SPARC minimum hardware specification
Hello, I never used a SPARC machine but I recall there are some people on the list doing this. What are the minimum requirements for a "decent" SPARC machine? I mean by that a machine who is able to run OpenBSD as a desktop. I am currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my needs. Tell me please the CPU or the machine name, I will search the prices :-). Thanks
Re: SPARC minimum hardware specification
> > From: Mihai Popescu > Sent: Thu Jul 16 20:09:30 CEST 2015 > To: > Subject: SPARC minimum hardware specification > > > Hello, > > I never used a SPARC machine but I recall there are some people on the > list doing this. > > What are the minimum requirements for a "decent" SPARC machine? I mean > by that a machine who is able to run OpenBSD as a desktop. I am > currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my > needs. Tell me please the CPU or the machine name, I will search the > prices :-). > > Thanks > Hello Getting a cool little sparc machine is a nice project. They are quite expensive compared to their setup & a little bit hard to find, if you're looking for a specific model. I think you may not try to get a too old machine (ultras2/60/80) they are mythic, they are cool but too rare to be cheap. and not so much powerfull ... for cheaper price : Maybe somme v4xx or v2xx should be a good choice for budget...like v490 this one should be easy to find : maybe a t5220 (8 cores 16Gb ram & 1 disk should be about 250$/300$ max.. you may know they make big noise (at boot time in particular). & ask lot power I hope it can help you... Cordialement Francois Pussault 10 chemin de négo saoumos apt 202 - bat 2 31300 Toulouse +33 6 17 230 820 +33 5 34 365 269 fpussa...@contactoffice.fr
Re: SPARC minimum hardware specification
> What are the minimum requirements for a "decent" SPARC machine? I mean > by that a machine who is able to run OpenBSD as a desktop. I am > currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my > needs. Tell me please the CPU or the machine name, I will search the > prices :-). I doubt any sparc hardware would fit. Not even a Sun Blade 2500.
Re: SPARC minimum hardware specification
On 2015-07-16 Thu 21:09 PM |, Mihai Popescu wrote: > > I never used a SPARC machine but I recall there are some people on the > list doing this. > Platform Specific Lists sp...@openbsd.org OpenBSD/sparc and OpenBSD/sparc64 ports http://www.openbsd.org/mail.html
Re: SPARC minimum hardware specification
On Thu, Jul 16, 2015 at 08:25:20PM +0200, Francois Pussault wrote: >> What are the minimum requirements for a "decent" SPARC machine? I mean >> by that a machine who is able to run OpenBSD as a desktop. <...> >for cheaper price : >Maybe somme v4xx or v2xx should be a good choice for budget...like v490 I hope you are making a joke. You definately don't want to use such a noisy beast as a desktop. -- Maurice
Re: SPARC minimum hardware specification
On Thu, Jul 16, 2015 at 8:09 PM, Mihai Popescu wrote: > What are the minimum requirements for a "decent" SPARC machine? I mean > by that a machine who is able to run OpenBSD as a desktop. I am > currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my The problem is that Intel is really on top with CPU performance -- well let's not-consider POWER8 for now. This P4 @ 3.2 is quite powerful, I'm even not sure if decent SPARC64 like SPARC64 VII+ 2.7 GHz will get to your P4 performance on single-threaded. It'll probably wipe the flour with it on multi-threading benchmark especially if you consider 4 core CPU which provides 2 threads per core. So well, you would like to know my dream machine? SPARC M3000. One of the most silent (47 dBA!), but yet powerful SPARCs. Unfortunately see prices on eBay. No way... If you do have server room, then M4000 is way much affordable but be prepared for big power consumption (around 1kW) and noise is on 59 dBA and weight? ~80kg... The advantage of SPARC64 is that it's way more powerful than even the fastest UltraSPARC built into Ultra45 which was the last Sun's dual-cpu SPARC workstation. Again see prices. More affordable are Blade 1500/2500 silver/red with nearly the same cpu like last Ultra's 25/45. Well, check prices for yourself. Have fun! Karel
Re: SPARC minimum hardware specification
> > From: Maurice Janssen > Sent: Thu Jul 16 22:55:05 CEST 2015 > To: > Subject: Re: SPARC minimum hardware specification > > > On Thu, Jul 16, 2015 at 08:25:20PM +0200, Francois Pussault wrote: > >> What are the minimum requirements for a "decent" SPARC machine? I mean > >> by that a machine who is able to run OpenBSD as a desktop. > <...> > >for cheaper price : > >Maybe somme v4xx or v2xx should be a good choice for budget...like v490 > > I hope you are making a joke. You definately don't want to use such > a noisy beast as a desktop. > > -- > Maurice > Ironic... It's possible but as you say it's just mad to do so Cordialement Francois Pussault 10 chemin de négo saoumos apt 202 - bat 2 31300 Toulouse +33 6 17 230 820 +33 5 34 365 269 fpussa...@contactoffice.fr
Re: SPARC minimum hardware specification
Hi, Mihai Popescu wrote: What are the minimum requirements for a "decent" SPARC machine? I mean by that a machine who is able to run OpenBSD as a desktop. I am currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my needs. Tell me please the CPU or the machine name, I will search the prices:-). if "desktop" includes "browsing" the answer is "none". None that can won't be too big or too noisy. Your P4 is still a mighty machine as a CPU, but it is single-core and has a terribly long pipeline too: today for browsing and most web-based stuff it is just not enough, even a slower, but dual-core cpu works better. And you need a decent graphics card too. The only non-intel machines that could be used as a desktop were PPC macs, but even those suffer today. For all the rest, including development, compilation, email, newsgroup... even a decently fitted Ultra workstation could do. I used to have "non intel" hardware for everything, but today it is not feasible anymore. You just need an intel laptop for that convenience. Not only for the power, but also because Firefox & friends often show bugs on non-intel. Furthermore if you do need Skype or other proprietary crap, you need Linux at least or directly Windows/Mac. I love all my alternative architectures and OS's... but I need to have a standard piece of hardware handy. A pity, but for my even basic needs it is that way. SSH to your sparc box, export display. Riccardo
