Date: Thu, 12 Jul 2001 15:33:02 -0700 From: David Greenman <dg@root.com> To: Bsdguru@aol.com Cc: isp@freebsd.org Subject: Re: Intel ISP1100 or similar 1U experience with 4.3 stable Message-ID: <20010712153302.J40357@nexus.root.com> In-Reply-To: <da.8f012a9.287f7590@aol.com>; from Bsdguru@aol.com on Thu, Jul 12, 2001 at 05:50:08PM -0400 References: <da.8f012a9.287f7590@aol.com>
next in thread | previous in thread | raw e-mail | index | archive | help
>In a message dated 07/12/2001 2:07:02 PM Eastern Daylight Time, dg@root.com >writes: > >> >There is less than 1" of space above the cpu in a 1U unit, which doesnt >> leave >> >enough space for an approproate heat-sink and fan for high Mhz cpu. Yes >> they >> >will run, and may never fail, but they are not as reliable as 2U units >> which >> >allow for a more appropriate heat sink and larger fans. >> >> Uhh, that just isn't true. It's true that _most_ low profile CPU coolers >> will not cool adequately, but certainly not all of them. There are some >that >> are large but not tall that have a powerful fan and are made partly or >> completely of copper that do a fine job of cooling. CPUs in our 1U systems, >> for example, run worst case about 20C above ambient temperature, usually >> less than 10C above - i.e. about 35C, which is well below the 70C thermal >> requirement of the 1GHz FCPGA PIII and fairly close to the performace of >the >> large aluminum heat sink that Intel provides with the boxed processor. >> > >Firstly, we are talking about dual systems, or at least my original post was. >Even on the slimist fan, the clearance above the fan is too small, 1/8 to >1/4" usually. Its not enough space, no matter how powerful the fan. You'll >have hot spots, particularly near the cpus, which can cause failures over >time. > >Are you really arguing with me that a 2U enclosure is more reliable and >offers better cooling, or only that its possible to use a 1U (or in dgs case >are you just trying to justify your product...i suspect you dont have a 2U >solution to offer)? Im not sure what is wrong with you people. Very few >people cant spare the extra 2" of space,and the result is more reliable, >which was and is my point. There is zero question that 2U offers better >cooling than 1U, so what exactly are all of you home-spun "engineers" trying >to say? You're really starting to sound like a troll. I'm not a 'home-spun' engineer. I don't need to justify our products; they are built based on sound engineering and have extremely low failure rates - well below 1%. We've never had a cooling related failure. We have both 1U and 2U solutions. The cooling in the 2U solution is better than the 1U, but the 1U is well within the acceptable range. We measure both cabinet temperature and the on-die diode CPU temperature and our spec is for the CPU to operate 20C or more below the recommended maximum operating temperature of 70C. We've engineered and tested the 1U solution for dual CPUs, but don't currently offer that due to reasons that have nothing to do with cooling. What *I'm* trying to say is that, while the 2U solutions may have slightly better cooling, it doesn't matter because the 1U cooling is just fine and provides a large margin for reliable operation in idiot environments where the ambient temperature is high enough to roast any humans nearby. -DG David Greenman Co-founder, The FreeBSD Project - http://www.freebsd.org President, TeraSolutions, Inc. - http://www.terasolutions.com Pave the road of life with opportunities. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010712153302.J40357>