Date: Tue, 15 Dec 2009 17:01:48 +0200 From: Alexander Motin <mav@FreeBSD.org> To: Wes Peters <barnaclewes@gmail.com> Cc: freebsd-current@freebsd.org, Norikatsu Shigemura <nork@freebsd.org> Subject: Re: How do I use NCQ of Intel X25-E(SSD) on ahci(4)? Message-ID: <4B27A4DC.7070908@FreeBSD.org> In-Reply-To: <f83770800912150136u1fbabd2ak7b0cd15f85c0099f@mail.gmail.com> References: <20091119004651.7432a6e4.nork@FreeBSD.org> <f83770800912150136u1fbabd2ak7b0cd15f85c0099f@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Wes Peters wrote: > On Wed, Nov 18, 2009 at 7:46 AM, Norikatsu Shigemura <nork@freebsd.org> wrote: >> I have a Intel 64GB SSD(X25-E) and a Western Digital Caviar Green >> 1TB HDD (WD10EADS), and use them on ahci(4). ahci(4) can use NCQ >> of WD10EADS, but doesn't use NCQ of X25-E. > > How useful is NCQ on an SSD? Isn't the primary thrust of NCQ to allow > the disk to reorder writes to optimize the seek pattern? Many SSDs have several flash channels. Multiple parallel requests running allow controller to effectively use them in parallel. On my OCZ Vertex I am receiving about 30% speedup from using NCQ on many short read requests. If you look into Intel SSD's datasheet, you'll also see that all benchmarks are provided for maximal request queue. -- Alexander Motin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4B27A4DC.7070908>