Re: SPARC minimum hardware specification
On Fri, Jul 17, 2015 at 1:03 PM, Riccardo Mottola wrote: > The only non-intel machines that could be used as a desktop were PPC macs, > but even those suffer today. Do you have any experience with SPARC64 VI/VII? I'm curious since I've thought this is the only chip which can give some hope for this so I'm curious if I've been wrong on this. Thanks, Karel
Re: SPARC minimum hardware specification
On Thu, 16 Jul 2015 21:09:30 +0300 Mihai Popescu wrote: > Hello, > > I never used a SPARC machine but I recall there are some people on the > list doing this. > > What are the minimum requirements for a "decent" SPARC machine? I mean > by that a machine who is able to run OpenBSD as a desktop. I am > currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my > needs. Tell me please the CPU or the machine name, I will search the > prices :-). > > Thanks > Hello Misc, As a new user, I find myself in the same position as the OP: very interested in non-Intel products. But there seems to be a vacuum of information around this topic. (Which tells me I'm on the right track) No operating system can protect itself from its own hardware. So how does one protect one's family in these hostile and trying times and what-not? The replies to the OP seem discouraging. If not Oracle, and not Fujitsu, then what? If not a sparc desktop, then what about a sparc router? A RISC anything?? I have yet to achieve that warm fuzzy feeling, and I fear that it is not within my reach. Long live Puffy! All the best, Keith Larsen CPS Coatings 318-222-6100
Re: SPARC minimum hardware specification
Have you seen TemLib[1]? That's may be the way for SPARC, otherwise MIPS64/Octeon or ARM. Good luck! Karel [1]: http://temlib.org/site/ On Fri, Jul 17, 2015 at 6:15 PM, BSD wrote: > On Thu, 16 Jul 2015 21:09:30 +0300 > Mihai Popescu wrote: > >> Hello, >> >> I never used a SPARC machine but I recall there are some people on the >> list doing this. >> >> What are the minimum requirements for a "decent" SPARC machine? I mean >> by that a machine who is able to run OpenBSD as a desktop. I am >> currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my >> needs. Tell me please the CPU or the machine name, I will search the >> prices :-). >> >> Thanks >> > > Hello Misc, > > As a new user, I find myself in the same position as the OP: very > interested in non-Intel products. But there seems to be a vacuum of > information around this topic. (Which tells me I'm on the right track) > > No operating system can protect itself from its own hardware. So how > does one protect one's family in these hostile and trying times and > what-not? > > The replies to the OP seem discouraging. If not Oracle, and not > Fujitsu, then what? If not a sparc desktop, then what about a sparc > router? A RISC anything?? > > I have yet to achieve that warm fuzzy feeling, and I fear that it is > not within my reach. Long live Puffy! > > > All the best, > > Keith Larsen > CPS Coatings > 318-222-6100
Re: SPARC minimum hardware specification
On 2015-07-17, BSD wrote: > As a new user, I find myself in the same position as the OP: very > interested in non-Intel products. But there seems to be a vacuum of > information around this topic. You're 15 years too late. x86 has won. -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: SPARC minimum hardware specification
> On Jul 17, 2015, at 1:45 PM, Christian Weisgerber wrote: > > On 2015-07-17, BSD wrote: > >> As a new user, I find myself in the same position as the OP: very >> interested in non-Intel products. But there seems to be a vacuum of >> information around this topic. > > You're 15 years too late. x86 has won. I couldn’t have said it better. To my knowledge - the fastest non-x86 desktop type system that OpenBSD runs on in the PowerMac G5, it’s 10 years old, and a nicely specked can cost almost as much as a much faster x86 box. Things like the BeagleBones and Edge Routers are great for what they are - be it a router or embedded platform or simple web server, etc. - but the modern desktop environment and things like firefox are just to slow on anything non intel esp. w/o accelerate graphics. Just how it is. -bp @creepingfur
Re: SPARC minimum hardware specification
On Fri, Jul 17, 2015 at 08:45:23PM +, Christian Weisgerber wrote: > On 2015-07-17, BSD wrote: > > > As a new user, I find myself in the same position as the OP: very > > interested in non-Intel products. But there seems to be a vacuum of > > information around this topic. > > You're 15 years too late. x86 has won. Generally yes but there's still POWER8 and OpenPOWER group. They claim docs should be open and TYAN[1] has a cheap server with POWER8. IBM won't kill this HW as they invested money to have OpenStack running on it, they also ported KVM on POWER8. [1] http://www.tyan.com/solutions/tyan_openpower_system.html j.
Re: SPARC minimum hardware specification
> On Fri, Jul 17, 2015 at 08:45:23PM +, Christian Weisgerber wrote: > > On 2015-07-17, BSD wrote: > > > > > As a new user, I find myself in the same position as the OP: very > > > interested in non-Intel products. But there seems to be a vacuum of > > > information around this topic. > > > > You're 15 years too late. x86 has won. > > Generally yes but there's still POWER8 and OpenPOWER > group. They claim docs should be open and TYAN[1] has > a cheap server with POWER8. > > IBM won't kill this HW as they invested money to have > OpenStack running on it, they also ported KVM on POWER8. > > [1] http://www.tyan.com/solutions/tyan_openpower_system.html I suspect your diff was MIME encoded, and therefore got removed... Please be realistic.
