Re: Clarification on problem 18.9
0 views
Skip to first unread message
Shawn Blanton
Apr 5, 2013, 8:21:17 PM4/5/13
to ericbr...@cmu.edu, 18340...@googlegroups.com, 18340-stu...@googlegroups.com
Hi Eric,
the weight of the GRS bits, do they really matter in your opinion?
The remaining statements you make are reasonable however.
sb
On 4/4/2013 9:36 PM, ericbr...@cmu.edu wrote:
> Hi,
>
> Maybe this is obvious, but I wanted to get a clarification before I
> started thinking about the problem for a long time. Problem 18.9 says we
> get "an immediate 2's-complement result ... with guard, round, and sticky
> bits ... at hand." I assume this means that the normalization phase
> produced a 2's complement result that is being fed to the round stage and
> that this result has three bits past the 24 that we would normally have
> i.e. int_part.GRS where S is the or of all bits in positions with weights
> 1/8 and less?
>
> Thanks,
> Eric
>
the weight of the GRS bits, do they really matter in your opinion?
The remaining statements you make are reasonable however.
sb
On 4/4/2013 9:36 PM, ericbr...@cmu.edu wrote:
> Hi,
>
> Maybe this is obvious, but I wanted to get a clarification before I
> started thinking about the problem for a long time. Problem 18.9 says we
> get "an immediate 2's-complement result ... with guard, round, and sticky
> bits ... at hand." I assume this means that the normalization phase
> produced a 2's complement result that is being fed to the round stage and
> that this result has three bits past the 24 that we would normally have
> i.e. int_part.GRS where S is the or of all bits in positions with weights
> 1/8 and less?
>
> Thanks,
> Eric
>
Reply all
Reply to author
Forward
0 new messages