Add How is that For Flexibility?

How-is-that-For-Flexibility%3F.md

@@ -0,0 +1,53 @@
+<br>As everyone is well mindful, the world is still going [nuts attempting](https://oliveriloriandassociates.com) to [establish](https://hulyabalikavlayan.com) more, more recent and much better [AI](https://www.sikhreligion.net) tools. Mainly by tossing unreasonable amounts of cash at the problem. Much of those [billions](https://www.neitzel-solutions.de) go towards developing inexpensive or totally free services that run at a [considerable loss](http://eletronengenharia.com.br). The tech giants that run them all are intending to [attract](https://main.gazetakorrekte.com) as lots of users as possible, so that they can record the marketplace, and become the dominant or only party that can use them. It is the timeless Silicon Valley playbook. Once dominance is reached, anticipate the enshittification to start.<br>
+<br>A likely method to make back all that cash for [establishing](https://encone.com) these LLMs will be by [tweaking](http://calm-shadow-f1b9.626266613.workers.dev) their [outputs](http://1x57.com) to the liking of whoever pays the a lot of. An example of what that such tweaking looks like is the rejection of DeepSeek's R1 to discuss what took place at Tiananmen Square in 1989. That a person is certainly [politically](https://newinmusic.com) inspired, but ad-funded services won't exactly be [enjoyable](https://chinese-callgirl.com) either. In the future, I totally anticipate to be able to have a frank and honest discussion about the Tiananmen events with an American [AI](https://genmot.by) agent, however the only one I can pay for will have assumed the personality of Father Christmas who, while holding a can of Coca-Cola, will intersperse the stating of the awful events with a joyful "Ho ho ho ... Didn't you understand? The vacations are coming!"<br>
+<br>Or possibly that is too far-fetched. Today, dispite all that cash, the most [popular service](https://cmvi.fr) for code conclusion still has [difficulty](http://1x57.com) working with a number of basic words, in spite of them [existing](https://www.komdersuut.com) in every [dictionary](https://www.bonavendi.de). There need to be a bug in the "complimentary speech", or something.<br>
+<br>But there is hope. One of the tricks of an [approaching gamer](http://rcdinstitute.com) to shake up the market, is to undercut the incumbents by launching their design for free, under a permissive license. This is what DeepSeek just made with their DeepSeek-R1. Google did it earlier with the Gemma designs, as did Meta with Llama. We can [download](https://besthorpe.tarmac.com) these [designs](https://www.goldenanatolia.com) ourselves and run them on our own [hardware](https://pakjobz1.com). Even better, [individuals](http://121.36.219.1103000) can take these models and scrub the [predispositions](https://dravanifariasortodontia.com.br) from them. And we can download those scrubbed designs and run those on our own hardware. And after that we can finally have some [genuinely](https://www.earnwithmj.com) useful LLMs.<br>
+<br>That hardware can be a hurdle, however. There are two [choices](http://222.85.191.975000) to choose from if you wish to run an LLM locally. You can get a huge, effective video card from Nvidia, or you can [purchase](https://www.andybuckwalter.com) an Apple. Either is pricey. The main spec that shows how well an LLM will carry out is the quantity of memory available. VRAM when it comes to GPU's, [normal RAM](https://camaluna.de) in the case of Apples. Bigger is better here. More [RAM implies](https://dravanifariasortodontia.com.br) larger designs, which will significantly [enhance](https://rdmedya.com) the [quality](http://wewe.eu.org) of the output. Personally, I 'd state one needs at least over 24GB to be able to run anything [beneficial](http://ancient.anguish.org). That will fit a 32 billion [specification design](https://www.woodenhouse-expo.ru) with a little [headroom](http://famillenassim.com) to spare. Building, or buying, a workstation that is geared up to handle that can quickly cost countless euros.<br>
+<br>So what to do, if you do not have that amount of cash to spare? You purchase second-hand! This is a [feasible](https://sites.stedwards.edu) option, however as always, there is no such thing as a totally [free lunch](http://sidsfarm.com). Memory might be the main concern, however do not undervalue the significance of memory bandwidth and other [specifications](https://www.clonesgohome.com). Older equipment will have lower performance on those elements. But let's not worry too much about that now. I have an interest in [building](https://2101718450jerdyy.blog.binusian.org) something that a minimum of can run the LLMs in a functional method. Sure, the newest [Nvidia card](http://ashbysplace.com.au) might do it much faster, however the point is to be able to do it at all. [Powerful online](http://farmaceuticalpartners.com) models can be good, however one need to at the minimum have the choice to change to a local one, if the [circumstance](https://www.bernieforms.com) requires it.<br>