Re: SPARC minimum hardware specification
On Fri, Jul 17, 2015 at 11:29 PM, Jiri B wrote: > Generally yes but there's still POWER8 and OpenPOWER > group. They claim docs should be open and TYAN[1] has > a cheap server with POWER8. I would not bet on POWER8 on desktop, it's too big beast, rather on smaller ARMv8 which is already winning on mobile...
Re: SPARC minimum hardware specification
On Fri, 17 Jul 2015 09:15:14 -0700, BSD wrote: The replies to the OP seem discouraging. If not Oracle, and not Fujitsu, then what? If not a sparc desktop, then what about a sparc router? A RISC anything?? You might be interested in Bunny's Novena project [1] [2] [1] http://www.mail-archive.com/misc%40openbsd.org/msg126490.html [2] https://www.crowdsupply.com/sutajio-kosagi/novena
Re: SPARC minimum hardware specification
If I’m not mistaken the PS3 had a PPC as well. Many of the intel alternatives do a better job at math & calculations in my opinion while PPC & SPARC may need additional time to execute operations often there are fewer errors and the results are far more accurate. This quite obvious on the PS3. Things like X & Y coordinate mappings were often very accurate. Seemingly intel aims to disregard this type of accuracy. Regards Patrick > On Jul 18, 2015, at 9:19 AM, Seth wrote: > > On Fri, 17 Jul 2015 09:15:14 -0700, BSD wrote: >> The replies to the OP seem discouraging. If not Oracle, and not >> Fujitsu, then what? If not a sparc desktop, then what about a sparc >> router? A RISC anything?? > > You might be interested in Bunny's Novena project [1] [2] > > [1] http://www.mail-archive.com/misc%40openbsd.org/msg126490.html > > [2] https://www.crowdsupply.com/sutajio-kosagi/novena
Re: SPARC minimum hardware specification
On Fri, Jul 17, 2015 at 11:15:14AM -0500, BSD wrote: > On Thu, 16 Jul 2015 21:09:30 +0300 > Mihai Popescu wrote: > > > Hello, > > > > I never used a SPARC machine but I recall there are some people on the > > list doing this. > > > > What are the minimum requirements for a "decent" SPARC machine? I mean > > by that a machine who is able to run OpenBSD as a desktop. I am > > currently use a Pentium 4 3.2GHz with 2 GB DDR and it barely meets my > > needs. Tell me please the CPU or the machine name, I will search the > > prices :-). Sun Fire servers are cheap to buy but not to run. A V210 is a 1U box and with dual 1.35 CPUs it is fast enough for desktop use. It's not something most people with families or without a flightline headset are going to want sitting next to their desk though and you will need some air conditioning. You didn't say much about your needs so it's hard to know why your current setup isn't satisfactory but 2G of RAM is often not enough for a desktop these days. All it takes is a fairly recent version of Firefox with a bunch of tabs open and a few big PDFs open in Acroreadto use up that amount of RAM. > The replies to the OP seem discouraging. If not Oracle, and not > Fujitsu, then what? If not a sparc desktop, then what about a sparc > router? A RISC anything?? OpenBSD mips64el runs oustandingly well on the Lemote boxes. See here: http://www.openbsd.org/loongson.html I don't think anybody will be happy with a Loongson as a desktop box but they do shine tiny servers. /jl -- ASCII ribbon campaign ( ) Powered by Lemote Fuloong against HTML e-mail X Loongson MIPS and OpenBSD and proprietary/ \http://www.mutt.org attachments / \ Code Blue or Go Home! Encrypted email preferred PGP Key 2048R/DA65BC04
Re: SPARC minimum hardware specification
So what are good mips or arm motherboards nowadays? I have an openbsd box at home I need to upgrade. Might as well take a look at non-x86 stuff, as long as they can take SATA... :)
Re: SPARC minimum hardware specification
On Sun, Jul 19, 2015 at 11:03 AM, John Long wrote: > Sun Fire servers are cheap to buy but not to run. A V210 is a 1U box and > with dual 1.35 CPUs it is fast enough for desktop use. It's not something > most people with families or without a flightline headset are going to want > sitting next to their desk though and you will need some air conditioning. Could you clarify this? http://www.andovercg.com/datasheets/sun-fire-v210-server.pdf Suggests that we're talking 320 watts, and 7.3 db acoustic noise. http://www.gcaudio.com/resources/howtos/loudness.html suggests that whispers are about 30 db. And, 320 watts is not too far from what you would expect from a 3 bulb light fixture. Gamer rig PCs can run hotter than that. So perhaps I'm looking at the wrong information? Thanks, -- Raul
Re: SPARC minimum hardware specification
On 2015-07-19, John Long wrote: > OpenBSD mips64el runs oustandingly well on the Lemote boxes. See here: > http://www.openbsd.org/loongson.html Given that only about 2/3 of the ports tree can be built on loongson, I'm questioning this "outstandingly well". -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: SPARC minimum hardware specification
2015-07-19 17:03 GMT+02:00 John Long : > > OpenBSD mips64el runs oustandingly well on the Lemote boxes. See here: > http://www.openbsd.org/loongson.html > > I don't think anybody will be happy with a Loongson as a desktop box but > they do shine tiny servers. > > /jl Where could one acquire one of the machines mentioned on the link ? I've seen no option to buy one on their site and I got no luck on ebay. -- Cordialement, Coues Ludovic +336 148 743 42
Re: SPARC minimum hardware specification
