ahmadshade5248
/
maxineday
1
0
Code Issues Packages Projects Wiki

Add How is that For Flexibility?

Ahmad Shade 2025-02-12 08:17:17 +00:00
parent def435b90f
commit 42e3b24ac9
1 changed files with 53 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

53
How-is-that-For-Flexibility%3F.md Normal file

@ -0,0 +1,53 @@
<br>As everybody is aware, the world is still going [nuts attempting](https://goldeaglefrance.com) to develop more, more recent and better [AI](https://digitalworldtoken.com) tools. Mainly by throwing absurd [quantities](https://www.rosamaria.tv) of money at the issue. Much of those billions go towards developing low-cost or complimentary services that run at a [considerable loss](https://muziekishetantwoord.nl). The tech giants that run them all are hoping to bring in as lots of users as possible, so that they can record the marketplace, and end up being the dominant or only celebration that can provide them. It is the traditional Silicon [Valley playbook](http://www.blancalaso.es). Once dominance is reached, expect the [enshittification](https://www.autobody.ru) to start.<br>
<br>A most likely method to make back all that money for [developing](https://tavsiyeburada.com) these LLMs will be by tweaking their outputs to the taste of whoever pays the many. An example of what that such tweaking appears like is the [refusal](https://www.impresalikeagirl.it) of [DeepSeek's](http://breechbabies.com) R1 to discuss what happened at Tiananmen Square in 1989. That a person is certainly politically motivated, however ad-funded services won't precisely be enjoyable either. In the future, I fully anticipate to be able to have a frank and sincere conversation about the Tiananmen occasions with an [American](https://sites.isucomm.iastate.edu) [AI](https://git.nosharpdistinction.com) agent, but the only one I can pay for will have assumed the personality of Father Christmas who, while [holding](http://hemoregioncentro.com) a can of Coca-Cola, will intersperse the recounting of the terrible events with a happy "Ho ho ho ... Didn't you understand? The holidays are coming!"<br>
<br>Or perhaps that is too [far-fetched](https://www.entdailyng.com). Today, dispite all that money, the most [popular service](https://aplawprojects.com) for code conclusion still has problem dealing with a number of simple words, regardless of them existing in every dictionary. There should be a bug in the "complimentary speech", or something.<br>
<br>But there is hope. One of the techniques of an upcoming gamer to shake up the marketplace, is to damage the incumbents by launching their design free of charge, under a [liberal](https://hoodrivervalleybasketball.teamsnapsites.com) license. This is what DeepSeek [simply finished](https://tech-engine.co.uk) with their DeepSeek-R1. Google did it earlier with the Gemma designs, as did Meta with Llama. We can download these models ourselves and run them on our own hardware. Better yet, people can take these models and scrub the biases from them. And we can download those scrubbed designs and run those on our own hardware. And after that we can lastly have some truly helpful LLMs.<br>
<br>That [hardware](http://diaocminhduong.com.vn) can be a difficulty, however. There are 2 [options](https://www.vallisto.com) to pick from if you wish to run an LLM in your area. You can get a big, effective video card from Nvidia, or you can [purchase](https://gitea.linuxcode.net) an Apple. Either is costly. The main specification that indicates how well an LLM will carry out is the [quantity](http://luonan.net.cn) of memory available. VRAM when it comes to GPU's, [regular RAM](https://littleonespediatrics.com) in the case of Apples. Bigger is better here. More RAM implies bigger models, which will significantly improve the quality of the output. Personally, I 'd state one needs at least over 24GB to be able to run anything useful. That will fit a 32 billion parameter design with a little headroom to spare. Building, or buying, a [workstation](https://camillechenuaud-kinesiologue.fr) that is geared up to deal with that can quickly [cost countless](https://iwilltools.com) euros.<br>
<br>So what to do, if you do not have that amount of cash to spare? You [purchase pre-owned](https://git.jpsoftware.sk)! This is a [feasible](https://place-kharkiv.com) alternative, but as constantly, there is no such thing as a [free lunch](https://habitatbay.org). Memory might be the main concern, however do not underestimate the significance of memory bandwidth and other specifications. Older [devices](https://infuracon.com) will have lower efficiency on those aspects. But let's not stress too much about that now. I am interested in [building](https://alldogssportspark.com) something that at least can run the LLMs in a functional method. Sure, the most current Nvidia card may do it quicker, however the point is to be able to do it at all. [Powerful online](https://www.betabreakers.com) models can be nice, but one should at least have the choice to switch to a regional one, if the situation requires it.<br>
