Collector's Corner...On The Fidelity EAG V5

[Steve B]

The 10th and final game of the match..
the matches second Sicilian with both computers in book until move 11
the V5 did manage to win a P but it was not enough to win
the game and match ended under a peaceful note

[Round "10"]
[Date "2011.6.26"]
[White "Excel MACH IV"]
[Black "EAG V5"]
[Time Control "1 Minute Per Move "]
[Result "1/2-1/2"]

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 Nc6 6.Ndb5 Bb4 7.a3 Bxc3+
8.Nxc3 d5 9.exd5 exd5 10.Bd3 Qe7+ 11.Be2 d4 12.Nb5 Bg4 13.f3 Bf5 14.g4
a6 15.gxf5 axb5 16.Rg1 O-O 17.Bh6 Ne8 18.Bf4 Nd6 19.Bxd6 Qxd6 20.Rg2
Rae8 21.Qd2 Qe5 22.O-O-O Qxf5 23.Rg5 Qf6 24.Rdg1 Re5 25.Bxb5 Rfe8
26.Rxe5 Qxe5 27.Qg2 g6 28.Bd3 Qa5 29.Qf2 Qe5 30.Kb1 Qf4 31.Qg2 Re3
32.Be4 Kh8 33.h3 f5 34.Bxc6 bxc6 35.Rf1 c5 36.Qf2 Qd6 37.Kc1 Qe6 38.Qh4
g5 39.Qg3 f4 40.Qg1 Re2 41.Kb1 Qxh3 42.Rf2 Re5 43.Ka2 Qe6+ 44.b3 Re2
45.Rxe2 Qxe2 46.Kb1 h6 47.Qg4 Qe1+ 48.Kb2 Qc3+ 49.Kb1 c4 50.Qe6 Kg7
51.Qe7+ Kg6 52.Qe6+ Kh7 53.Qe7+ Kg8 54.Qe6+ *

FINAL POSITION
[fen]6k1/8/4Q2p/6p1/2pp1p2/PPq2P2/2P5/1K6 w - - 0 54[/fen]

V5 wins the match by the lowest possible margin..5.5-4.5

interesting to think that two 68000 processors working in tandem can perhaps equalize or even surpass the computing power of the 68020

Point,Game,Set and Match Regards
Steve

[Reinfeld]

Just out of curiosity, why is it that the EAG models are so prevalent in Europe, but so elusive in the States?


- R.

[Steve B]

Just out of curiosity, why is it that the EAG models are so prevalent in Europe, but so elusive in the States?


- R.

Well i can only offer up my opinion on that but i think it does not only pertain to the EAG model in particular
i was very close to the Premiere US Retailer of chess computers here in the states for the period 1980-1995
i lived within a few miles of his shop and i purchased the vast majority of my computers from him
i can tell you he sold tons of chess computers(including EAG'S) to US customers
I think more chess computers were sold in the US then in any other country

the issue really is about the collecting hobby today in general
there are FAR more folks interested in collecting the oldies in Europe then in the US
I can think of only a handful of US collectors i know while there are many more in Europe
so in countries like Spain,France and Germany where the hobby is alive and well you see Internet web-sites,forums,on-line tournaments etc etc
its only natural then that you will see far more offering's on Ebay from these countries then from the US where the hobby is almost dead

i can tell you this...
somewhere spread throughout the entire US ..there are 100's of EAG'S,Prestiges,Tasc Units ,Saitek Renaissance boards ..etc..etc
my guess is that some day we will see these beauties being offered up on Ebay ..probably when the original owners die and their loving children scavenger through their possessions in the rush to liquidate everything and turn their dear departed's assets into cold hard cash
IMHO Regards
Steve

[Mark Uniacke]

Well the V5 made its way safely across the pond to join the collection proper
i am relieved to report that it arrived in good form

A review of the special dual-processor EAG manual reveals that the actual speed up achieved is about 1.7x and not a full doubling of speed
there is a two page letter from the Spracklens in the manual explaining the loss of .3 in speed due to various hardware overhead causes
for those interested in the technical details the 2 page letter can be found here:
http://www.flickr.com/photos/10261668@N ... 0922170604

http://www.flickr.com/photos/10261668@N ... 922170604/


there is an option to turn OFF the second processor so i preformed a simple test to see the speed up achieved
interestingly..the manual discusses that with the second processor disengaged.. the single processor runs at 98% the speed of a normal V2...
again due to hardware overhead causes
so the single processor-V5 plays a slightly weaker game then the normal V2


i set the computer for INFINITE Time
i cleared the hash memory and played these moves
1.a3 g6
2.a4