On Sun, Jul 19, 2015 at 7:51 PM, Raul Miller wrote: > On Sun, Jul 19, 2015 at 11:03 AM, John Long wrote: >> Sun Fire servers are cheap to buy but not to run. A V210 is a 1U box and >> with dual 1.35 CPUs it is fast enough for desktop use. It's not something >> most people with families or without a flightline headset are going to want >> sitting next to their desk though and you will need some air conditioning. > > Could you clarify this? > > http://www.andovercg.com/datasheets/sun-fire-v210-server.pdf > > Suggests that we're talking 320 watts, and 7.3 db acoustic noise. 320 watts should not be showstopper, but 7.3 db certainly is. I do have T2000 which is ~62 dB(A) and this is simply horrible while being in the same room. I'm using ear protectors like man uses while working with chainsaw and still after few hours I'm head is going in circles...But to be honest I'm probably over-sensitive to level of noise so you may be in different situation. > http://www.gcaudio.com/resources/howtos/loudness.html suggests that > whispers are about 30 db. Pay attention especially to: Perceptions of Increases in Decibel Level Imperceptible Change 1dB Barely Perceptible Change 3dB Clearly Noticeable Change 5dB About Twice as Loud 10dB About Four Times as Loud 20dB Anyway those boxes are cheap these days, just win one on ebay and enjoy. :-) But if you are going this noisy route, then I would certainly recommend SPARC64 VI/VII(+) box. If you are lucky it may be for the same price and the cpu should be way faster then ordinary 1.3GHz UltraSPARC -- just search thorough spec.org.
Re: SPARC minimum hardware specification
On Sun, Jul 19, 2015 at 09:56:35PM +0200, Karel Gardas wrote: > On Sun, Jul 19, 2015 at 7:51 PM, Raul Miller wrote: > > Suggests that we're talking 320 watts, and 7.3 db acoustic noise. > > 320 watts should not be showstopper, but 7.3 db certainly is. I do > have T2000 which is ~62 dB(A) and this is simply horrible while being > in the same room. Indeed. If you're a "my neighbour listens to heavy metal all day, whether he wants to or not" kind of person, then these machines are for your home. Else, you'll need to find a rack in some data centre to put them in. My T1000 is louder than my hoover turned all the way up to full blast. It spent just enough time in my flat to be ready for remote install once mounted in its rack.
Re: SPARC minimum hardware specification
> Indeed. If you're a "my neighbour listens to heavy metal all day, whether > he wants to or not" kind of person, then these machines are for your home. Well, ny neighbours listen to heavy metal every day whether they want it or not, but I'd hate for a loud computer to impair that heavy metal experience.
Re: SPARC minimum hardware specification
On 19.07.2015. 19:51, Raul Miller wrote: http://www.andovercg.com/datasheets/sun-fire-v210-server.pdf Suggests that we're talking 320 watts, and 7.3 db acoustic noise. http://www.gcaudio.com/resources/howtos/loudness.html suggests that whispers are about 30 db. And, 320 watts is not too far from what you would expect from a 3 bulb light fixture. Gamer rig PCs can run hotter than that. So perhaps I'm looking at the wrong information? Info is all right, but is says 7.3 B = 73 dB = very loud.
Re: SPARC minimum hardware specification
> On Jul 20, 2015, at 1:53 AM, Graham Stephens > wrote: > > On 19/07/2015 20:56, Karel Gardas wrote: >> On Sun, Jul 19, 2015 at 7:51 PM, Raul Miller wrote: >>> On Sun, Jul 19, 2015 at 11:03 AM, John Long wrote: Sun Fire servers are cheap to buy but not to run. A V210 is a 1U box and with dual 1.35 CPUs it is fast enough for desktop use. It's not something most people with families or without a flightline headset are going to want sitting next to their desk though and you will need some air conditioning. >>> >>> Could you clarify this? >>> >>> http://www.andovercg.com/datasheets/sun-fire-v210-server.pdf >>> >>> Suggests that we're talking 320 watts, and 7.3 db acoustic noise. >> >> 320 watts should not be showstopper, but 7.3 db certainly is. I do >> have T2000 which is ~62 dB(A) and this is simply horrible while being >> in the same room. I'm using ear protectors like man uses while working >> with chainsaw and still after few hours I'm head is going in >> circles...But to be honest I'm probably over-sensitive to level of >> noise so you may be in different situation. >> >>> http://www.gcaudio.com/resources/howtos/loudness.html suggests that >>> whispers are about 30 db. >> >> Pay attention especially to: >> >> Perceptions of Increases in Decibel Level >> Imperceptible Change 1dB >> Barely Perceptible Change 3dB >> Clearly Noticeable Change 5dB >> About Twice as Loud 10dB >> About Four Times as Loud 20dB >> >> Anyway those boxes are cheap these days, just win one on ebay and >> enjoy. :-) But if you are going this noisy route, then I would >> certainly recommend SPARC64 VI/VII(+) box. If you are lucky it may be >> for the same price and the cpu should be way faster then ordinary >> 1.3GHz UltraSPARC -- just search thorough spec.org. >> > > Another thing to bear in mind is the pitch of the noise; I find that loudish > but low-frequency sound (like from 4-inch+ fans) isn't that uncomfortable, > but the whine from 1U 1 inch fans get unbearable REALLY quickly. A prime example of this is the Sun T1000 or the 1U T2 boxes - those fans screech and seem to cut thru any sort of noise or sound deadening at a pitch which is almost painful. Awful stuff. The v2xx boxes aren’t nearly as bad, in comparison - but are still rather loud, especially if warm.
