretive about giving out any games to Kasparov. As part of the presentation, Hsu addressed this by pointing out that the chess chips for DB for the first match arrived in January for the February match. DB was assembled for the first time just two weeks before the match, and so they were not likely to give out games played in that period with the match coming up, and the software not fully debugged. The new chess chips for the rematch were made in 6 months and debugged in two months, thus there was almost a half year of time before the second match, which could have afforded them time to play some games and give them to Kasparov. To explain why they did not, Hsu used the "laughingstock" rule, which he said meant that in determining what requests of Kasparov's to accept or reject, they decided that if he requested anything that would make him a "laughingstock" if the tournament were for the world championship, they would deny the request. He stated that no player vying for the world championship would release games in the six months before the championship. They generally go into seclusion, he said, and study new openings, etc., to try out in the match. Therefore, if Kasparov were playing "the number two player in the world, Vishy Anand," and he demanded that Anand give him games played in preparation, he would be a laughingstock. This, then, is why they would not release any games. I must admit that I found this to be the strangest argument he used in his presentation. I told him afterward that I felt that not to have any games at all to study put Kasparov at a distinct disadvantage, since the opening books designed for Deep Blue's rematch were based on Kasparov's games that they had available. He stated that there are many games available from Deep Thought's matches that he could study, and I asked if he seriously thought that studying the game DT lost to Fritz in '95 would really help Kasparov. He was quite adamant, and stated that Anand doesn't play the same way now that he did when he was younger, either. I left it at this point, since I had futher questions, and didn't want to push the point since he obviously felt that their team was being quite fair. To be evenhanded, I must point out that there are apparently not very many games against the full DB, anyway, since most practice games were played by Joel Benjamin against DB, Jr. According to Hsu, most of these training (for DB) games were also not actual games (40/2 tournament style); rather Benjamin would play until he got behind, and then take back moves until he returned to a relatively neutral position, and try different tacks in this manner until he found a weakness, and could get the machine into a lost position. He would then explain the weakness to Hsu, and the others, and they would tweak the evaluation until it wouldn't play the poor line anymore. So most of those games were, in the end, losses, and might have revealed a lot more than a string of regular games. 6. I asked him about whether DB had gone into "Panic mode" during move 35 of Game 2. He said that it apparently had, because it took so much time to play the move, but that he didn't know what was causing it, since he was not allowed to use the terminal during the match for purposes of finding out what was causing it to take so long. He was curious at the time, himself. I asked about Amir Ban's interpretation of the printouts, and he stated that he did not think that Amir knew how to read the outputs. He said that it was ridiculous to think that axb5 had been pondered for less than a second. In the parralel-processing method of Deep Blue that move had been being processed right along with Qb6, but never made it to the top of the list until Qb6 was rejected at the last minute. He said he did not know why it was rejected at the last minute. I asked if there were logs stored to disk as the game was played, and he said there were, but they we somewhat basic, and there was no way to log what all the chess processors were doing during that time. Therefore, he would not be able to tell from the log. He seemed to say that the only way he could have found out what was happening was to use the terminal while the move was going on, and that this was not allowed. 7. I inquired as to when he might publish some papers on the advances in computer chess achieved during his work on DT/DB, since they were moving on to other things. He said he had written an article on DB's chess processor, but perhaps it was published in the wrong place, a chip design journal. He said he had no plans to write papers on such specifics as Evaluation techniques, etc., as with 8000 separate features, individually tunable, in the Eval section of the chess processor, this would require a pretty massive book, and that there were only a handful of people in the world who would be interested. I guess that handful would be you guys. He said he is too busy writing a book on the development of the project to publish anything, and when, after he said he wouldn't know a good place to publish something like that anyway, I suggested JICCA, he said that that would not be a good place. So I wouldn't hold my breath on receiving any startling revelations from that corner. He is more interested in the hardware side of the problem, since that was his chosen field, at least this was my impression. Well, I'm going to upload this part now, before my ISP gets too busy, and I have trouble logging on. I'll try to get the rest tonight, but it may have to wait until tomorrow. kp Sorry for the delay. Here is Part III of my Deep Blue Report: 8.There was a rather long thread