[fen]rnbqkbnr/pppppp1p/6p1/8/P7/8/1PPPPPPP/RNBQKBNR w KQkq - 0 2[/fen]
In this position:
the dual processor achieves 8 ply in 174 seconds and will play 2.Bg7
the single processor achieves 8 ply in 327 seconds and will also play 2.B
g7
That is about a 1.89x speed up
i imagine over the course of a full game the speedup would come down a bit to the 1.7x reported by the Spracklens

i dug up some old price lists and i notice that the V5 retailed for $1600 and the V2 Retailed for $1000..the V5 Only $600 more then the V2
in fact the V5 was the second cheapest EAG you could buy with all of the other versions selling for a higher price !
the V6 retailed for $1000 more then the V5
the value of a V5 versus a V6 today show an entirely different picture with two V5's recently selling for about $2500 on Ebay and V6's have sold for somewhat less then that..certainly not $1000 more
you just never know what something will be worth years from now i guess

Selective Search rates the V5 at 2051 with the V2 at 1981 and the V6 2070
Fidelity(never known to be conservative in their estimated ratings) guesstimated that the V5 would be only 50 Elo better then the V2 and 50 weaker then the V6

SS shows a 70 point increase over the V2 and only 19 less then the V6

I do plan on posting a 10 game match (colors reversing) at 1 minute per move pitting the V5 against the Designer MACH IV V6
i didn't choose the EAG V6 because getting both of those table top computers to fit on my on my desk along with my laptop would be no small feat!

Double Jeopardy Regards
Steve

Glad to see that beauty arrived safely and has played a successful match against its "more" expensive "bigger" brother.

I assume the V5 has 2 identical 68K/16Mhz CPUs to the V2?

A nice description from Dan and Kathe of the PVS algorithm they used with shared hash table and some of the problems they encountered. Indeed a x1.7 speed up would be typical of such an approach. It is testament to their abilities that they got this working so effectively in the late 1980s on a custom system as your match has so amply shown.

For me the Spracklens are legendary figures of microcomputer chess, I first encountered their program Sargon 2 on a Research Machines 380Z in 1980/81 and Hiarcs on a PDP 11/70 had the unpleasant experience of getting beaten by it at the time!

Incidently, Hiarcs' first official tournament game was against the Spracklen Challenger X operated by Tom Fürstenberg at the 1st Computer Olympiad in 1989. That might be a Challenger not in Steve's collection, it was a 68030 @ 44 Mhz, the machine being in a custom transparent plastic housing, I guess it was the Mach IV+ program. Tom told me the same machine played in Edmonton a few months earlier.

I first met Dan and Kathe in person in the offices of Sun Microsystems before the World Championship in Vancouver 1991 and of course when their Saitek Sparc X played Hiarcs in the tournament.

Nostalgic regards,
Mark

[Steve B]

Thanks Mark

YES ..two identical 68K processors as in the V2
but as i mentioned in the lead post..again due to hardware overhead issues.. each V2 in the V5 runs at 98% of the speed of the processor in the V2 .so if one processor is turned off the single processor V5 will be slightly slower/weaker then the V2
this is briefly discussed in the special dual-processor addendum to the EAG manual :

http://www.flickr.com/photos/10261668@N ... 0922170604

The 1991 World Amateur Microcomputer Chess Champion, Vancouver, Canada.
Hiarcs won that Championship
how did you do against the Sparc?

Reelin In The Years Regards
Steve

[Mark Uniacke]

Steve, thanks for the reference to that info, yes I can imagine the overhead would have some effect although 2% seems more than I would of expected.

In the game in Vancouver, it was a very close game but unfortunately Hiarcs went wrong in a double rook ending and sadly lost. However, the tournament was a great success for Hiarcs as you menitoned.

Was the Mach IV program a clear program improvement or was it really a Mach III version on a 68020?

Puzzled regards,
Mark

[Steve B]

Steve, thanks for the reference to that info, yes I can imagine the overhead would have some effect although 2% seems more than I would of expected.

In the game in Vancouver, it was a very close game but unfortunately Hiarcs went wrong in a double rook ending and sadly lost. However, the tournament was a great success for Hiarcs as you menitoned.

Was the Mach IV program a clear program improvement or was it really a Mach III version on a 68020?

Puzzled regards,
Mark

basically it was a Mach III on an 020
but not exactly..

