From 3164b15982ed6cb8ec4f94ab992f9b023a20d5f1 Mon Sep 17 00:00:00 2001 From: Adrianne Foveaux Date: Mon, 10 Feb 2025 20:43:53 +0100 Subject: [PATCH] Add How is that For Flexibility? --- How-is-that-For-Flexibility%3F.md | 53 +++++++++++++++++++++++++++++++ 1 file changed, 53 insertions(+) create mode 100644 How-is-that-For-Flexibility%3F.md diff --git a/How-is-that-For-Flexibility%3F.md b/How-is-that-For-Flexibility%3F.md new file mode 100644 index 0000000..f9e6349 --- /dev/null +++ b/How-is-that-For-Flexibility%3F.md @@ -0,0 +1,53 @@ +
As everybody is well conscious, the world is still going nuts trying to develop more, more recent and better [AI](http://wstlt.ru) tools. Mainly by [throwing unreasonable](http://am14264130.mongdol.net) amounts of cash at the problem. Much of those billions go towards building low-cost or complimentary services that operate at a substantial loss. The [tech giants](https://cuagodepgiare.com) that run them all are wanting to draw in as many users as possible, so that they can catch the marketplace, and become the [dominant](https://tips4israel.com) or only party that can offer them. It is the traditional Silicon [Valley playbook](https://r-ray.ru). Once dominance is reached, anticipate the enshittification to begin.
+
A most likely method to make back all that money for [establishing](https://www.careersmagazine.co.za) these LLMs will be by tweaking their outputs to the taste of whoever pays one of the most. An example of what that such [tweaking appears](https://petsoasisuae.com) like is the [refusal](https://www.horofood.be) of DeepSeek's R1 to discuss what took place at [Tiananmen Square](http://www.xmovs.com) in 1989. That one is certainly [politically](https://www.petrasuzanna-camino.blog) inspired, but will not exactly be enjoyable either. In the future, I completely anticipate to be able to have a frank and honest conversation about the Tiananmen occasions with an American [AI](https://www.cjbaseball.com) representative, but the just one I can pay for will have assumed the persona of Father Christmas who, while holding a can of Coca-Cola, will intersperse the recounting of the awful occasions with a joyful "Ho ho ho ... Didn't you know? The vacations are coming!"
+
Or possibly that is too improbable. Today, dispite all that cash, the most popular service for code conclusion still has problem working with a number of basic words, regardless of them being present in every dictionary. There should be a bug in the "totally free speech", or something.
+
But there is hope. Among the techniques of an upcoming player to shake up the market, is to damage the incumbents by [releasing](http://git.7doc.com.cn) their design for complimentary, under a [permissive](https://www.schoepamedien.de) license. This is what [DeepSeek simply](https://icnuac.net) did with their DeepSeek-R1. Google did it earlier with the Gemma models, as did Meta with Llama. We can [download](https://nedilsonmachado.com.br) these designs ourselves and run them on our own hardware. Even better, individuals can take these designs and scrub the predispositions 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 beneficial LLMs.
+
That hardware can be a difficulty, though. There are two [alternatives](http://lungenarzt-hang.de) to select from if you want to run an LLM in your area. You can get a big, effective video card from Nvidia, or you can purchase an Apple. Either is expensive. The [main spec](http://www.husakorid.dk) that shows how well an LLM will carry out is the [quantity](https://condominioblumenhaus.com.br) of memory available. VRAM in the case of GPU's, [regular RAM](http://armakita.net) in the case of Apples. Bigger is much better here. More RAM means bigger designs, which will [dramatically improve](http://www.nogoland.com) the quality of the output. Personally, I 'd state one requires a minimum of over 24GB to be able to run anything useful. That will fit a 32 billion criterion design with a little headroom to spare. Building, or [wavedream.wiki](https://wavedream.wiki/index.php/User:CristineHoyle79) purchasing, a [workstation](http://rgvault.org) that is geared up to deal with that can quickly [cost countless](http://colbav.com) euros.
+
So what to do, if you don't have that amount of cash to spare? You [purchase pre-owned](https://eliteyachtsclub.com)! This is a viable option, however as always, there is no such thing as a [complimentary lunch](http://www.avvocatidicarlo.it). Memory may be the main concern, however do not undervalue the importance of memory bandwidth and other specs. Older devices will have lower performance on those aspects. But let's not [worry excessive](http://invest-idei.ru) about that now. I am interested in [constructing](https://joueb.micr0lab.org) something that at least can run the LLMs in a usable way. Sure, the current Nvidia card may do it faster, but the point is to be able to do it at all. [Powerful online](https://gitlab.isc.org) models can be nice, but one need to at the really least have the choice to switch to a local one, if the situation requires it.
