HN user

gerdesj

7,705 karma
Posts3
Comments3,613
View on HN

"Sounds like a plane taking off."

Helicopter. I live in Yeovil, Somerset, UK - there's a helicopter factory just down the road. I had a IBM "AS/400" or whatever they are called now in our computer room rack for a customer and it made nearly as much noise as everything else put together. It was clearly tuned for start up noise to impress because they would fire up in sequence, rise to a crescendo and then slow down in sequence to just a din instead of painfully loud.

A switch or PC server on boot will normally run up cooling fans instantly to max as a default protection mechanism until the "OS" has started and sensors read and then the fans will slow down to deal with the actual thermal load.

If you switch off your air conn, it gets noisy, quickly. Recently in the UK we are seeing routine temperatures around 30C and we broke 200 odd year records for temperatures a few weeks back. I know its even worse elsewhere but our infrastructure is not designed for this. Here we are at the same latitude as Calgary AB!

Actually, its ~119GB usable (I have one). You can shutdown a lot of unneeded services if you only use it remote which will trim a lot more fat. You can enable the RDP service if you don't want to sit in front of it but get a desktop interface.

The "shitty" network is 10Gb/s and wifi7! You get a twin QSFP28DDlol+++ (I jest) that each run at 200Gb/s - not for the casual home user but handy at work, although I "only" have 40Gb/s on my switches sigh. With and no switch two you can do a three node cluster with some careful networking. If you want to do more then a switch is needed and it will need to be pretty funky! That said you could wire them up in a circle and use VLANs and MSTP and accept less than 200Gb/s per link. You'll probably need Openvswitch and a lie down afterwards.

I'm not a fan of the Gnome desktop but it works well enough and I think the Nvidia customised Ubuntu is well thought out. You get all the complicated NVidia extras pre-installed, along with docker (full fat, not the Ubuntu one) for a fairly quick start. It includes Ubuntu Pro which is free for five systems anyway but its nice to see it pre-installed.

You can run quite decent models on this thing see: https://spark-arena.com/ Also see "DS4".

We blew abut £4000 on one and it will pay for itself in a few months. I tried pricing up an Apple thingie and the Store wouldn't offer me more than 96Gb of RAM and a delivery date in Q3 at the earliest. Our Spark rocked up next day. They seem to come in 1TB or 4TB SSD variants. 1TB is enough for me and saves a lot of cash - keep an eye on your model downloads and ruthlessly delete old experiments. docker system prune.

We went for the Asus variant that has active cooling and I stuck it in the ceiling cable tray over our computer room racks. It sits on 1½" stainless steel mesh with lots of clearance in an actively cooled environment.

Replace the word "ejected" with "removed" as appropriate and the error becomes apparent.

No one "ejects" an external hard disc, you remove or disconnect it. You do eject a CD or a DVD.

Eject is a forceful removal ... if my memory is not totally shagged - iacto (Latin - I throw) or similar.

To be fair I have seen a few ejections or the aftermath of discs (hard/floppy/CD/DVD) and not considered them ... a disconnect event 8)

I suggest Apple go back to school and dispense with the lazy error reporting.

"most important technology in human history"

So far it rates "quite important" and certainly not "paradigm shift".

I've just spent most of the day wading through vibed blogs and what not to glean info on ... setting up LLMs! The content on the web is already in a parlous state and rapidly getting nicely but rather the same presentation and mostly right but somewhat wrong, often in crucial parts.

On the bright side I did find an absolute belter of a vibed site and it was only mildly wrong but I was able to sort that bit out and to be fair I think it was written up by an expert with help from a LLM and had a genuine human mistak in it.

Anyway, I suggest you might like to look at the internet as a whole as a paradigm shift and reserve judgement on LLMs and this decade's version of AI.

My old man used to say: "an army marches on its stomach".

Most people hereabouts will be wittering on about how best to kill people/do stuff with drones and forget that logistics is rather more complicated than killing things.

Fundamentally: people need food, water and shelter. Vehicles need fuel, mechanics and stuff. Killing people and destroying stuff needs ammunition or at least something sharp (and that will need something to keep it sharp).

Then you need to coordinate all this stuff, along with comms and a lot more details that I've not mentioned.

Then you need to persuade your troops to do their medieval best on the opposition and hope it works.

Given you find this is _normal_ for a six mile ambulance ride: "and $11,670 as a “base rate.”"

What on earth would you consider normal for a helicopter ride from Exeter to London?

That's roughly 150 miles as the crow flies. Pilot, co-pilot and a medic, minimum crew for say 1.5 hours. Each way, so 300 miles of fuel and aircraft lifetime and three hours of crew cost, not to mention ground crew etc.