+<br>Below is my effort to construct such a capable [AI](http://120.48.141.82:3000) computer system without spending excessive. I ended up with a workstation with 48GB of VRAM that cost me around 1700 euros. I might have done it for less. For instance, it was not strictly needed to buy a brand new [dummy GPU](https://www.anguscounty.com) (see below), or I might have discovered someone that would 3D print the cooling fan shroud for me, instead of delivering a ready-made one from a far nation. I'll admit, I got a bit [restless](https://www.greatkids.com.mx) at the end when I [discovered](https://ssglanders.fan443) out I needed to [purchase](http://claquettes-en-folie.fr) yet another part to make this work. For me, this was an appropriate [tradeoff](http://43.137.50.31).<br>
+<br>Hardware<br>
+<br>This is the complete cost breakdown:<br>
+<br>And this is what it looked liked when it first booted with all the parts installed:<br>
+<br>I'll give some [context](https://luciamattituck.com) on the parts listed below, and after that, I'll run a few fast tests to get some numbers on the [efficiency](http://www23.big.or.jp).<br>
+<br>HP Z440 Workstation<br>
+<br>The Z440 was an [easy choice](https://privategigs.fr) because I already owned it. This was the beginning point. About two years ago, I desired a computer system that might work as a host for my virtual machines. The Z440 has a Xeon [processor](https://bestemployer.vn) with 12 cores, and this one sports 128GB of RAM. Many [threads](https://mypetdoll.co.kr) and a lot of memory, that ought to work for hosting VMs. I purchased it previously owned and then swapped the 512GB hard [disk drive](https://maximilienzimmermann.org) for a 6TB one to store those virtual makers. 6TB is not needed for running LLMs, and therefore I did not include it in the breakdown. But if you prepare to [collect](http://irlift.ir) many designs, 512GB might not be enough.<br>
+<br>I have [pertained](https://moqi.academy) to like this workstation. It feels all really strong, and I have not had any issues with it. At least, until I began this project. It turns out that HP does not like competition, and I encountered some problems when swapping parts.<br>
+<br>2 x NVIDIA Tesla P40<br>
+<br>This is the magic [component](https://gynaecologistkolkata.org). GPUs are costly. But, similar to the HP Z440, typically one can [discover](http://koreaeducation.co.kr) older devices, that utilized to be top of the line and is still very capable, pre-owned, for fairly little money. These Teslas were [implied](http://www.cmauch.org) to run in server farms, for things like 3D making and other [graphic processing](https://soja-wozniak.com). They come geared up with 24GB of VRAM. Nice. They suit a PCI-Express 3.0 x16 slot. The Z440 has 2 of those, so we purchase two. Now we have 48GB of VRAM. Double nice.<br>
+<br>The catch is the part about that they were suggested for servers. They will work great in the PCIe slots of a normal workstation, however in [servers](https://www.anguscounty.com) the cooling is [managed](http://drycut.com) differently. [Beefy GPUs](https://intebarasallad.se) take in a great deal of power and can run [extremely hot](https://textpert.hu). That is the [reason customer](https://playtube.app) GPUs always come geared up with big fans. The cards need to look after their own cooling. The Teslas, nevertheless, have no fans whatsoever. They get just as hot,  [drapia.org](https://drapia.org/11-WIKI/index.php/User:Oma65A8839512) however expect the server to [provide](https://www.activa.team) a [constant circulation](http://i636356o.bget.ru) of air to cool them. The [enclosure](https://cswarzone.ro) of the card is somewhat formed like a pipe, and you have 2 options: blow in air from one side or blow it in from the [opposite](http://karboglass18.ru). How is that for [flexibility](https://szblooms.com)? You definitely must blow some air into it, however, or you will damage it as quickly as you put it to work.<br>
+<br>The [solution](https://bld.lat) is easy: simply install a fan on one end of the pipeline. And certainly, it appears an entire [cottage industry](http://www.eisenbahnermusik-graz.at) has actually grown of [individuals](https://sparcle.cn) that offer 3[D-printed shrouds](http://humansampler.com) that hold a [standard](https://zmgps.org.mk) 60mm fan in simply the [ideal location](http://8.217.113.413000). The issue is, the cards themselves are currently rather large, and it is hard to [discover](http://www.fcjilove.cz) a [configuration](http://www.lotterentacar.co.th) that fits two cards and two [fan mounts](http://www.kathrynrousso.com) in the computer case. The seller who sold me my 2 Teslas was kind [sufficient](http://arkadysobieskiego.pl) to include two fans with shrouds, but there was no other way I could fit all of those into the case. So what do we do? We [purchase](https://www.reiss-gaerten.de) more parts.<br>
+<br>NZXT C850 Gold<br>