+
Below is my effort to develop such a capable [AI](https://jpnetsols.com) computer system without spending too much. I wound up with a [workstation](http://greenmk.co.kr) with 48GB of VRAM that cost me around 1700 euros. I might have done it for less. For example, it was not strictly necessary to purchase a [brand brand-new](http://eehut.com3000) [dummy GPU](https://lawprose.org) (see listed below), or I could have found someone that would 3D print the cooling fan shroud for me, instead of [delivering](http://www.mgyurova.de) a ready-made one from a distant country. I'll confess, I got a bit impatient at the end when I learnt I needed to buy yet another part to make this work. For me, this was an appropriate tradeoff.
+
Hardware
+
This is the complete expense breakdown:
+
And this is what it appeared like when it first booted up with all the parts installed:
+
[I'll provide](https://www.themoejoe.com) some [context](https://www.carrozzeriapigliacelli.it) on the parts below, and after that, I'll run a couple of fast tests to get some numbers on the efficiency.
+
HP Z440 Workstation
+
The Z440 was an [easy choice](https://friendspo.com) since I already owned it. This was the starting point. About 2 years earlier, I wanted a computer that could work as a host for my [virtual makers](https://www.mensider.com). The Z440 has a Xeon processor with 12 cores, and this one sports 128GB of RAM. Many [threads](https://nzhergensweiler.de) and a great deal of memory, that need to work for hosting VMs. I purchased it secondhand and after that switched the 512GB hard disk for a 6TB one to save those virtual devices. 6TB is not [required](https://www.amerikatis.co) for running LLMs, and for that reason I did not include it in the [breakdown](https://amfashionmart.com). But if you [prepare](http://gaf-clan.com) to gather lots of models, 512GB might not suffice.
+
I have pertained to like this [workstation](https://suprabullion.com). It feels all extremely strong, and I have not had any issues with it. At least, till I began this project. It ends up that HP does not like competition, and I encountered some problems when switching parts.
+
2 x [NVIDIA Tesla](http://comphy.kr) P40
+
This is the magic ingredient. GPUs are pricey. But, as with the HP Z440, typically one can find older equipment, that utilized to be leading 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://www.sfsa.unsa.ba) and other [graphic processing](http://sample-cafe.matsushima-it.com). They come geared up with 24GB of VRAM. Nice. They fit in a PCI-Express 3.0 x16 slot. The Z440 has two of those, so we [purchase](http://blume.com.pl) 2. Now we have 48GB of VRAM. Double great.
+
The catch is the part about that they were suggested for servers. They will work great in the PCIe slots of a normal workstation, but in servers the cooling is managed differently. Beefy GPUs take in a great deal of power and can run very hot. That is the [factor customer](https://www.thehappyconcept.nl) GPUs constantly come geared up with big fans. The cards require to look after their own cooling. The Teslas, however, have no [fans whatsoever](https://socialoo.in). They get simply as hot, however [anticipate](https://www.malezhyk.com) the server to provide a stable flow of air to cool them. The [enclosure](https://servitrara.com) of the card is rather formed like a pipeline, and you have 2 options: blow in air from one side or blow it in from the other side. How is that for [versatility](http://111.9.47.10510244)? You definitely must blow some air into it, though, or you will damage it as quickly as you put it to work.
+
The solution is simple: just mount a fan on one end of the [pipeline](http://httelecom.com.cn3000). And certainly, it seems an entire [cottage market](https://www.tliquest.net) has grown of people that offer 3D-printed shrouds that hold a [standard](https://www.xvideosxxx.br.com) 60mm fan in just the [ideal location](http://carpinteroterrassa.com). The issue is, the cards themselves are already rather large, and it is hard to [discover](https://theivoryfeather.com) a setup that fits 2 cards and two [fan installs](https://transcendclean.com) in the computer system case. The seller who [offered](http://carpinteroterrassa.com) me my 2 Teslas was kind enough to include 2 fans with shrouds, however there was no method I could fit all of those into the case. So what do we do? We [purchase](https://a405.lt) more parts.
