amirhirsch

Shout out to all my SF 5am crew checking if their overnight prs passed CI. Real 597 “member of technical staff” energy. I guess we should expect this, it is a Tuesday!

I like this too. I have been intentionally -maxxingmaxxing to get the meme out there. It's a good canary to sort out who gets the spicy takes from the pedestrians who probably still copy-paste into the ChatGPT web app like a psychopath.

I would love to sign up for antigravity cli but when I click on Get Plan it says: “This account isn't eligible for Google Antigravity benefits included with Google AI plans Google Antigravity benefits included with Google AI plans aren't available in some countries or for people under a certain age. Learn more about Google One feature eligibility.” With a button that says “Explore Google AI Plans” that when I click on it takes me to my Drive.

I can’t believe our Google account setup is different from any other startup in SF. Anyone have success with this? Do they even have a bot at Google that tracks this attrition?

ROTFLMAO made my day!

Let’s reduce this to absurdity:

I think you only wanted clarification of CTF (Capture the Flag) and not AI (Artificial Intelligence) and not GPT-4 (Generative Pre-Trained Transformer version 4) and not CLI (Command Line Interface) and not MCP (Model Context Protocol) and not LLM (Large Language Model)

Quoting TFA (The Fucking Article): “just adapt bro”

lol at the BGP example

MIT should spin up MIT.ai for-profit, raise 100B @ 1T, then go buy big computer.

At least in my experience (MIT ’06) many of the people most comfortable gaming academics ended up in finance.

I've always felt that it was these kind of folks that caused the 2008 financial crisis

Peter Turchin’s theory of “elite overproduction” suggests this is a cause for social instability and revolutions

Please have an endpoint for up-to-minute billing with hooks for notifications and setting limits, importantly have it correlate-with/be the endpoint that shows usage analytics.

Previous-co could never get argo billing to match argo analytics, and with no support from CF over months we backed away from CF completely in fear that scale-up would present new surprise unknown/untraceable costs

Previous-previous-co is probably the largest user of web worker

Dude is on hacker news on a Sunday. half the GDP of the world is competing with him. What metrics would you like to see?

Gp was referring to AN/TPY2 which is the THAAD radar.

Iron dome has nothing to do with that systems.

Any sufficiently advanced incompetence is indistinguishable from malice. If you are the editor of Gray’s Anatomy, incompetence is malice.

Critically, Amdahl’s Law remains a conjecture until we have the maths to disjoin complexity class P from NC.

Not sure about the conclusion regarding NVidia value capture. I imagine the context for many applications will come from a physical simulation environment running in dramatically more GPUs than the AI part.

Beat me to it. Now I have to delete my reply.

> S&S Deli in Cambridge

Good lunch spot for a nudnik

Here’s some other hints: combine hash stages 2 and 3, it can be two muladds and a XOR

For the first several rounds (when every tree value is in use) Combine the stage 5 XOR with the subsequent round’s tree XORs. You can determine even/odd in hash stage 5 starting with a ^ (a>>16) without Xoring the constant, then you can only need one XOR, this saves you a ton of XORs

Create separate instruction bundles for the first round, rounds 1-5 (combining hash stages 5 XOR with next round tree XORs) and 6-9 (not every tree node is used anymore), round 10 round 11-14 and round 15 and combine them.

you can use add_imm in parallel to load consts. stage 0 you have to do load the tree first and the vals, by later stages when everything is in scratch, you could use 12 scalar XORs and 6 vector XORs on scratch. once you vload vals, you can start to do XORs but can only advance so much at a time, so I’m starting to work on getting hash stages moving to different rounds faster to hide the initial vloads and get to the heavy load section sooner and spread the load pain.

I think I can hit #1 (current #1 is 1000). sub 900 not possible though.