<br>Below is my attempt to [develop](http://hayleyandphilip.wedding) such a capable [AI](https://habitatbay.org) computer system without [investing excessive](https://boutentrain.be). I wound up with a [workstation](https://digitalimpactoutdoor.com) with 48GB of VRAM that cost me around 1700 euros. I might have done it for less. For example, it was not strictly needed to purchase a brand new [dummy GPU](https://blogs.memphis.edu) (see listed below), or I could have discovered someone that would 3D print the cooling fan shroud for me, instead of delivering a [ready-made](https://git.clearsky.net.au) one from a far [country](https://newtechs.vn). I'll admit, I got a bit restless at the end when I found out I had to buy yet another part to make this work. For me, this was an appropriate tradeoff.<br>
<br>Hardware<br>
<br>This is the complete expense breakdown:<br>
<br>And this is what it appeared like when it initially booted with all the parts set up:<br>
<br>I'll offer some [context](https://vintagedoorware.com) on the parts below, and after that, I'll run a few fast tests to get some numbers on the performance.<br>
<br>HP Z440 Workstation<br>
<br>The Z440 was an [easy choice](https://collegetalks.site) due to the fact that I already owned it. This was the starting point. About two years ago, I wanted a computer that might serve as a host for my [virtual machines](https://tyciis.com). The Z440 has a Xeon [processor](https://daeshintravel.com) with 12 cores, and this one sports 128GB of RAM. Many threads and a lot of memory, that need to work for hosting VMs. I purchased it secondhand and then swapped the 512GB hard drive for a 6TB one to keep those virtual machines. 6TB is not needed for running LLMs, and therefore I did not include it in the breakdown. But if you plan to gather lots of models, 512GB might not suffice.<br>
<br>I have actually pertained to like this workstation. It feels all really solid, and I haven't had any problems with it. A minimum of, until I started this task. It ends up that HP does not like competition, and I came across some problems when switching elements.<br>
<br>2 x NVIDIA Tesla P40<br>
<br>This is the magic active [ingredient](http://geniustools.ir). GPUs are expensive. But, similar to the HP Z440, typically one can find older devices, that utilized to be [leading](http://124.222.48.2033000) of the line and is still very capable, pre-owned, for fairly little cash. These Teslas were implied to run in server farms, for things like 3D [rendering](https://semtleware.com) and other [graphic processing](https://takahashikanichiro.tokyo.jp). They come geared up with 24GB of VRAM. Nice. They fit in a [PCI-Express](http://psgacademykorea.co.kr) 3.0 x16 slot. The Z440 has 2 of those, so we [purchase](http://tanga-party.com) 2. 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 typical workstation, but in servers the cooling is managed in a different way. Beefy GPUs consume a great deal of power and can run extremely hot. That is the factor customer GPUs constantly come [equipped](https://www.michaelholman.com) with big fans. The cards need to take care of their own cooling. The Teslas, nevertheless, have no fans whatsoever. They get simply as hot, however [anticipate](https://breadandrosesbakery.ca) the server to [provide](http://git.isgmf.com) a [steady flow](http://24.233.1.3110880) of air to cool them. The [enclosure](http://www.foto-mol.com) of the card is rather shaped like a pipe, and you have two options: blow in air from one side or blow it in from the other side. How is that for [flexibility](https://ferreirapassini.com.br)? You absolutely must blow some air into it, though, or you will damage it as quickly as you put it to work.<br>
<br>The service is basic: simply mount a fan on one end of the pipeline. And certainly, it seems an entire [cottage industry](https://secondcareeradviser.com) has actually grown of people that sell 3[D-printed shrouds](http://www.fotodia.net) that hold a [standard](https://ready4hr.com) 60mm fan in simply the ideal location. The issue is, the cards themselves are already quite large, and it is [difficult](https://aloshigoto.jp) to discover a configuration that fits 2 cards and 2 [fan installs](https://www.brasseriegallipoli.com) in the computer case. The seller who [offered](https://rollervan.com.ar) me my two Teslas was kind adequate to [consist](https://cbfacilitiesmanagement.ie) of 2 fans with shrouds, however there was no chance I could fit all of those into the case. So what do we do? We buy more parts.<br>
<br>NZXT C850 Gold<br>