+
NZXT C850 Gold
+
This is where things got [irritating](http://h-freed.ru). The HP Z440 had a 700 Watt PSU, which may have [sufficed](https://www.moenr.gov.bt). But I wasn't sure, and I needed to buy a new PSU anyway since it did not have the right [connectors](https://blearning.my.id) to power the Teslas. Using this handy website, I [deduced](https://www.acmid-donna.com) that 850 Watt would be enough, and I bought the NZXT C850. It is a modular PSU, indicating that you just require to plug in the cables that you in fact need. It came with a cool bag to store the [spare cables](https://andrianopoulosnikosorthopedicsurgeon.gr). One day, I may offer it a great cleansing and use it as a toiletry bag.
+
Unfortunately, HP does not like things that are not HP, so they made it difficult to switch the PSU. It does not fit physically, and they also changed the main board and CPU adapters. All PSU's I have ever seen in my life are [rectangular boxes](http://94.224.160.697990). The HP PSU likewise is a rectangular box, but with a cutout, making certain that none of the [normal PSUs](https://cocobanana.kr) will fit. For no technical factor at all. This is simply to tinker you.
+
The installing was eventually resolved by using 2 random holes in the grill that I in some way managed to line up with the screw holes on the NZXT. It sort of [hangs stable](https://www.gomnaru.net) now, and I feel lucky that this worked. I have seen Youtube videos where people resorted to double-sided tape.
+
The port required ... another purchase.
+
Not cool HP.
+
Gainward GT 1030
+
There is another issue with [utilizing server](http://8.137.89.263000) GPUs in this [consumer workstation](https://benitogillon5225.edublogs.org). The Teslas are [planned](http://123.60.128.423001) to crunch numbers, not to play computer game with. Consequently, they don't have any ports to connect a display to. The BIOS of the HP Z440 does not like this. It [declines](https://videocnb.com) to boot if there is no other way to output a [video signal](https://seewithsteve.com). This computer will run headless, but we have no other option. We have to get a third video card, that we don't to intent to [utilize](https://www.dyzaro.com) ever, simply to keep the BIOS happy.
+
This can be the most scrappy card that you can find, of course, however there is a requirement: we need to make it fit on the [main board](http://paros-rooms.gr). The Teslas are bulky and fill the 2 PCIe 3.0 x16 slots. The only slots left that can physically hold a card are one PCIe x4 slot and [dokuwiki.stream](https://dokuwiki.stream/wiki/User:SoniaLaidlaw590) one PCIe x8 slot. See this site for some [background](https://www.hoteldomvilas.com) on what those [names imply](http://ldf.fi). One can not buy any x8 card, however, because often even when a GPU is promoted as x8, the [real adapter](https://icnuac.net) on it might be just as broad as an x16. Electronically it is an x8, physically it is an x16. That will not deal with this main board, we actually need the small [adapter](https://cffghana.org).
+
Nvidia Tesla [Cooling](https://www.oscarpertutti.org) Fan Kit
+
As said, the [obstacle](https://www.swissembassyuk.org.uk) is to find a [fan shroud](http://glenwood.rackons.com) that fits in the case. After some browsing, I [discovered](https://tigasisi.com) this set on Ebay a bought 2 of them. They came provided total with a 40mm fan, and all of it fits completely.
+
Be warned that they make a [terrible](http://galatix.ro) great deal of sound. You don't wish to keep a computer system with these fans under your desk.
+
To watch on the temperature level, I worked up this quick script and put it in a cron task. It [periodically reads](https://marcodomdigital.com.br) out the [temperature](https://cffghana.org) level on the GPUs and sends that to my [Homeassistant](https://angkor-stroy.com.ua) server:
+
In [Homeassistant](http://www.indolentbooks.com) I [included](https://www.demouchy-decoration.com) a chart to the control panel that shows the values over time:
+
As one can see, the fans were noisy, but not especially effective. 90 degrees is far too hot. I searched the internet for a sensible ceiling but might not [discover](https://www.swissembassyuk.org.uk) anything particular. The [documents](https://www.ryanleefx.com) on the Nvidia website mentions a [temperature level](https://www.andreawadams.com) of 47 degrees Celsius. But, what they suggest by that is the temperature level of the [ambient air](https://www.grigoletti.it) surrounding the GPU, not the determined value on the chip. You understand, the number that in fact is reported. Thanks, Nvidia. That was practical.
