Date: Thu, 26 Mar 1998 11:20:20 -0500 From: Omar Thameen <omar@clifford.inch.com> To: freebsd-hardware@FreeBSD.ORG Subject: Re: DPT and ECC memory Message-ID: <19980326112020.04093@clifford.inch.com> In-Reply-To: <19980323193054.62953@clifford.inch.com>; from Omar Thameen on Mon, Mar 23, 1998 at 07:30:54PM -0500 References: <19980323193054.62953@clifford.inch.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Mar 23, 1998 at 07:30:54PM -0500, I wrote: > > Is there anything special about the DPT ECC memory that they sell, [for their RAID controllers] > apart from it being ECC? DPT lists the price at over $1100 for 16M, > but my vendor is telling me that he sells 32M of 70 pin ECC memory for > around $70. Why the disparity? Here's the company line I got from DPT: DPT ECC Regular ECC ------------------------------------------------------------------------ 1. It will correct full 32 bit Based on Hamming Coding. word errors in a 512 byte page 2. Hardware implemetation - Must access data in 4 or 8 byte no overhead on Motorola chunks processor on the HBA - only when corrections needed. (not 512 byte chunks) 3. Corrects multi-bit errors Corrects only single byte. in word. 4. Overhead -3% (Overhead Has 12.5 to 21.9% overhead +system resource usage) 5. Raid Tower hot swaps can cause (sometimes) a glitch and crash the system _ DPT ECC SIMMS can prevent this. (electrical spike becomes odd bit which is taken care of by correction). I can't comment on the the claims, but at >$1100 vs. $70, we'll be going with regular ECC. -- Omar To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980326112020.04093>