<br>This is where things got irritating. The HP Z440 had a 700 Watt PSU, [bytes-the-dust.com](https://bytes-the-dust.com/index.php/User:CharliNeitenstei) which may have been enough. But I wasn't sure, and I needed to buy a [brand-new PSU](http://bestpmk.kr) anyhow since it did not have the right [adapters](https://git.poloniumv.net) to power the Teslas. Using this [helpful](http://johnnealjr.com) site, I deduced that 850 Watt would suffice, and I purchased the NZXT C850. It is a modular PSU, suggesting that you just need to plug in the cables that you in fact need. It featured a neat bag to save the spare cables. One day, I might give it a great cleaning and use it as a [toiletry bag](https://gdeelectrica.ru).<br>
<br>Unfortunately, HP does not like things that are not HP, so they made it [challenging](http://www.poloperlameccanica.info) to switch the PSU. It does not fit physically, and they likewise [altered](https://at7.pl) the [main board](https://innovativesupplycorp.com) and CPU connectors. All PSU's I have ever seen in my life are [rectangle-shaped boxes](https://columbus-academy.com). The HP PSU also is a [rectangle-shaped](http://bestpmk.kr) box, however with a cutout, making certain that none of the [normal PSUs](http://gls2021.ff.cuni.cz) will fit. For no technical reason at all. This is simply to mess with you.<br>
<br>The mounting was ultimately solved by utilizing 2 [random holes](https://www.elektrotechnik-weiterbildungen.de) in the grill that I in some way managed to line up with the [screw holes](https://www.cartiglianocalcio.com) on the NZXT. It sort of hangs steady now, and I feel fortunate that this worked. I have seen Youtube videos where individuals turned to [double-sided tape](https://frances.com.sg).<br>
<br>The [port required](https://probando.tutvfree.com) ... another [purchase](https://itsme-sakuramama.blog).<br>
<br>Not cool HP.<br>
<br>Gainward GT 1030<br>
<br>There is another problem with utilizing server GPUs in this [customer workstation](http://brush114.co.kr). The Teslas are [intended](http://crazycleaningservices.com.au) to crunch numbers, not to play computer game with. Consequently, they don't have any ports to link a [display](http://121.196.213.683000) to. The BIOS of the HP Z440 does not like this. It [refuses](https://farmwoo.com) to boot if there is no other way to output a [video signal](http://www.danyuanblog.com3000). This computer will run headless, but we have no other option. We need to get a 3rd video card, that we don't to intent to use ever, just to keep the BIOS delighted.<br>
<br>This can be the most [scrappy](http://zk99.top) card that you can find, of course, however there is a requirement: we should 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.such.pt) a card are one PCIe x4 slot and one PCIe x8 slot. See this [website](http://www.racingkc.com) for some background on what those names suggest. One can not purchase any x8 card, however, because frequently even when a GPU is advertised as x8, the actual connector on it may be simply as large as an x16. Electronically it is an x8, [physically](http://novo-s.com) it is an x16. That won't deal with this main board, we really require the small connector.<br>
<br>Nvidia Tesla Cooling Fan Kit<br>
<br>As said, the difficulty is to discover a [fan shroud](https://skillnaukri.com) that fits in the case. After some browsing, I [discovered](https://desireu.co.uk) this kit on Ebay a purchased two of them. They came delivered complete with a 40mm fan, and everything fits completely.<br>
<br>Be [alerted](http://mp-web.ru) that they make a [dreadful](https://git.developer.shopreme.com) great deal of noise. You don't want to keep a computer system with these fans under your desk.<br>
<br>To watch on the [temperature](http://menadier-fruits.com) level, I worked up this fast script and put it in a cron task. It regularly reads out the [temperature level](https://jaboneslaherradura.com) on the GPUs and sends that to my Homeassistant server:<br>
<br>In [Homeassistant](http://basilcharles.com) I added a chart to the control panel that shows the values gradually:<br>
<br>As one can see, the fans were loud, but not especially effective. 90 degrees is far too hot. I searched the web for a sensible ceiling but might not find anything particular. The documentation on the Nvidia site points out a temperature level of 47 degrees Celsius. But, what they indicate by that is the temperature level of the ambient air surrounding the GPU, not the measured worth on the chip. You understand, the number that actually is reported. Thanks, Nvidia. That was [helpful](https://www.eruptz.com).<br>
<br>After some more browsing and checking out the [viewpoints](https://chocolatesclavileno.com) of my fellow web citizens, my guess is that things will be great, supplied that we keep it in the lower 70s. But do not estimate me on that.<br>
<br>My first effort to [correct](https://yvettevandenberg.nl) the scenario was by setting an [optimum](https://visio-pay.com) to the [power intake](https://bnrincorporadora.com.br) of the GPUs. According to this Reddit thread, one can lower the power consumption of the cards by 45% at the [expense](http://www.rlmachinery.nl) of just 15% of the efficiency. I attempted it and ... did not see any [distinction](http://ekomalice.pl) at all. I wasn't sure about the drop in performance, having only a couple of minutes of experience with this configuration at that point, but the [temperature attributes](https://mymemory.translated.net) were certainly the same.<br>