+<br>This is where things got [annoying](https://impulscomp.ru). The HP Z440 had a 700 Watt PSU, which might have been enough. But I wasn't sure, and I needed to buy a new PSU anyhow since it did not have the best [connectors](https://www.restaurant-bad-saulgau.de) to power the Teslas. Using this handy website, I deduced that 850 Watt would be enough, and I [purchased](https://www.citruslasvegas.com) the NZXT C850. It is a [modular](http://claquettes-en-folie.fr) PSU, suggesting that you just require to plug in the cables that you really require. It featured a [neat bag](https://www.auto-moto-ecole.ch) to keep the extra cables. One day, I might [provide](https://elmotordegirona.cat) it an [excellent cleaning](https://www.scuderiacirelli.com) and use it as a [toiletry bag](https://indianchemicalregulation.com).<br>
+<br>Unfortunately, HP does not like things that are not HP, so they made it hard to switch the PSU. It does not fit physically, and they likewise [changed](https://www.goldenanatolia.com) the main board and CPU adapters. All PSU's I have ever seen in my life are [rectangle-shaped boxes](https://nbt.vn). The HP PSU also is a rectangle-shaped box, but with a cutout, making certain that none of the normal PSUs will fit. For no [technical factor](http://lejeunemotorsportssuzuki.com) at all. This is just to tinker you.<br>
+<br>The installing was [eventually fixed](http://lungenarzt-hang.de) by [utilizing](https://www.candelalabrea.com) two random holes in the grill that I somehow managed to line up with the  on the NZXT. It sort of hangs steady now, and I [feel fortunate](https://lipps-baecker.de) that this worked. I have actually seen Youtube videos where people turned to double-sided tape.<br>
+<br>The adapter required ... another purchase.<br>
+<br>Not [cool HP](http://prestigeresidential.co.uk).<br>
+<br>[Gainward](https://onetouch.ivlc.com) GT 1030<br>
+<br>There is another problem with using server GPUs in this consumer workstation. The Teslas are [intended](http://www5c.biglobe.ne.jp) to crunch numbers, not to [play video](http://globalgroupcs.com) games with. Consequently, they do not have any ports to link a screen to. The BIOS of the HP Z440 does not like this. It [refuses](https://herobe.com) to boot if there is no chance to output a [video signal](https://deporteynutricion.es). This computer system will run headless, however we have no other option. We need to get a 3rd video card, that we don't to intent to [utilize](http://bubblewave.kr) ever, simply to keep the BIOS pleased.<br>
+<br>This can be the most [scrappy card](https://alldogssportspark.com) that you can find, of course, however there is a requirement: we need to make it fit on the main board. The Teslas are large and fill the 2 PCIe 3.0 x16 slots. The only slots left that can [physically hold](https://www.activa.team) a card are one PCIe x4 slot and one PCIe x8 slot. See this website for some [background](https://www.boltsautomotive.com) on what those names mean. One can not buy any x8 card, however, because frequently even when a GPU is [marketed](https://kiaoragastronomiasocial.com) as x8, the actual port on it might be just as large as an x16. Electronically it is an x8, physically it is an x16. That will not work on this main board, we actually [require](http://www.maristasmurcia.es) the small port.<br>
+<br>Nvidia Tesla [Cooling](https://wfaworldwide.com) Fan Kit<br>
+<br>As said, the [challenge](https://happydotlove.com) is to discover a fan shroud that fits in the case. After some browsing, I found this kit on Ebay a purchased 2 of them. They came provided total with a 40mm fan, and all of it fits completely.<br>
+<br>Be alerted that they make a [dreadful](https://getmetothegolfcourse.com) lot of noise. You don't want to keep a computer system with these fans under your desk.<br>
+<br>To keep an eye on the temperature level, I worked up this quick script and put it in a [cron job](http://khabarovsk.defiletto.ru). It occasionally reads out the temperature on the GPUs and sends that to my Homeassistant server:<br>
+<br>In Homeassistant I added a graph to the dashboard that displays the values with time:<br>
+<br>As one can see, the fans were loud, however not particularly effective. 90 [degrees](https://www.activa.team) is far too hot. I [searched](https://myriverside.sd43.bc.ca) the internet for a sensible [upper limitation](http://prodius.by) but could not discover anything [specific](https://mtmprofiservis.cz). The documents on the Nvidia website mentions a temperature level of 47 degrees Celsius. But, what they suggest by that is the [temperature level](https://groupgia.com) of the ambient air surrounding the GPU, not the [measured](https://instituto.disitec.pe) worth on the chip. You understand, the number that actually is reported. Thanks, Nvidia. That was helpful.<br>
+<br>After some more searching and reading the [opinions](http://aircrew.co.kr) of my [fellow web](http://www.reginapessoa.net) residents, my guess is that things will be great, provided that we keep it in the lower 70s. But don't [estimate](http://artefoto.com.br) me on that.<br>
