Wilco (Wilco.Dijkstra.delete@this.ntlworld.com) on July 12, 2013 4:08 am wrote:
>
> This is what Primate labs said on AnandTech forums:
>
> "John from Primate Labs here (the company behind Geekbench).
>
> I wanted to provide some details about what's going on with the floating point workloads
> the Silvermont architect referenced. Two of the Geekbench 2 floating point workloads (Sharpen
> Image and Blur Image) have a fencepost error. This error causes the workloads to read uninitialized
> memory, which can contain denorms (depending on the platform). This causes a massive drop
> in performance, and isn't representative of real-world performance.
>
> We only found out about this issue a couple of months ago. Given that Geekbench 3 will
> be out in August, and fixing the issue in Geekbench 2 would break the ability to compare
> Geekbench 2 scores, we made the call not to fix the issue in Geekbench 2.

Tough call.
Not fixing is bad, fixing is worse yet.
I'd simply add new entries, so the result sheet will show both "Sharpen Image" and "Sharpen Image (New)". May be, even two separate sections in the report, with separate aggregate scores - FP(legacy) and FP2013.

Not that I personally care. I don't even look at FP section of geekbench. It's too "microbenchy" for my liking.
Not that I have something against "microbenchy" microbenchmarks in general. In fact I like them and had written many myself. But, IMHO, "microbenchy" microbenchmarks absolutely have to be open-source and are 100% useless otherwise.

>
> If you've got any questions about this (or about anything Geekbench) please let me know and I'd be happy to
> answer them. My email address is john at primatelabs dot com if you'd prefer to get in touch that way."
>
> So I guess that explains the accidental use of denormals.
>

Yes, now it's cleared. Which does not help the outcome.

> Wilco