Re: SPARC minimum hardware specification
Let's cut through some of this crap. If you want a reasonably quiet sparc64 designed to be put on a desk, your fastest choices are a Sun Blade 100 or Sun Blade 150 (~20% faster) Of course these machines are 15+ years old and something like Pentium II speed. There are also Sun Blade [12]xxx models that are more powerful, but they don't get any quieter. The Blade 150 is now my noisiest machine at home. My Xeon-based Dell PowerEdge T20 is quieter. -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: SPARC minimum hardware specification
On Mon, Jul 20, 2015 at 09:53:13AM +0100, Graham Stephens wrote: > Another thing to bear in mind is the pitch of the noise; I find that > loudish but low-frequency sound (like from 4-inch+ fans) isn't that > uncomfortable, but the whine from 1U 1 inch fans get unbearable > REALLY quickly. I agree with this. The 1U units are very shrill and annoying. I have 4U machines and they are louder but somehow easier to tolerate. You get to choose boiling whistling tea kettle or Hoover... /jl -- ASCII ribbon campaign ( ) Powered by Lemote Fuloong against HTML e-mail X Loongson MIPS and OpenBSD and proprietary/ \http://www.mutt.org attachments / \ Code Blue or Go Home! Encrypted email preferred PGP Key 2048R/DA65BC04
Re: SPARC minimum hardware specification
Yeah, I misread that. Thanks, -- Raul On Sun, Jul 19, 2015 at 10:04 PM, Zeljko Jovanovic wrote: > On 19.07.2015. 19:51, Raul Miller wrote: > >> http://www.andovercg.com/datasheets/sun-fire-v210-server.pdf >> >> Suggests that we're talking 320 watts, and 7.3 db acoustic noise. >> >> http://www.gcaudio.com/resources/howtos/loudness.html suggests that >> whispers are about 30 db. And, 320 watts is not too far from what you >> would expect from a 3 bulb light fixture. Gamer rig PCs can run hotter >> than that. >> >> So perhaps I'm looking at the wrong information? > > > > Info is all right, but is says 7.3 B = 73 dB = very loud.
Re: SPARC minimum hardware specification
On Sun, Jul 19, 2015 at 01:51:34PM -0400, Raul Miller wrote: > On Sun, Jul 19, 2015 at 11:03 AM, John Long wrote: > > Sun Fire servers are cheap to buy but not to run. A V210 is a 1U box and > > with dual 1.35 CPUs it is fast enough for desktop use. It's not something > > most people with families or without a flightline headset are going to want > > sitting next to their desk though and you will need some air conditioning. > > Could you clarify this? > > http://www.andovercg.com/datasheets/sun-fire-v210-server.pdf > > Suggests that we're talking 320 watts, and 7.3 db acoustic noise. The power figure is correct but I guess the noise must be 73 db? I would guess the average power consumption of a V210 would be higher than that of your P4 but I haven't tested it. /jl -- ASCII ribbon campaign ( ) Powered by Lemote Fuloong against HTML e-mail X Loongson MIPS and OpenBSD and proprietary/ \http://www.mutt.org attachments / \ Code Blue or Go Home! Encrypted email preferred PGP Key 2048R/DA65BC04
Re: SPARC minimum hardware specification
On Sun, Jul 19, 2015 at 05:59:17PM +, Christian Weisgerber wrote: > On 2015-07-19, John Long wrote: > > > OpenBSD mips64el runs oustandingly well on the Lemote boxes. See here: > > http://www.openbsd.org/loongson.html > > Given that only about 2/3 of the ports tree can be built on loongson, > I'm questioning this "outstandingly well". I tested my first Lemote Mini by doing about 5 complete builds back to back of OpenBSD over a period of a couple of weeks with no failures and no issues of any kind. The box remained nearly silent and was cool to the touch the whole time. Whether LinTel apps compile may be one thing but OpenBSD certainly runs outstandingly well on these boxes. I have one setup as an FTP server and it pushes 9+ MB/s out the door with the stock 160G SATA. I think for what you get for how much you pay these boxes are a screaming deal. As I said I don't think anybody is going to want to use one as a desktop but as tiny green servers they are a great value and work fantastically. /jl -- ASCII ribbon campaign ( ) Powered by Lemote Fuloong against HTML e-mail X Loongson MIPS and OpenBSD and proprietary/ \http://www.mutt.org attachments / \ Code Blue or Go Home! Encrypted email preferred PGP Key 2048R/DA65BC04
Re: SPARC minimum hardware specification
On Sun, Jul 19, 2015 at 09:09:56PM +0200, ludovic coues wrote: > 2015-07-19 17:03 GMT+02:00 John Long : > > > > OpenBSD mips64el runs oustandingly well on the Lemote boxes. See here: > > http://www.openbsd.org/loongson.html > > > > I don't think anybody will be happy with a Loongson as a desktop box but > > they do shine tiny servers. > > > > /jl > > Where could one acquire one of the machines mentioned on the link ? > I've seen no option to buy one on their site and I got no luck on ebay. As far as I know the only way left is directly through Lemote. They have an Aliexpress shop but it is usually offline. If you email them they open their shop again. If you can't find the address on their website let me know offline and I'll scrounge up some contact info. /jl -- ASCII ribbon campaign ( ) Powered by Lemote Fuloong against HTML e-mail X Loongson MIPS and OpenBSD and proprietary/ \http://www.mutt.org attachments / \ Code Blue or Go Home! Encrypted email preferred PGP Key 2048R/DA65BC04
Re: SPARC minimum hardware specification
On Sat, Jul 18, 2015 at 5:45 AM, Christian Weisgerber wrote: > On 2015-07-17, BSD wrote: > >> As a new user, I find myself in the same position as the OP: very >> interested in non-Intel products. But there seems to be a vacuum of >> information around this topic. > > You're 15 years too late. x86 has won. I know I'm persona non-grata on the list these days, and I doubt I'm going to make much sense in an argument, but it's the way Intel won that has some of use willing to take a small hit on performance or price. Besides, it's only a small hit on much of what I do. RAM and hard disk speed make up for quite a bit. -- Joel Rees Be careful when you look at conspiracy. Arm yourself with knowledge of yourself, as well: http://reiisi.blogspot.jp/2011/10/conspiracy-theories.html