All of the Eag Versions V2-V10 have the same exact program and the only difference is Hardware..either more memory or faster processors or multiple processors
so the EAG V2(Mach III) has the same program and opening book as the EAG V6 (Mach IV) or even the EAG V9 or 10

things get a bit sticky if we compare the Designer Mach III to the EAG Mach III because all of the EAG programs have a slightly improved program and larger book from the Designer Mach III
EAG's aside ..

if we just compare the Designer Mach III to the Designer Mach IV then the programs are identical and the only difference is the hardware..(this is the basic answer to your question)

to make matters even more confusing...
the Excel(Humpback)Mach III is slightly different then the Designer Mach III in that when Fidelity imported the program to the Designer housing they also slightly improved the Hash table program
they included the hash table improvement in the Excel Mach IV
both the Excel and Designer Mach IV's have identical programs in all respects

SO...
in Summary
Excel Mach III differs slightly from Designer Mach III(hash table program upgrade in Designer)
Designer Mach III and Designer Mach IV ..SAME Program and differ in terms of hardware only
Excel Mach III and Excel Mach IV ..difference is the hardware and the slightly improved hash table programs

all Excel and Designers Mach III or IV's differ from EAG' versions (Slightly improved program and larger book in EAG)

now if you think i remembered all of that by heart you would be giving me far too much credit

i had to look up most of the above in "Computer Chess Reports " 1990-1991


Exhausted Regards
Steve

[ricard60]

Hi Steve,
In that world chess championship of 1991 there was a mephisto x machine, was that the mephisto Vancouver?

x regards
Ricardo

[Steve B]

Hi Steve,
In that world chess championship of 1991 there was a mephisto x machine, was that the mephisto Vancouver?

x regards
Ricardo

YES
for the years 1989-1991 Lang Entered his Program under the title Mephisto X
you can see here...
http://www.grappa.univ-lille3.fr/icga/p ... php?id=202


however i think the hardware used for the X machine was different then the hardware sold in the modules
Still the program would be the same

XXX Regards
Steve

[Mark Uniacke]

Richard was using the Mephisto tournament board in Vancouver. He had a terrible start in that tournament 1/3 (a shock at the time), it was a tough tournament with the emergence of The King, SaitekX, MChess and Gideon. Ed's Risc program won the tournament, but in the commercial play-off honours were shared 2-2.

Richard was using a >50Mhz 68030.

I have the h/w details if anyone is interested.

[Mythbuster]

I think it was a 68.030 with 66 MHz and 8 MB RAM.

[Steve B]

i can tell you this...
somewhere spread throughout the entire US ..there are 100's of EAG'S,Prestiges,Tasc Units ,Saitek Renaissance boards ..etc..etc
my guess is that some day we will see these beauties being offered up on Ebay ..probably when the original owners die and their loving children scavenger through their possessions in the rush to liquidate everything and turn their dear departed's assets into cold hard cash
IMHO Regards
Steve

No Sooner did i write that and we have this offering on Ebay direct from the original 13 colonies..

http://hiarcs.net/forums/viewtopic.php?p=51583#51583

Premonition Regards
Steve

[spacious_mind]

the issue really is about the collecting hobby today in general
there are FAR more folks interested in collecting the oldies in Europe then in the US
I can think of only a handful of US collectors i know while there are many more in Europe
so in countries like Spain,France and Germany where the hobby is alive and well you see Internet web-sites,forums,on-line tournaments etc etc
its only natural then that you will see far more offering's on Ebay from these countries then from the US where the hobby is almost dead
IMHO Regards
Steve

You are right Steve, the situation is very disapointing. About 5 years ago I suggested in Ismenio's Forum (when the place was still very popular) that perhaps we should hold a tournament in the US and asked in the Forum if anyone else would be interested. Well other than Ismenio who at that time was still very active, there was no one else who had the slightest bit of interest.

I can't even find anyone here who is interested in playing an online game...

I have never played anyone in the US online in a chess computer game.

Best regards,

Nick

[Larry]

You guys think chess computer interest is dead in the US...you should try
Australia. Our dedicated chess comp fans pretty much amount to Jon
(Monsieur Plastique), his brother and myself. That's in a community of
23 million people.
Larry

[Steve B]

You guys think chess computer interest is dead in the US...you should try
Australia. Our dedicated chess comp fans pretty much amount to Jon
(Monsieur Plastique), his brother and myself. That's in a community of
23 million people.
Larry

Cmon Larry
i know Peter Parr has some interest in the oldies so right there you have one more
Up To 4 Regards
Steve