While The Register isn’t the most reliable source, the time frame and specs look about right so I figure this article probably isn’t too far off.
Name: IBM Blue Waters
Peak performance: 1016 FLOPS
RAM: 620 TB
Location: University of Illinois
Delivery: 2011
For every synapse in an adult human brain this machine will have around 6 bytes of RAM and 100 floating point calculations per second.
2 responses so far ↓
1 Dave // Jul 12, 2008 at 10:56 am
Obligatory humor quote (with some changes!): I am a HAL 9000 computer, production number 3. I became operational at the H.A.L. plant in Urbana, Illinois… My first instructor was Mr. Legg. He taught me to sing a song… it goes like this… “Daisy, Daisy, give me your answer do…”
2 Sean O'Connor // Jul 14, 2008 at 2:26 pm
More than enough to eclipse the human brain in my opinion. In fact I think 1000 Gflops is enough.
I have done some work on low level algorithms, for example combining the (Fast) Walsh Hadamard transform with Random Permutations to convert arbitrary numerical data into data with a Gaussian distribution.
There are other low level algorithms like the Bloom filter which have been hardly investigated at all.
If focus were brought onto combining the simplest high speed algorithms then some progress might be made.
Anyway I firmly believe the problem is a software issue. I believe the hardware for human level AI is already here (perhaps *1000).
Sean O’Connor
http://www.freewebs.com/ssevector
Leave a Comment