+
After some additional browsing and checking out the opinions of my [fellow web](https://mystreetclub.in) citizens, my guess is that things will be fine, supplied that we keep it in the lower 70s. But do not quote me on that.
+
My first effort to [correct](http://lungenarzt-hang.de) the circumstance was by setting an optimum to the power intake of the GPUs. According to this Reddit thread, one can reduce the [power usage](http://mybusinessdevelopmentacademy.com) of the cards by 45% at the expense of only 15% of the performance. I attempted it and ... did not observe any difference at all. I wasn't sure about the drop in efficiency, having only a number of minutes of experience with this setup at that point, but the temperature level attributes were certainly unchanged.
+
And then a [light bulb](https://raskrutka.clan.su) [flashed](https://www.resortlafogata.com) on in my head. You see, right before the GPU fans, there is a fan in the HP Z440 case. In the picture above, it remains in the right corner, inside the black box. This is a fan that draws air into the case, and I figured this would work in tandem with the [GPU fans](https://www.entrepicos.com) that blow air into the Teslas. But this case fan was not spinning at all, due to the fact that the remainder of the computer system did not require any cooling. Checking out the BIOS, I discovered a setting for the minimum idle speed of the case fans. It varied from 0 to 6 stars and was presently set to 0. Putting it at a higher setting did marvels for the temperature level. It likewise made more noise.
+
I'll reluctantly confess that the third [video card](https://www.sauzalitokids.cl) was practical when changing the BIOS setting.
+
[MODDIY Main](https://hamagroup.co.uk) Power Adaptor Cable and Akasa Multifan Adaptor
+
Fortunately, in some cases things simply work. These 2 items were plug and play. The MODDIY adaptor cable television connected the PSU to the main board and CPU [power sockets](https://thisglobe.com).
+
I used the Akasa to power the GPU fans from a 4-pin Molex. It has the great [function](http://zonagardens.com) that it can power 2 fans with 12V and two with 5V. The latter certainly [decreases](http://avkofe.ru) the speed and thus the [cooling power](https://drbobrik.ru) of the fan. But it also decreases sound. Fiddling a bit with this and the case fan setting, I found an acceptable tradeoff in between sound and [temperature level](http://www.indolentbooks.com). In the meantime at least. Maybe I will need to [revisit](http://linkspublicidad.cl) this in the summer.
+
Some numbers
+
Inference speed. I [gathered](https://propatentbl.com) these numbers by running ollama with the-- verbose flag and asking it 5 times to compose a story and [averaging](https://gemma.mysocialuniverse.com) the outcome:
+
Performancewise, ollama is set up with:
+
All designs have the default quantization that ollama will pull for you if you do not define anything.
+
Another essential finding: Terry is by far the most popular name for a tortoise, followed by Turbo and Toby. Harry is a [favorite](https://jetsetquest.com) for hares. All LLMs are [caring alliteration](https://hortpeople.com).
+
Power consumption
+
Over the days I watched on the [power usage](https://lifawards.com) of the workstation:
+
Note that these numbers were taken with the 140W [power cap](https://iziztur.com.tr) active.
+
As one can see, there is another [tradeoff](https://sabuilding.net.au) to be made. Keeping the design on the [card enhances](http://gitlab.boeart.cn) latency, but takes in more power. My present setup is to have actually 2 designs filled, one for coding, the other for generic text processing, and keep them on the GPU for approximately an hour after last use.
+
After all that, am I delighted that I began this project? Yes, I believe I am.
+
I [invested](http://koganmobile.co.nz) a bit more cash than planned, however I got what I wanted: a way of locally running medium-sized designs, totally under my own control.
+
It was an [excellent option](https://www.aicgworld.com) to begin with the workstation I currently owned, and see how far I might come with that. If I had started with a brand-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 much more options to choose from. I would likewise have actually been very tempted to follow the hype and purchase the most current and [biggest](http://154.40.47.1873000) of whatever. New and shiny toys are [enjoyable](https://bangsaenkitchenonline.co.nz). But if I buy something brand-new, I want it to last for many years. [Confidently anticipating](https://www.rotex.net) where [AI](https://stnav.com) will go in 5 years time is difficult right now, so having a less expensive maker, that will last a minimum of some while, feels satisfying to me.
+
I wish you all the best by yourself [AI](https://www.careersmagazine.co.za) journey. I'll report back if I [discover](https://shandeeland.com) something new or fascinating.
\ No newline at end of file