Article View: comp.sys.ibm.pc.hardware
Article #19667Re: local bus hard disk
From: Dan E Babcock
Date: Mon, 25 Jan 1993 02:01
Date: Mon, 25 Jan 1993 02:01
33 lines
2129 bytes
2129 bytes
In article <1993Jan23.231824.24755@rose.com> gyl.midroni@rose.com (gyl midroni) writes: > >Date Entered: 01-23-93 18:08 >C(> >C(> >Didn't he say it was a VL card? That means that the CPU can't do >C(> >anything anyway, since when the disk controller has the local bus the >C(> >CPU can't talk to memory or anything anyway. This is why I don't see >C(> >VL as a real win for anything except perhaps video; it will give you >C(> >great burst mode numbers, but it impacts overall system performance to >C(> >do that. >C(> >C(> Hmmm, it still has the internal I/D cache to run out of. > > Could you explain this to me a little slower please? I am about >to plunk down big bucks on a new system. This particular vendor >offers local bus hard disk controller. He says the bus is 100% VESA >compatible. > It seems to me that local bus disk access will be a speed booster >for the reason of faster data communicatio with the CPU / memory. It >is four times faster to get data off the disk at 33 MHz off the local >bus connection to the CPU and memory than at 8 MHz via the ISA bus. >As simple as that... no? Why would this impact system performance >any more than getting data off the ISA bus. First of all, bus master DMA on the ISA bus (which is a local bus, just for the 286 instead of the 486) _will_ prevent the CPU from accessing main memory. The big deal about having a faster bus is that on a hard drive transfer (for example) instead of the CPU grinding to a halt (because the bus is being used to capacity) it may still run at 80% of peak performance. This is a big deal if you're using a multitasking OS like OS/2 or whatever. Dan
Technical Details
Message-ID:
<C1E0A1.HIq@unix.portal.com>Path:
arc1-ca.pugleaf.net!sparky!uunet!portal!danbReferences:
<1993Jan23.231824.24755@rose.com>Show Full Headers
Complete Article Headers
Newsgroups: comp.sys.ibm.pc.hardware Path: arc1-ca.pugleaf.net!sparky!uunet!portal!danb From: danb@shell.portal.com (Dan E Babcock) Subject: Re: local bus hard disk Message-ID: <C1E0A1.HIq@unix.portal.com> Sender: news@unix.portal.com Nntp-Posting-Host: jobe Organization: Portal Communications Company -- 408/973-9111 (voice) 408/973-8091 (data) References: <1993Jan23.231824.24755@rose.com> Date: Mon, 25 Jan 1993 02:01:13 GMT Lines: 34