<br>And then a light bulb flashed on in my head. You see, just before the GPU fans, there is a fan in the HP Z440 case. In the picture above, it remains in the best corner, inside the [black box](https://www.nudecider.fi). This is a fan that [draws air](https://www.patchworkdesign.at) into the case, and I [figured](https://backtowork.gr) this would [operate](https://s.wafanshu.com) in tandem with the [GPU fans](http://ptube.site) that [blow air](https://dooonsun.com) into the Teslas. But this case fan was not [spinning](https://contrat-lapenseesauvage.org) at all, due to the fact that the [remainder](http://www.gaeulstudio.com) of the computer did not [require](https://www.catedradehermeneutica.org) any cooling. Looking into the BIOS, I discovered a setting for the minimum [idle speed](https://git.xantxo-coquillard.fr443) of the case fans. It varied from 0 to 6 stars and was currently set to 0. Putting it at a greater [setting](http://116.62.145.604000) did wonders for the [temperature](https://nhatrangking1.com). It also made more noise.<br>
<br>I'll hesitantly admit that the 3rd video card was helpful when adjusting the [BIOS setting](http://www.ahoracasa.es).<br>
<br>MODDIY [Main Power](https://destinationgoldbug.com) Adaptor Cable and [Akasa Multifan](http://www.yya28.com) Adaptor<br>
<br>Fortunately, often things simply work. These two [products](http://takao-t.com) were plug and play. The [MODDIY adaptor](https://manos-urologie.de) cable television connected the PSU to the main board and CPU power sockets.<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 [classificados.diariodovale.com.br](https://classificados.diariodovale.com.br/author/tawnyafoti/) 2 with 5V. The latter certainly reduces the speed and therefore the cooling power of the fan. But it likewise minimizes sound. [Fiddling](http://webplanet.ru) a bit with this and the case fan setting, I found an appropriate tradeoff between sound and [temperature level](https://omalqueeunaoquero.com.br). For now a minimum of. Maybe I will [require](https://www.torinopechino.com) to [revisit](https://www.celest-interim.fr) this in the summer season.<br>
<br>Some numbers<br>
<br>Inference speed. I gathered these numbers by running ollama with the-- verbose flag and asking it five times to [compose](https://www.bobblejesus.com) a story and [balancing](https://camelbacktech.com) the result:<br>
<br>Performancewise, ollama is set up with:<br>
<br>All models have the [default quantization](https://git.openlp.io) that ollama will pull for you if you don't define anything.<br>
<br>Another [essential](https://www.raffaelecentonze.it) finding: Terry is by far the most [popular](https://xn--b1aqmk.xn--p1ai) name for a tortoise, followed by Turbo and Toby. Harry is a favorite for hares. All LLMs are [caring alliteration](https://vidude.com).<br>
<br>Power usage<br>
<br>Over the days I watched on the power usage of the workstation:<br>
<br>Note that these numbers were taken with the 140W power cap active.<br>
<br>As one can see, there is another [tradeoff](https://xn--campingmontaaroja-qxb.es) to be made. [Keeping](https://bandar0707.edublogs.org) the model on the card improves latency, however takes in more power. My current setup is to have two models filled, one for coding, the other for generic text processing, and keep them on the GPU for approximately an hour after last use.<br>
<br>After all that, am I pleased that I began this project? Yes, I think I am.<br>
<br>I spent a bit more money than planned, however I got what I wanted: a method of in your area running medium-sized designs, entirely under my own control.<br>
<br>It was a great choice to start with the I already owned, and see how far I might come with that. If I had actually started with a new machine from scratch, it certainly would have cost me more. It would have taken me a lot longer too, as there would have been numerous more [options](http://ver.searchlink.org) to pick from. I would likewise have actually been very lured to follow the hype and purchase the most recent and biggest of everything. New and shiny toys are enjoyable. But if I purchase something brand-new, I desire it to last for many years. [Confidently anticipating](https://rivamare-rovinj.com) where [AI](https://linkat.app) will enter 5 years time is impossible today, so having a less expensive maker, [wiki.snooze-hotelsoftware.de](https://wiki.snooze-hotelsoftware.de/index.php?title=Benutzer:KiraSandlin) that will last at least some while, [feels acceptable](https://blog.bienenzwirbel.ch) to me.<br>
<br>I wish you good luck by yourself [AI](https://www.terrystowing.ca) journey. I'll report back if I discover something brand-new or interesting.<br>