Re: SPARC minimum hardware specification
2015-07-21 1:57 GMT+02:00 Joel Rees : > On Sat, Jul 18, 2015 at 5:45 AM, Christian Weisgerber > wrote: >> On 2015-07-17, BSD wrote: >> >>> As a new user, I find myself in the same position as the OP: very >>> interested in non-Intel products. But there seems to be a vacuum of >>> information around this topic. >> >> You're 15 years too late. x86 has won. > > I know I'm persona non-grata on the list these days, and I doubt I'm > going to make much sense in an argument, but it's the way Intel won > that has some of use willing to take a small hit on performance or > price. > > Besides, it's only a small hit on much of what I do. RAM and hard disk > speed make up for quite a bit. > I doubt anyone will argue about intel being cheaper and faster. But people might be interested in other thing. Power consumption come in my mind. That's one of the reason why most smartphone run on ARM and wikipedia claim the cell processor was used for making to most efficient server regarding floating operation per second versus power consumption. Another thing is to find bug. Each platform is different and some have already exposed bug which existed on x86 or amd64 but where harder to produce. Here is a metaphor. Intel produce hammer. Really nice hammer, work really well on nails, can also be used to stick a screw in wood. But sometimes, you aren't looking for a hammer. Cordialement, Coues Ludovic +336 148 743 42
Re: SPARC minimum hardware specification
On Mon, Jul 20, 2015 at 6:14 PM, Christian Weisgerber wrote: > Let's cut through some of this crap. If you want a reasonably quiet > sparc64 designed to be put on a desk, your fastest choices are a > > Sun Blade 100 or > Sun Blade 150 (~20% faster) Following this: http://unixhq.com/websgt/sunblade150.pdf -- it's 5.5 bells (is that 55 dB?). M3000, on the other hand is 47 dB(A), reference here: https://www.fujitsu.com/global/Images/ds-sparcenterprise-m3000-ww-en.pdf Does that mean that lying on desk, Sun Blade 150 is more noisy than M3000? I'm asking someone who does have real practical experience with both boxes, since comparison of those noise values is nearly impossible since I'm afraid they both are measured using different metrics. Thanks! Karel
Re: SPARC minimum hardware specification
On Tue, Jul 21, 2015 at 09:14:30AM BST, Karel Gardas wrote: > Following this: http://unixhq.com/websgt/sunblade150.pdf -- it's 5.5 > bells (is that 55 dB?). Yes - it's a standard SI prefix[0]. However, 'bel'(B), *not* 'bell', is not used very often and 'decibel'(dB) is the actual unit. [0] http://www.bipm.org/en/measurement-units/prefixes.html Raf
Re: SPARC minimum hardware specification
On 2015-07-21, Karel Gardas wrote: > Does that mean that lying on desk, Sun Blade 150 is more noisy than > M3000? Coincidently, we yesterday lugged two M3000s into the hackroom here at c2k15. When turned on, these make a hellish noise and you want them in an insulated server room far away. The Blade 150 is a desktop machine. It's not as quiet as a modern quiet PC, but it's par for the course for an average PC from fifteen years ago. (It is also slower than my Soekris net6501-50.) -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: SPARC minimum hardware specification
On 2015-07-21, Graham Stephens wrote: > These machines were not fast when new, but I will say that if you do try > one of these you *need* the proper memory for them (IIRC, registered). You need the proper memory for _any_ machine. And you misremember. spdmem0 at iic0 addr 0x50: 512MB SDRAM ECC PC133CL2 spdmem1 at iic0 addr 0x51: 512MB SDRAM ECC PC133CL2 spdmem2 at iic0 addr 0x52: 256MB SDRAM ECC PC133CL2 > You can run them on cheaper (non registered) memory, but they run *MUCH* > slower than with the supplied memory. That doesn't make any sense. -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: SPARC minimum hardware specification
On 2015-07-20, Joel Rees wrote: > I know I'm persona non-grata on the list these days, and I doubt I'm > going to make much sense in an argument, but it's the way Intel won > that has some of use willing to take a small hit on performance or > price. The irony is that I've probably run more non-x86 hardware than the mouth flappers here. (For one, I've had four different Alphas over the years.) Nowadays there are no alternatives to x86 in the desktop market. None. There are choices in the (big) server market and there are choices in the embedded market, but there is nothing in the desktop segment. Being "willing to take a small hit on performance or price" does not magically will such alternatives into existence; it just makes you sound delusional. And anybody considering OpenBSD on non-x86 be better prepared to pitch in with development, add support and fix problems. If you just want to use it, you're better off with x86. (For instance, and getting vaguely back to topic, the Blade 150 suffers both ohci(4) and gem(4) lockups if you hit the right usage pattern. And we have tons of build logs from ports that fail to build on various archs.) -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: SPARC minimum hardware specification