My dad got that on the firm when the shit hit the fan and he needed to be seen by specialists in the Royal Brompton and Royal Devon and Exeter decided that was his best shot at life. That was 15 years ago.

Anyway, the OP's bill looked pretty normal until the 11,000 base rate nonsense. How can that possibly be justified?

Pint in England 17 days ago

Perhaps (citation needed)! Barrel usage in Britain is a lot older than the Roman occupation of Britain. A cup is not a barrel.

If I was to bet, cooper is probably an anglicized form of a brythonic language word. Any Welsh, Scottish, Irish, Cornish or even Cumbric (int al) speakers hanging around here?

You are playing with language badly. The three letters c-o-w sometimes end up close to each other in English but they don't mean a literal cow (moo)!

You might examine, say: coworker. A word I never use (because I'm a Brit) but it looks to me like cow-orker.

"informal, disparaging + offensive : an unpleasant, annoying, or vacuous woman cowy ˈkau̇-ē adjective"

That seems rather parochial (and misogynistic). Have you ever heard or said "cowy" yourself?

We do spend out quite a lot here in the UK for the BBC. They could easily dump a couple of expensive presenters and use the savings for vacuum tubes, if that is what is needed.

No idea where vacuum tubes were invented but I'm sure the BBC could find someone to make them.

"I imagine my work could be completed with AI assistance in a matter of days—maybe hours."

Would some one with tokens to burn mind checking that statement out and post back. Be sure to use long dashes too.

This tells you all you need to know about the "fox":

"This report was reviewed by Legal, who have asked us to clarify that the fox was depicted as over eighteen and that the sunglasses remained on throughout."

Senatus Populus ... Que Romani! The senate and people of Rome.

Italy, per se, is a modern construct but the capital is still Rome and I think it fair to draw associations to the old empire.

I love seeing the SPQR logo and I think it is a really useful link to the past.

I would encourage this sort of thing in my company. I'm not google. I'm not legally beholden to anyone except myself and my business partners ... and my own sense (which is worryingly odd!)

Google can never be exciting or interesting evermore by design and intent. They dived on in and went "money" full on. They exist to generate revenue for their shareholders. They dumped the "Don't be evil" thing without blushing.

That's recipient testing based on mailbox name. I don't recommend that for spammers - its so trite and early 2000s.

I wont allow you to test deliverability to my email domains without you sending an email I can analyze and decide to allow or drop mid stream. I also get to drop it before you consider it sent. I obviously drop connections that just establish from and to and go weird after that.

Get your buddy with his smoking Mac to allow multiple concurrent connections and see how it gets on compared to your Spark. Don't ever use a single "chat" test to derive performance - try running say 10 or more.

You might also notice that your Spark has a pair of QSFP28 or DD (not sure yet) type interfaces as well as the 10Gb/s ethernet - that network card is a right old beast and adds quite a lot to the cost. It is capable of either 200 or 400Mb/s and can be split into two lots of four. Your mate's Mac probably has a wifi connection and is too cool for ethernet 8)

That NIC is there for a good reason - the Spark wants some friends to cluster with and you will absolutely spank any Mac when you spaff Mac style money on say three of these beasts and some cables and cluster them up. If you want four or more, you will need a switch and Mikrotik and others have them.

Casual "tokens per second" in AI is a bit like gamers whittering on about "ping" when they are using TCP and UDP for their games. ICMP request/response is a handy way of testing network paths and can give some indications towards potential performance limitations.

The real problem is that OP is or wants to be an old school disruptor working at what used to be an exciting and disruptive employer (but isn't any more - its just a boring old money maker).

OP crank out a pretty decent and well received, by the community, product and get absolutely canned because they are well out of touch of how Google now works. You don't do risk (without reward) at Google and you certainly don't show a bit of ankle or look exciting. Google are well out of the market for being interesting (outside of the balance sheet and P&L for those who fetishise in accountancy.

Unfortunately: going viral isn't always a good thing as anyone who has experienced a nasty virus will attest.

A nvidia spark thingie has 128GB unified RAM. They also have a dual port version of one of these things: https://www.nvidia.com/content/dam/en-zz/Solutions/networkin.... ie 2 x 100GB/s ports, they may even be 2 x 200GB/s. Once I've got my paws on one, I'll know more.

You can cluster these beasts too. Two and three (with two IP subnets) is fairly obvious. Four or more might need a switch depending on how much network latency affects things.

Apple seem to have forgotten about M series with gobs of RAM. I can't get the Apple shop to show more than 96GB of unified RAM and that costs a kidney.