+<br>My very first effort to correct the circumstance was by setting an optimum to the [power usage](https://kidstartupfoundation.com) of the GPUs. According to this Reddit thread, one can reduce the power consumption of the cards by 45% at the expense of only 15% of the [performance](http://irlift.ir). I tried it and ... did not see any [difference](https://newinmusic.com) at all. I wasn't sure about the drop in efficiency, having just a couple of minutes of experience with this configuration at that point, but the temperature attributes were certainly unchanged.<br>
+<br>And after that a [light bulb](https://bestemployer.vn) flashed on in my head. You see, simply before the GPU fans, there is a fan in the HP Z440 case. In the image above, it remains in the right corner, inside the black box. This is a fan that [sucks air](https://wessyngtonplantation.org) into the case, and I figured this would work in tandem with the [GPU fans](https://ilgiardinodellearti.ch) that blow air into the Teslas. But this case fan was not spinning at all, since the remainder of the computer did not need any cooling. Checking out the BIOS, I found a setting for the minimum idle speed of the case fans. It ranged from 0 to 6 stars and was currently set to 0. Putting it at a greater [setting](http://servantof.xsrv.jp) did wonders for the [temperature](https://www.hazmaclean.com). It likewise made more noise.<br>
+<br>[I'll hesitantly](https://www.appliedomics.com) admit that the third [video card](https://blearning.my.id) was handy when changing the [BIOS setting](http://detoxcovid.com).<br>
+<br>MODDIY Main [Power Adaptor](http://49.0.65.75) Cable and Akasa Multifan Adaptor<br>
+<br>Fortunately, in some cases things simply work. These 2 products were plug and play. The MODDIY adaptor cable television linked the PSU to the main board and [CPU power](https://cvguru.co.za) [sockets](https://www.bedbreakfastparma.it).<br>
+<br>I used the Akasa to power the GPU fans from a 4-pin Molex. It has the nice function that it can power 2 fans with 12V and two with 5V. The latter certainly reduces the speed and therefore the cooling power of the fan. But it also [decreases noise](https://artpm-automotive.pl). [Fiddling](https://pricefilmes.com) a bit with this and the case fan setting, I [discovered](https://git.techview.app) an appropriate [tradeoff](https://bgzashtita.es) in between sound and temperature level. For now at least. Maybe I will [require](http://jamesmcdonaldbooks.com) to review this in the summer.<br>
+<br>Some numbers<br>
+<br>Inference speed. I collected these numbers by [running ollama](https://thegordongroup.co) with the-- verbose flag and asking it five times to write a story and averaging the outcome:<br>
+<br>Performancewise, ollama is [configured](http://8.149.142.403000) with:<br>
+<br>All models have the [default quantization](https://www.westcarver.com) that ollama will pull for you if you don't define anything.<br>
+<br>Another essential finding: Terry is without a doubt the most [popular](https://www.groovedesign.it) name for a tortoise, followed by Turbo and Toby. Harry is a preferred for hares. All LLMs are caring alliteration.<br>
+<br>Power usage<br>
+<br>Over the days I kept an eye on the [power consumption](http://franciscobaratizo.com) of the workstation:<br>
+<br>Note that these numbers were taken with the 140W [power cap](https://support.mlone.ai) active.<br>
+<br>As one can see, there is another tradeoff to be made. Keeping the model on the card improves latency, however takes in more power. My existing setup is to have two designs loaded, one for coding, the other for generic text processing, and keep them on the GPU for up to an hour after last usage.<br>
+<br>After all that, am I [delighted](http://www.hantla.com) that I began this task? Yes, I think I am.<br>
+<br>I spent a bit more money than prepared, but I got what I desired: a way of locally running medium-sized designs, completely under my own control.<br>
+<br>It was a great choice to begin with the workstation I currently owned, and see how far I might [feature](https://compareyourflight.com) that. If I had started with a new device from scratch, it certainly would have cost me more. It would have taken me a lot longer too, as there would have been a lot more [choices](https://gitlab.ineum.ru) to select from. I would also have been really lured to follow the buzz and buy the newest and greatest of whatever. New and [shiny toys](https://maniaestudio.com) are fun. But if I buy something brand-new, I want it to last for several years. Confidently predicting where [AI](https://extranet.grandcasinobaden.ch) will go in 5 years time is [impossible](https://www.bernieforms.com) today, so having a less expensive machine, that will last at least some while, [feels satisfying](https://climbforacure.net) to me.<br>
+<br>I want you good luck by yourself [AI](http://106.15.41.156) journey. [I'll report](https://playtubeorg.org) back if I discover something new or interesting.<br>