Robert Myers <rbmyersusa@[EMAIL PROTECTED]
> wrote in part:
> If hardware latency matters, you're in deep doo-doo, to quote George
Bush.
Then that is were you are, at least on some apps.
> The only thing that matters is *predicatbility*.
Yes, that is the Intel thinking since the i440BX. They have
put in hardware into their Northbridges/MCH to detect memory
access patterns and trigger speculative prefetches. Clever.
There are two problems with this: 1) not all fetches are
predictable (relational DB is particularly bad) and 2)
OS memory fragmentation by 4 KB pages hurts predictability.
Of course some apps are much worse than others (SETI@[EMAIL PROTECTED]
).
I view all apps as a combination of CPU time, bandwidth time
and latency time. Sometime those times can be overlapped.
But as CPU and bandwidth improvements drive those times down,
the unimproved latency becomes more im****tant for all apps.
Still, the situation is not at all bad for pure memory rippers.
300 CPU clocks of latency is tiny when spread out over a 4KB
page: 0.07 clk/byte. No problems for video.
> If you can't predict memory fetches with some accuracy, you're
> going to spend all your time stalled. There is no hardware fix.
Sure there is: AMD has a 30% solution with an integrated memory
controller. Ultimately, we will go to some flavor of SRAM. Yes,
it takes 2-6x the transistors per cell, but we're already having
trouble finding other things to do with the xtors.
> The energy directed at latency mostly has to go into faking
> it rather than to fixing it, and faking it works rather well,
> otherwise "modern" processors would be no better than not
> modern processors.
Guess what? For some things they aren't. For very few apps
is performance a direct function of CPU clock or mem bandwidth.
Double speed seldom yields double performance.
-- Robert
|
