From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mout.gmx.net (mout.gmx.net [212.227.17.21]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "mout.gmx.net", Issuer "TeleSec ServerPass DE-1" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 7C49421F0D3 for ; Mon, 19 Jan 2015 15:44:59 -0800 (PST) Received: from hms-beagle.home.lan ([217.237.73.95]) by mail.gmx.com (mrgmx103) with ESMTPSA (Nemesis) id 0MYfJW-1YIO2v0V86-00VOmy; Tue, 20 Jan 2015 00:44:55 +0100 Content-Type: text/plain; charset=windows-1252 Mime-Version: 1.0 (Mac OS X Mail 7.3 \(1878.6\)) From: Sebastian Moeller In-Reply-To: Date: Tue, 20 Jan 2015 00:44:53 +0100 Content-Transfer-Encoding: quoted-printable Message-Id: References: <63C60796-0C39-474C-B5A4-F2FCACC4592A@gmx.de> To: =?windows-1252?Q?Dave_T=E4ht?= X-Mailer: Apple Mail (2.1878.6) X-Provags-ID: V03:K0:AM/B+TCQ/oZwM1UoOjbPvMph94vFZ4K16M2TsZscBl+Zk5EbAQx Pt95Yfiwyfx7nBeo53cyrMGF4QXcarYdzpdt7WSmynQMtApDMx8uxHqgoIwyVZzFy1HbTnB 90FkxUN/qEvn4WJ3gZlUVew2Er/jEdTmL4cWSSZIx4zE/ymPs20l2TAFiIo6qCjtJ4c/HaS U1Ll3xbtcwofJSaTAB7Iw== X-UI-Out-Filterresults: notjunk:1; Cc: cerowrt-devel Subject: Re: [Cerowrt-devel] another contender for WNDR replacement X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: Development issues regarding the cerowrt test router project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 19 Jan 2015 23:45:28 -0000 Hi Dave, hi list, On Jan 19, 2015, at 17:37 , Dave Taht wrote: > oh, they are using a nice ethernet chip from intel? That will help. >=20 > I did feel the apu1c was improvable, well, for what it is worth the apu1c is EOL=92d and replaced by = the apu1d, no idea what the difference is, both use an AMD low = power/performance from last generation ;) > but the intel "dma to cache" > since ivy bridge feature is a tremendous advantage for intel chips. But the fitlet uses a slightly more modern AMD design (I think = its jaguar/puma instead of bobcat, whatever that means, technologically = I mean, I am on top of the biologic difference between the two ;) ) I = would not bet money that this can saturate its gigabit ethernet ports, = but I think the immediate goal is a reasonably low priced device that = allows SQM on current bandwidths <=3D 300Mbps combined. Maybe this is = it... >=20 > Sigh, more benchmarking=85. More waiting ;) I think the dual and quad ethernet devices are = not out yet =85 Best Regards Sebastian >=20 > On Mon, Jan 19, 2015 at 5:46 AM, Sebastian Moeller = wrote: >> Hi List, >> the following might just be what we need to replace the WNDR as a = hig(er)-speed capable traffic shaper: >>=20 >> = http://www.fit-pc.com/web/products/specifications/fitlet-models-specificat= ions/?model%5B%5D=3Dfitlet-B+%28TBA%29&model%5B%5D=3Dfitlet-X+%28TBA%29&mo= del%5B%5D=3Dfitlet-i+%28TBA%29 >>=20 >> The entry level model fitlet-B is estimated at $129 (requires ram and = msata), no information about the more interesting fitlet-i (2 GE ports). = The AMD Jaguar SoC is supposedly more powerfull then the AMD G series = T40E in the pc-engines api-1d ( http://www.pcengines.ch/apu1d.htm ) that = was not that hot network wise ( = http://planet.ipfire.org/post/pc-engines-apu1c-a-review ). Improvements = include twice the L2, slightly faster DRAM, slightly tweaked core, an = potentially =93turbo=94 up to 1.6Ghz from 1GHz base, intel GE chip i211 = with BQL support. Downside the intel AC wifi card most likely will not = allow proper tweaking (so maybe not suited for the make-wifi-fast = project) >> So most likely this thing will also not be able to do SQM/Cake = at 300Mbps combined or better. Would it not be sweet to finally be able = to =93retire=94 the wndrs=85 >>=20 >> Best Regards >> Sebastian >>=20 >=20 >=20 >=20 > --=20 > Dave T=E4ht >=20 > thttp://www.bufferbloat.net/projects/bloat/wiki/Upcoming_Talks