Date: Mon, 5 Apr 1999 19:32:40 -0700 (PDT) From: <unknown@riverstyx.net> To: "Dragon Knight ][" <dragonknight@dtgnet.com> Cc: FreeBSD Questions <questions@FreeBSD.ORG> Subject: Re: K6-2/333, was: Re: Debug kernel by default (was: System sizewith -g) Message-ID: <Pine.LNX.4.04.9904051917550.19515-100000@hades.riverstyx.net> In-Reply-To: <37097B00.2186EB92@dtgnet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Presumably the logic here is that it eliminates wait-states while the two
busses sync. If everything's running on the same base with different
multipliers it's easier. 33*2=66, so it's either a go, or wait one cycle.
With 100/66/33, and you're trying to move data between a 66 bus and a 100
bus at any random moment, it's either a go (1/3 of the time), wait one
cycle (1/3 of the time), or wait 2 cycles (1/3 of the time).
WARNING: cheesy ascii art approaching!
a) |----------|---------|
2 5
b) |-------|------|-----|
1 3 4
at position one, you must wait 2 cycles before you're in sync again. at
3, you wait one cycle. at 4, you don't wait. average is 1 cycle wasted
waiting. that's if your 100 wants to send. when the other side's picking
the moment, you're waiting 1 cycle at position 2, and not waiting at 5.
if you have just 33/66, the most you'll ever have to wait is one cycle
before both are in sync. that means only 1/2 cycle wasted
on average. statistically, it's better.
i'm just flailing here, if anyone can correct me, please do. this is just
cute math, i don't know how it would actually work.
---
tani hosokawa
river styx internet
On Mon, 5 Apr 1999, Dragon Knight ][ wrote:
> > While this is mathematically and theoretically sound thinking, tests
> > have shown that there is little CPU/memory performance gain with a
> > 100MHz bus. Just take a look at www.tomshardware.com. As for my own
> > systems, I run K6-2 333s at 5x66 just because it sets the PCI and AGP
> > clocks at their spec'd rate of 33 and 66MHz, respectively, while
> > providing the CPU's spec'd 333MHz.
> >
>
>
> As you say, there is a little CPU/Memory performance increase at 100MHz
> fsb.
> So I do not see your logic in setting your chips at 5x66 because it puts
> your
> PCI and AGP where they should be. One of the 'specs' of the 100MHz fsb
> is
> that PCI and AGP cards will run at their normal speed of 33 and 66MH. I
> believe
> this is also true of the 95MHz busses.
>
>
> Samuel Greear
>
>
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-questions" in the body of the message
>
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.04.9904051917550.19515-100000>