I have dismantled my Sun Blade 100, circa 2002 era, and I have the (4) 500 MG memory sticks, keyboard, mouse, monitor, CD player, and the two original stock 15 GB IDE drives. If anyone wants the parts, let me know. It was occasionally a noisy box. I kept mine in a metal enclosure for medium sized tower computers, bought via catalog. The enclosure really deadened the sound for around $35. This machine worked well with OpenBSD. Version 5.3 was very good on it. Sent from my BlackBerry 10 smartphone on the Verizon Wireless 4G LTE network. Original Message From: Graham Stephens Sent: Tuesday, July 21, 2015 2:31 PM To: na...@mips.inka.de Cc: misc@openbsd.org Subject: Re: SPARC minimum hardware specification On 21/07/2015 17:10, Christian Weisgerber wrote: > On 2015-07-21, Graham Stephens wrote: > >> These machines were not fast when new, but I will say that if you do try >> one of these you *need* the proper memory for them (IIRC, registered). > > You need the proper memory for _any_ machine. And you misremember. > > spdmem0 at iic0 addr 0x50: 512MB SDRAM ECC PC133CL2 > spdmem1 at iic0 addr 0x51: 512MB SDRAM ECC PC133CL2 > spdmem2 at iic0 addr 0x52: 256MB SDRAM ECC PC133CL2 > >> You can run them on cheaper (non registered) memory, but they run *MUCH* >> slower than with the supplied memory. > > That doesn't make any sense. > Now that you've called me out, I had to do a bit of digging to remember some of the facts... I have the alternative memory in working machines at the moment, so I can't take them apart to check the specs; it may be that the clock timings are different, I don't recall. I was right, however, about the memory being registered - I just had it the wrong way round. The OEM memory is ECC unbuffered. By changing a jumper on the motherboard it allows the use of ECC registered, which is easier to come by and hence cheaper - at least it was when I was looking last. This may explain the difference in speeds.
Re: SPARC minimum hardware specification
On Tue, Jul 21, 2015 at 6:06 PM, Christian Weisgerber wrote: > Coincidently, we yesterday lugged two M3000s into the hackroom here > at c2k15. When turned on, these make a hellish noise and you want > them in an insulated server room far away. Christian, this is really first hand experience I needed to hear. It basically means that it's useless to purchase a lot of expensive M3k instead of more powerful, yet noisy and a lot heavier M4k. Thanks a lot for this! > The Blade 150 is a desktop machine. It's not as quiet as a modern > quiet PC, but it's par for the course for an average PC from fifteen > years ago. (It is also slower than my Soekris net6501-50.) :-) good to know. I guess M3k is several times faster than blade 150. My bet is 5-7x. is that right? Thanks a lot for all your information provided! Karel
Re: SPARC minimum hardware specification
On Tue, Jul 21, 2015 at 04:09:58PM +0100, Raf Czlonka wrote: > On Tue, Jul 21, 2015 at 09:14:30AM BST, Karel Gardas wrote: > > > Following this: http://unixhq.com/websgt/sunblade150.pdf -- it's 5.5 > > bells (is that 55 dB?). > > Yes - it's a standard SI prefix[0]. However, 'bel'(B), *not* 'bell', is > not used very often and 'decibel'(dB) is the actual unit. The wonders of metric logic: a decimeter is one-tenth of a meter, but a decibel is ten times a bel? Erling > [0] http://www.bipm.org/en/measurement-units/prefixes.html > > Raf
Re: SPARC minimum hardware specification
On Tue, Jul 21, 2015 at 11:18 PM, Erling Westenvik wrote: > On Tue, Jul 21, 2015 at 04:09:58PM +0100, Raf Czlonka wrote: >> On Tue, Jul 21, 2015 at 09:14:30AM BST, Karel Gardas wrote: >> >> > Following this: http://unixhq.com/websgt/sunblade150.pdf -- it's 5.5 >> > bells (is that 55 dB?). >> >> Yes - it's a standard SI prefix[0]. However, 'bel'(B), *not* 'bell', is >> not used very often and 'decibel'(dB) is the actual unit. > > The wonders of metric logic: a decimeter is one-tenth of a meter, but a > decibel is ten times a bel? decibel is also one-tenth of bell, isn't it? Anyway, by different metic logic I've been more thinking about all those measurements on bystander position versus common position etc. See http://www.spectra.com/wp-content/uploads/coolthreadst1000.pdf -- and search for "noise". You will see: Declared Operating/Idling Acoustic Acoustic Noise 7.7B (LwAd,1B=10dB) 66dB (LpAm, bystander positions) -- this LwAd and LpAm is what I'm talking about here and what makes comparison of noise from Blade 100/150 and M3000 so difficult. Fortunately Christian give me this information from his first hand experience. Karel
Re: SPARC minimum hardware specification
On Tue, 21 Jul 2015 23:18:40 +0200 Erling Westenvik wrote: > On Tue, Jul 21, 2015 at 04:09:58PM +0100, Raf Czlonka wrote: > > On Tue, Jul 21, 2015 at 09:14:30AM BST, Karel Gardas wrote: > > > > > Following this: http://unixhq.com/websgt/sunblade150.pdf -- it's 5.5 > > > bells (is that 55 dB?). > > > > Yes - it's a standard SI prefix[0]. However, 'bel'(B), *not* 'bell', is > > not used very often and 'decibel'(dB) is the actual unit. > > The wonders of metric logic: a decimeter is one-tenth of a meter, but a > decibel is ten times a bel? Mindgames...? 1 dB == 0.1 Bel 1 dm == 0.1 m 1 Bel == 10 dB 1 m == 10 dm > Erling > > > [0] http://www.bipm.org/en/measurement-units/prefixes.html > > > > Raf > Greetings Ben
Re: SPARC minimum hardware specification
Hello again, As the OP I would say that I got the idea of "SPARC as a desktop". There were nice and insightdfull answers, so I thank you all. I have waited to see Nick Holland's answer, he is a real guru in such questions, but maybe he is busy for now. I can wait, no problem, but I think we can let this thread away since it had reached its purpose. There is no need for contradictions and such things. Let it be hackatons. And pictures from the current one, like Christian Weisgerber said. A little bit more like this one: http://undeadly.org/images/p2k10/serenade-full.jpg Accidentally I found these also: http://www.openbsd-france.org/mirrors/c2k3/index.html Thanks.
