Date: Tue, 30 Jan 1996 21:45:39 +0100 From: se@zpr.uni-koeln.de (Stefan Esser) To: Eric Chet <ec0@s1.ganet.net> Cc: questions@freebsd.org Subject: Re: various questions Message-ID: <199601302045.AA04431@Sysiphos> In-Reply-To: Eric Chet <ec0@s1.GANet.NET> "Re: various questions" (Jan 30, 15:26)
next in thread | previous in thread | raw e-mail | index | archive | help
On Jan 30, 15:26, Eric Chet wrote: } Subject: Re: various questions } } > The Dirty Tag RAM and write-back cache mode ought to make } > more of a difference on a DX4 system. But I doubt you'll } > see more than some 5% to 7% speedup. } } Thanks Stefan } } I was looking at the wrong numbers. Are there any problems } using a Intel DX4(which has write-back L1) on a SP3G with write-back } L2 cache? The SP3G is said to have problems, with write-back L1 && L2. The Saturn technical data book contains a section on supported cache configurations, and explicitly states, that ONLY ONE of the caches can operate in write-back mode. This is a design limititation. In fact it seems like a primary cache is not suypported at all. I.e. it can't be used with the Pentium Overdrive nor any other CPU, AFAIK. (I.e. I never heard of anybody using a i486DX4 with WB primary cache in a motherboard with Saturn chip set). I suppose a write-back primary cache had more effect on effective throughput than a WB secondary cache, but the Saturn was faster with both caches set to write-through than most other 486 chip sets with WB support, anyway. Regards, STefan -- Stefan Esser, Zentrum fuer Paralleles Rechnen Tel: +49 221 4706021 Universitaet zu Koeln, Weyertal 80, 50931 Koeln FAX: +49 221 4705160 ============================================================================== http://www.zpr.uni-koeln.de/~se <se@ZPR.Uni-Koeln.DE>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199601302045.AA04431>