World Class PC, Mac and Palm Chess Software on
your desktop and in the Palm of your hand!
HIARCS 10
The new PC HIARCS experimental prototype is now in testing.
It has been two long years since the release of HIARCS 9 for PC. Many
people have asked why has it taken so long...
Since its first release in December 1992 HIARCS chess software has always
tried to provide value for money. I have always taken the view that each new
version should be at least 40 Elo stronger than its predecessor. Unlike many
others in the industry I have not cashed in on yearly upgrades even when
some progress has been made. Instead I have continued to develop the
application until the improvement between versions is so significant that it
has to be released. It is worth reflecting on the HIARCS
progress made in chess strength over the years since the very first
commercial HIARCS release:
HIARCS 1 (1992)
HIARCS 2 (1993)
HIARCS 3 (1994)
HIARCS 4 (1995)
HIARCS 6 (1997)
HIARCS 7 (1998)
HIARCS 8 (2002)
HIARCS 9 (2003)
HIARCS 9.6 (2004)
Elo improvement
N/A
130
58
20
41
51
58
56
38
You can see from the above chart and table that the improvements have
been significant. There was a disappointment at HIARCS 4 due to the
introduction of learning. Despite this the average improvement in Elo between versions
has been 56.4 Elo. Based on this information Hiarcs 10 will not disappoint!
After numerous tests HIARCS 9.6, released for Apple Macintosh in January
2005, has been shown to have improved on Hiarcs 9 by about 40 Elo. Now I
know a lot of PC users wanted to see a new HIARCS 10 at this stage. I had
basically achieved my minimum required chess strength improvement and I did start to
investigate a possible release. I had been experimenting with some other
ideas in December 2004/January 2005 and I wanted to see if I could make it
work as it offered another step forward. With that in mind I tentatively
hoped to release in early March 2005. However, I was not successful in
resolving some problems with the implementation. Outside events took over
and I was not able to work on HIARCS during the period March and April 2005. This gap gave me time to reflect and I realised where the
implementation was going wrong and could be improved. This lead to further
significant improvements which I developed during May to July 2005.
HIARCS emerged from this around 20% slower(!) in nodes per second than
Hiarcs 9, but much much stronger. HIARCS now looks very deeply into positions
and its blitz, rapid and tournament rate play has improved tremendously.
Results were starting to impress all our test team and refinement of the new
program began. The results have steadily improved even further with comments
like sensational and brilliant being received from those who have seen and
tested it.
We are still in the refinement phase though - noticing ways the chess
engine can be improved for maximum performance. It is clear that the new
prototype is far stronger than both Hiarcs 9.6 and Hiarcs 9.0. I would love
to release it now, but more needs to be done to make it the complete
package.
I will next update this page with information about the new HIARCS' style
and opening book development.
"I've been in the chess
computer software biz for over a decade and I can tell you that I have never
heard more raves from users over any other chess program than I've heard over Hiarcs.
Users love this chess engine. I've received literally scores of
highly favorable e-mails and comments from happy Hiarcs owners over the
years. If you never buy another chess engine, you owe it to yourself to add Hiarcs
to your stable of chess playing programs. Playing against Hiarcs is very
much like playing against a strong human player." - Hiarcs 9 review
by Steve Lopez 17/10/2003
"HIARCS is much better positionally than Deep Blue." - Garry
Kasparov 1997 (after game 5 of the rematch)