[Beowulf] Power calculations , double precision, ECC and power of APU's

[Lux, Jim (337C)]

-----Original Message-----
From: Geoffrey Jacobs [mailto:gdjacobs at gmail.com] 
Sent: Monday, March 18, 2013 10:02 PM
To: Lux, Jim (337C)
Cc: beowulf at beowulf.org
Subject: Re: [Beowulf] Power calculations , double precision, ECC and power of APU's

On 03/18/2013 07:09 PM, Lux, Jim (337C) wrote:
> 
> If you were processing all the RF signals available in the HF band you might need DP.  The instantaneous dynamic range can be around 130dB within a small subband, and that's on the order of 21-22 bits, and if you look at the entire 2-30 MHz spectrum, you probably have more range than can be accommodated in the 24 bit mantissa of a SP floating point.
> 
> Jim Lux

I would be happy to be corrected, but isn't an SNR of 1000:1 considered to be excellent for a spectrograph (I'm thinking of particular datasets from Victor Blanco and CFHT). I don't believe any reduction pipeline or analysis for such data benefits from DP.


I think there are high quality CCD sensors with 80dB+ dB dynamic range (max charge vs dark current of a few electrons/pixel) 10,000:1 voltage on the sense amplifier.  Call it something like 14 bits.. So SP would be good enough for a lot of processing. (as would 32 bit fixed point.<grin>)