Re: SPARC minimum hardware specification
On Wed, Jul 22, 2015 at 2:08 AM, Christian Weisgerber wrote: > On 2015-07-20, Joel Rees wrote: > >[...] > Being "willing to take a small hit on performance or > price" does not magically will such alternatives into existence; > it just makes you sound delusional. Well, yeah, I am a dreamer, but I don't think I'm all that delusional to think, for instance, that I could put some ARM hardware in a lunchbox, attach a keyboard, display, an ethernet cable, a sata disk, and a fan, boot up an OS, and run the typical office applications, image editors, music players and editors, even Gramps. No Skype, I'm sure, but that's not a loss to me. Getting OpenBSD running on it would take a bit of planning in choosing parts, to bring the amount of code that would have to be written down to a reasonable level. Getting the X11 server code running is a blocker, isn't it? But not a permanent one. Just because that is presently a dream for me does not mean that I should assume that no one interested in OpenBSD has the time and spare money to do it. Maybe even I could do that much in another six months or so. (Ehrm, no. Not six months. Longer than that. Did I admit that I'm a dreamer?) > And anybody considering OpenBSD on non-x86 be better prepared to > pitch in with development, add support and fix problems. If you > just want to use it, you're better off with x86. Agreed. Although even x86 is no guarantee. I'm doing without wifi on this HP pavilion because the controller is too recent for the present driver in openbsd. And, yes, I'm trying to find time to work through the code until I have enough confidence to start dragging code in from kernel.org. And I should be re-reading the code now instead of arguing with you. :-/ > (For instance, and getting vaguely back to topic, the Blade 150 > suffers both ohci(4) and gem(4) lockups if you hit the right usage > pattern. And we have tons of build logs from ports that fail to > build on various archs.) I guess what you're saying is that the OP should expect a really rough road ahead, if looking to do desktop kinds of things on sparc hardware. I don't think I'd want to argue with you there. You'd be the one to know. Still, saying x86 has won it all is going a little farther than I wanted to sit quietly by and ignore. Now I've said my piece and I'll shut up. -- Joel Rees Be careful when you look at conspiracy. Arm yourself with knowledge of yourself, as well: http://reiisi.blogspot.jp/2011/10/conspiracy-theories.html
Re: SPARC minimum hardware specification
We're hurtling towards the 5.8 release and, as usual, ports and packages on non-x86 platforms are in dire shape. If you want to put your money where your mouth is, take a look at recent build logs and start fixing some of those problems. http://build-failures.rhaalovely.net/ sparc64, powerpc, alpha, hppa, ... Yes, this requires skill and effort. -- Christian "naddy" Weisgerber na...@mips.inka.de
Re: SPARC minimum hardware specification
On 2015-07-27, Christian Weisgerber wrote:
> We're hurtling towards the 5.8 release and, as usual, ports and
> packages on non-x86 platforms are in dire shape.
>
> If you want to put your money where your mouth is, take a look at recent
> build logs and start fixing some of those problems.
> http://build-failures.rhaalovely.net/
> sparc64, powerpc, alpha, hppa, ...
>
> Yes, this requires skill and effort.
Some of them probably don't require that much skill, just access to the
machine arch and a little effort. Also look for BROKEN- in Makefiles.
Unfortunately a lot of the errors in these build logs would just go away
if the build was reattempted ("Error: job failed 256") so it's hard to
spot which ports have real problems from the directory indices.
[OT] Metric/SI prefixes (Was: Re: SPARC minimum hardware specification)
On Tue, Jul 21, 2015 at 10:46:51PM BST, Karel Gardas wrote: > On Tue, Jul 21, 2015 at 11:18 PM, Erling Westenvik > wrote: > > On Tue, Jul 21, 2015 at 04:09:58PM +0100, Raf Czlonka wrote: > >> On Tue, Jul 21, 2015 at 09:14:30AM BST, Karel Gardas wrote: > >> > >> > Following this: http://unixhq.com/websgt/sunblade150.pdf -- it's 5.5 > >> > bells (is that 55 dB?). > >> > >> Yes - it's a standard SI prefix[0]. However, 'bel'(B), *not* 'bell', is > >> not used very often and 'decibel'(dB) is the actual unit. > > > > The wonders of metric logic: a decimeter is one-tenth of a meter, but a > > decibel is ten times a bel? > > decibel is also one-tenth of bell, isn't it? Yes, as Benjamin has already explained and as the name suggests (sic!), it is one-tenth of a *bel* ... not *bell* :^) Raf