going on about a month ago on rgcc, concerning whether DB was an example of Artificial Intelligence. During his presentation, Hsu gave his opinion on the subject. He stated that chess is considered a game of intelligent people, and DB was able to play the game against the best player in the world, and so therefore, it could be argued that DB had passed a Turing test of sorts, albeit a chess-specific Turing Test. However, Hsu continued, he did not think that this constituted intelligence. He did not directly support his decision, but did show a cartoon that occurred just after the match, in which Kasparov is playing Deep Blue and Kasparov's foot has slipped under the plug for Deep Blue. Both have "thought balloons" showing in the cartoon. DB's balloons show a bishop with diagonal arrows; a rook with horizontal and vertical arrows; and a King with short arrows going in all directions, etc. Kasparov's balloon shows his foot lifted, and the plug out of the socket. He said this represents the difference between a chess-specific intelligence, and real intelligence. If DB were losing, it would have no way to think of a different solution outside the bounds to which it had been programmed. 9. One of the longest-running arguments on rgcc and CCC has been how well micros might fare against Deep Blue. During the Deep Blue excitement last year the news slipped out that there had been a match between DB, Jr. and Rebel 8 and Genius. DB, Jr. was supposed to have been slowed down to match the PC's speed somehow. I asked Hsu about this 10-game match. He was quite familiar with the results. He confirmed that there had been 5 games against each opponent. He stated that there was only one chess processor used, and that it's clock speed had been halved. He also said that several pruning algorithms were turned off, with some selective extensions, in order to emulate the performance of the micro hardware as much as possible. They did this to see how well they did against the micros on an evaluation specific level, keeping the speed advantage down to the difference between what the micros could evaluate given their nps levels, and what could be accomplished in the chess specific processor evaluation, rather than how many nodes were searched. Since the speed of a single chess processor is about 2-2.5 million nodes per second, and Hsu estimated that the removal of the algorithms caused a 5-10 times reduction in nodes searched, the probable nps level for DB, Jr. was somewhere between 100,000 and 250,000 with the clock speed reduction factored in. This is similar to the fast searchers, but is probably 2-5 times faster than Rebel 8 at the time. In any case, I asked how the games went, and Hsu pulled no punches. He said that the performance of the micros was much poorer than he had imagined they would be. He said all 10 games were basically blowouts. When I asked for specifics, he mentioned two examples against Rebel that had surprised him as to how little understanding they had of endgames and King safety. In the first example, the ending was with bishops of opposite color and normally would have been a draw. Rebel allowed an exchange which gave DB two widely separated passed pawns, and there was no way to stop both. Rebel did not realize until a few more moves that it was in trouble. Hsu said this was the kind of thing that is in his evaluation routines, and he was surprised that it was not in Rebel's. The second example was where DB sacrificed a Rook for a pawn next to Rebel's King. After the exchange, Hsu reported, Rebel showed 2+ pawns advantage. DB showed a .5 pawn advantage. A couple moves later, DB went up to a much higher advantage, and Rebel still showed +2. After a few more moves, Rebel suddenly realized it was busted, and dropped its eval way down. Hsu thought this was due to a minimal King safety evaluation. He did state that even with this, he thought Rebel had a much better understanding of positional play than Genius did. I asked him if it were possible to get scores of these games. He said he did not want to release them, as he did not want to give out any help to future competitors. I mentioned that he had said the chance of Deep Blue ever giving another match were almost nil, and so there should not be any future competitors. He responded that if he got the rights to the chess processors, Rebel and Genius would likely be the future competitors, and he wanted to leave his options open. I stated that even so, once released, there would be thousands of games available rather quickly, and that these 10 would not make much difference. He said that he wasn't even sure if the game scores had been saved. I realized that he was not going to let them out, so I suggested that if he found them, not to erase them, as there were a lot of people interested in them, and I moved on. 10. Since we had been talking about evaluations and positional understanding, I took the time at this point to bring up my current favorite among the chess programs I have, Chess System Tal. I stated that I was impressed with the amount it accomplished within 3000 nodes per second searched. I said that if its evaluation was able to search at a much higher rate, that I thought it would be much better than the other micro programs. I was surprised by the enthusiasm Hsu showed about the program. He mentioned that how it handled King Safety was much more similar to Deep Blue than the other micro programs, although perhaps a little bit more extreme, and that he, too, was impressed with it. He said that many things in CSTal were implemented in Deep Blue, which I found strange, since