cousin_it comments on Intelligence Explosion analysis draft: Why designing digital intelligence gets easier over time

cousin_it 22 Nov 2011 23:34 UTC
8 points

For example, IBM’s Deep Blue played chess at the level of world champion Garry Kasparov in 1997 using about 1.5 trillion instructions per second (TIPS), but a program called Deep Junior did it in 2003 using only 0.15 TIPS. Thus, the power of the chess algorithms increased by a factor of 100 in only six years

I’m seeing a factor of 10...
- arundelo 23 Nov 2011 1:31 UTC
  8 points
  Parent
  Yep. The source says:
  
  Moravec claims [Deep Blue] to be equivalent to a general-purpose processor having throughput on the order of 1-3 trillion instructions per second (TIPS). [...] The host computer [that Deep Blue’s successor Deep Junior ran on] was capable of a peak throughput of approximately 15 billion instructions per second (GIPS).
  
  If we consider the Deep Blue machine to be a 1.5 TIPS machine for arithmetic convenience [....]
  
  So “only 0.15 TIPS” should have been “only 0.015 TIPS”.