By: Patrick Chase (patrickjchase.delete@this.gmail.com), November 14, 2014 10:36 am
Room: Moderated Discussions
Dan Fay (danielDOTfay.delete@this.gmail.com) on November 12, 2014 12:38 pm wrote:
> IIRC, the old trace cache academic papers pushed trace caches mainly as an optimization
> to provide high instruction fetch bandwidth and to cache/store ahead of time branch
> predictions.
Yes. In particular it allows the instruction stream to contain a much higher percentage of taken branches at any given issue rate (or higher issue rate at same percentage). Dynamo had a similar intent, but in SW instead of HW.
> IIRC, the old trace cache academic papers pushed trace caches mainly as an optimization
> to provide high instruction fetch bandwidth and to cache/store ahead of time branch
> predictions.
Yes. In particular it allows the instruction stream to contain a much higher percentage of taken branches at any given issue rate (or higher issue rate at same percentage). Dynamo had a similar intent, but in SW instead of HW.