it wasn't released until after the match. I didn't think of that until later, so I wasn't able to ask him about this. Perhaps he meant implemented similarly, or perhaps ideas from CSTal's style of play exhibited in games Thorsten posted. Or maybe he had access to a beta version, or he was referring to Complete chess system, I don't really know. I only mention this as it gives a little insight into the approach used in the evaluation. During the presentation Hsu stated that unlike other chess programs, DB's evaluation in not just a matter of adding weights together with bonuses to arrive at a score, but some functions were calculated non-linearly, through multiplication, or other "second level" methods. When asked about this, he said that some examples of the non-linear evaluations were in the method of calcualating a pawn's value based on it's advance, and its position relative to other pieces and pawns, and King safety, which was an example of what he had referred to as "second level" methods. This was a question and answer section at the end of the presentation, and since it wasn't my question, I could not ask him to expand on these generalisms. 11. There has been some question as to the endgame databases used during the match. Hsu stated that there were 20 gigabytes of endgame databases from Ken Thompson and Lewis Stiller on the hard drive. He said that they were all of the five-man and down, plus selected six-man endgame databases. To his knowledge, during the match, they were never accessed, but he was not sure of this. He said that since the chess processors have some of the engame databases built in (I have read that these are the 3-man set), he figured that it never got to the point where the SP2s would need to access the hard-disk-based databases. He said that it was probably a good psychological weapon for Kasparov to know that they were there, since, if he made one wrong move during the endgame, he would know that he would quickly look foolish in front of millions of people, and this would have to have an effect. Other differing reports about how many processors DB used were also answered. Deep Blue employed 30 SP2 Scalable Processors. The frames were capable of holding sixteen each, and there were two frames, but in each frame, two processors were tied together to form a master processor, which meant a total of 30 instead of thirty-two. Each SP2 had 16 chess processors attached, so that meant a total of 480 chess processors. Up until this point I had only heard 256 or 512. Hsu said that Deep Blue used "two-level parallelism" to process positions. He described this as the method of the master processor evaluating the first 4 moves, then sending the 1000 or so positions involved to the other SP2s, which would carry it out for another 4 moves, and then turn over the positions to the chess processors, which would go on for 4-5 more moves. He said that on average DB would reach to 30 ply in considering a move, but in certain cases it had reached, through selective extensions and pruning up to 70 ply, though this was rare. It would on average process 200 million chess positions per second, but that this reached as high as 400 million in certain cases. The chess processors made for the rematch were capable of 2-2.5 million nodes per second processed, and with improved evaluation with Joel Benjamin's help, and better selective search, the speed was improved by 3-10 times over the 1996 version. I asked how many cycles it took to evaluate a position, and was told that it varied. There was a short evaluation used approximately 80% of the time which took only one cycle, and there was a long evaluation used 20% of the time that took 8 cycles. Move generation took 4 cycles. There were 8000 adjustable evaluation features, and these included such things as the value of a rook on an unopened file which could later be forced open with a pawn exchange or sacrifice. He said this was one that was added through the help of GM Joel Benjamin, and he knew of one instance during the match when it had an effect. (I have not looked over the games to see where this would be, perhaps some helpful reader with more time could find this out.) It would be very interesting to know how these evaluations can be performed in hardware, but I am not sure that this will ever be covered, especially if Hsu is really thinking of a commercial version of the program. Since he also mentioned that he would be interested to see if a single-chip chess machine could be created to beat the world champion someday, he may not be forthcoming on his research, as would be hoped. 12. Hsu evidently had difficulty in convincing the rest of the team to switch to a redesigned chess processor between the match and rematch. Since the lead time for a chess processor was normally a year for design, testing, and debug, and since they only had 1 year and three months until the rematch, they were more interested in tweaking the program in the SPs, and leaving the chips alone. Hsu said he worked for 6 months, 70-100 hours per week, redesigning the chess processors. When he had them ready, and began the tests to see how well they performed relative to the older chips, the difference was so great that the rest of the team quickly agreed to switch to the new processors and so continued on from there. Well, there was more, but this concludes the report for this forum. Most of the rest is anecdotal, and not so informative, so I will stop here. I hope it was interesting. ----------------------------------------------------------------- kp
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.