Let me put down my thought process: You have to start to think of designing a 6-slot x8-len vector pipeline doing 48 hashes in parallel first which needs at least 10 steps —- if you convert three stages to multiply adds and do parallel XORs for the other three) —- the problem with 10 cycle hashing is you need to cram 96 scalar xors along side your vector pipeline, so that will use all 12 ALUs for 8 of those cycles. Leaving you only 24 more scalar ops per hash cycle which isn’t enough for the 48 tree value xors..

so you must use at least 11 steps per hash, with 96 xors (including the tree value xor) done in the scalar alus using 8 steps, and giving 3*12 Alu ops per hash cycle. You need 12 more ops per hash to do odd/even, so you must be 12 stages, and just do all of the hash ops in valu, 4 cycles of 12 alus doing modulo, 8 cycles x 12 alus free

With 12 steps and 48 parallel you’re absolute minimum could be 4096/48 x 12 = 1,024 cycles, since stage 10 can be optimized (you don’t need the odd/even modulo cycle, and can use some of those extra scalar cycles to pre-xor the constant can save you ~10 cycles. 1024 gonna be real hard, but I can imagine shenanigans to get it down to 1014, sub-1000 possible by throwing more xor to the scalar alus.

======================================================================

BROADCAST LOAD SCHEDULE

======================================================================

Round | Unique | Load Strategy

------|--------|------------------------------------------

   0  |    1   | 1 broadcast → all 256 items

   1  |    2   | 2 broadcasts → groups

   2  |    4   | 4 broadcasts → groups

   3  |    8   | 8 broadcasts → groups

   4  |   16   | 16 broadcasts → groups

   5  |   32   | 32 broadcasts → groups

   6  |   63   | 63 loads (sparse, use indirection)

   7  |  108   | 108 loads (sparse, use indirection)

   8  |  159   | 159 loads (sparse, use indirection)

   9  |  191   | 191 loads (sparse, use indirection)

  10  |  224   | 224 loads (sparse, use indirection)

  11  |    1   | 1 broadcast → all 256 items

  12  |    2   | 2 broadcasts → groups

  13  |    4   | 4 broadcasts → groups

  14  |    8   | 8 broadcasts → groups

  15  |   16   | 16 broadcasts → groups

Total loads with grouping: 839

Total loads naive: 4096

Load reduction: 4.9x

take advantage of index collisions, optimizing round 0 and 11, speculative pre-loading, and the early branch predictor (which now I am doing looking at bits output at stage 3)

I'm at 1137 with one hour with opus now... Pipelined vectorized hash, speculation, static code for each stage, epilogues and prologues for each stage-to-stage...

I think I'm going to get sub 900 since i just realized i can in-parallel compute whether stage 5 of the hash is odd just by looking at bits 16 and 0 of stage 4 with less delay.....

Very Nice. There is an issue with panning on the million point demo -- it currently does not redraw until the dragging velocity is below some threshold, but it should seem like the points are just panned into frame. It is probably enough to just get rid of the dragging velocity threshold, but sometimes helps to cache an entire frame around the visible range

I gave your reply the most generous interpretation and read it in the ironic way as you point out in the edit

this won't cost the city too much, there's only like a hundred kids under 6 in this city and 3% of them are mine.

I did write this 20 years ago https://fpgacomputing.blogspot.com/2006/05/methods-for-recon...

The vendor tools are still a barrier to the high-end FPGA's hardened IP

python won because of enforced whitespace. It solved a social problem that other languages punted to linters, baking readability into the spec

Probably also end of ZIRP and some “AI washing” to give the illusion of progress

The effect of these tools is people losing their software jobs (down 35% since 2020). Unemployed devs aren’t clamoring to go use AI on OSS.

Cool!

Constraint propagation from SICP is a great reference here:

https://sicp.sourceacademy.org/chapters/3.3.5.html

# Tell the driver to completely ignore the NVLINK and it should allow the GPUs to initialise independently over PCIe !!!! This took a week of work to find, thanks Reddit!

I needed this info, thanks for putting it up. Can this really be an issue for every data center?