BenoitP

56 ms·

511.
by BenoitP 6y ago
Thank you for your work! Are there current efforts in your research going in mainstream RDBMS (say postgres)? The space improvements are so great columns could just be indexed by default.
512.
by BenoitP 6y ago
UPMEM They're etching a small CPU in each block of 4GB of RAM. This means a vast compute power at low electrical power. An Intel CPU spends 650 pJ transporting data from RAM, and then 3 pJ doing an addition. Data IO bandwidth and laten
513.
by BenoitP 6y ago
> HW embedded within the memory that could be given a small (safe) executable program to process the memory in-place rather than evaluating it on the CPU Well, that seems to be the exact definition of what UPMEM is doing: https:/&#
514.
by BenoitP 6y ago
> put carbon taxes on BTC/fiat exchanges Almost the best solution that preserves bitcoin, which is: put a tax on BTC mining. The Bitcoin consensus rules rely on miners proving their faith in Bitcoin up to the amount of resources the
515.
by BenoitP 6y ago
What we want could be a complexity metric corresponding to a set of lines. I'd say that's something that could be ML-learnable with the right dataset. Given an answer from stackoverflow, ask an average developer to identify what w
516.
by BenoitP 6y ago
Their marketing material states: "Facilitating machines to go beyond pattern recognition and into cause-and-effect learning". I wonder what they are referring to. Are they accelerating what SHAP's GradientExplainer [1] does?
517.
by BenoitP 6y ago
Well, in case of a global thermonuclear war I think we'll have bigger problems than some transactions being reversed.
518.
by BenoitP 6y ago
Would you mind elaborating on that; besides running out of money, what were the things that went wrong? Or the things that didn't happen when needed to?
519.
by BenoitP 6y ago
> performance should dictate the algorithms, not the other way You want correctness first, and performance second. But these two are very much intertwined. And knowing exactly where the boundary is will help you co-design them. Some AWS
520.
by BenoitP 6y ago
Sorry for the harsh comment. I've been reading about your columnar compression pipeline [1], and it sort of makes sense if the comparison is against a regular row-oriented DB. AWS Timestream must really be doing something wrong here, o
521.
by BenoitP 6y ago
Pre-reading hypothesis: TimescaleDB is declared orders of magnitude faster because the benchmark is serving results they're computing at writing time? Is it just like the ClickHouse benchmark from earlier, where they read from a `CREAT
522.
by BenoitP 6y ago
The price surge being news, and the news driving the price surge, IMHO. Maisie Williams asked her 2.7M followers recently [1]: "should i go long on bitcoin" Maybe that was a paid tweet, maybe not. 900k people answered in the poll
523.
by BenoitP 6y ago
You'd need a lot of resources for pumping right now. Or maybe it's already happened quite a few times and everybody is doing it now, sort of like a meme?
524.
by BenoitP 6y ago
18 GB on-chip SRAM (at L1 latency), they say on the homepage. Looks like RAM with local compute. Another startup is sort of doing this (Arm cores etched in the usual DRAM silicon process [1]). I wonder if this is a long term trend in comput
525.
by BenoitP 6y ago
This is sort of what I'm getting at. The one-thread-per-core part should bring little to no benefit vs JDK+Loom, for example.
526.
by BenoitP 6y ago
Nice! It'd be great to know where do the improvements came from. Is it from a different architectural design, is it from better exploitation of new hardware, is it from JVM limitations? For example, when Valhalla and Loom land in the J
527.
by BenoitP 6y ago
Oh, I forgot to mention that megabases currently are CPU limited. A truly new kind of megabase would be enabled by it.
528.
by BenoitP 6y ago
My secret wish is that they start working on Factorio 2 with a new item state-memoizing game engine. Let me explain. Conway's game of life organisms' different states can be memoized. Starting the simulation slowly, an engine can
529.
by BenoitP 6y ago
Lots of talks about `Domains`. So I looked it up [1] and here is what this is about: > the virtual machine contains a number of domains, each running as a separate system thread in the same address space. Each domain has a relatively sma
530.
by BenoitP 6y ago
> I'm not convinced maintainers are perceived as valuable. I'm pretty sure an accountant classifies technical debt control as not being valuable. "A cost center is a department or function within an organization that does
531.
by BenoitP 6y ago
This is to be read in the context of the recent terrorist attacks (Conflans, Nice in France; and Vienna in Austria) in Europe. Which were flagged on HN.
532.
by BenoitP 6y ago
> are interoperable in Rust (block_on) Blocking an OS thread as a mean to be compatible is not exactly what we're trying to do here. > Except the performance difference between a kernel-mode context switch and a user-mode one is
533.
by BenoitP 6y ago
> Is there a mechanism to group some goroutines onto a single CPU thread to avoid locking? Python's GIL comes to mind :p
534.
by BenoitP 6y ago
They're not the solution to well-sliced data processing, but I can't see how they would hurt the performance.
535.
by BenoitP 6y ago
> Basically just an ergonomics issue. A viral one, and number-of-possible-states-exploding one at that. > First, in Rust this isn't really a problem. You can always turn async calls into blocking ones in Rust by calling block_on
536.
by BenoitP 6y ago
Or Kraken the cryptocurrency exchange ( https://www.kraken.com/ )
537.
by BenoitP 6y ago
I'll go against the grain here and say that async/await, wether implemented by one thread-per-core like here or by stackless coroutines is not the solution. Async/await will make complexity explode because of the colored func
538.
by BenoitP 6y ago
> tools take a lot to learn well and it rarely transfers between roles at the same job So, great lock-in capabilities? I can see why money is pouring into these tools.
539.
by BenoitP 6y ago
I have that that kind of divisive discussion. But we did compromise. What we ended up doing was using an ORM for CRUD operations, which were mostly user-facing ('row' operations); and SQL for reports / data monitoring ('
540.
Induced subgraphs of hypercubes and a proof of the Sensitivity Conjecture(arxiv.org)
3 points by BenoitP 6y ago|0 comments

More ›