rthnbgrredf

10 ms·

1. 121.▲
   by rthnbgrredf 3y ago

   I think it comes down to personal preference. I prefer instant and snappy interfaces with as minimal latency as possible. Please provide me with an option to disable all animations and I'm likely fine with your interface.
2. 122.▲
   by rthnbgrredf 3y ago

   The issue here isn't specifically about the classification of memory, be it "unified memory," RAM, or VRAM. The primary concern is ensuring there's enough memory capacity for the models required for inference. The real q
3. 123.▲
   by rthnbgrredf 3y ago

   I already have a cli client, but how to talk to multiple different LLMs at the same time? I guess I can script something with tmux.
4. 124.▲
   by rthnbgrredf 3y ago

   Could anyone recommend an open-source tool capable of simultaneously sending the same prompt to various language models like GPT-4, Gemini, and Claude, and displaying their responses side by side for comparison? I tried chathub in the past,
5. 125.▲
   by rthnbgrredf 3y ago

   Now I understand where the Nigerian prince scam scheme stems from.
6. 126.▲
   by rthnbgrredf 3y ago

   Would be interesting what happened if the west put a 1 bln bounty on Putin + free citizenship in the U.S.
7. 127.▲
   by rthnbgrredf 3y ago

   This. People talk about crazy valuations all the time. I hear about crazy Bitcoin valuations for years and recently AI, yet it is more likely Bitcoin will hit 100k than anything else. (My personal opinion is that BTC has a fair value of clo
8. 128.▲
   by rthnbgrredf 3y ago

   I don't know for Intel, but for TSMC the marketing numbers work like this: Marketing, Gate pitch, Metal Pitch, Year 7nm, 60nm, 40nm, 2018 5nm, 51nm, 30nm, 2020 3nm, 48nm, 24nm, 2022 2nm, 45nm, 20nm, 2024 1nm, 42nm, 16nm, 2026
9. 129.▲
   by rthnbgrredf 3y ago

   No matter if the data supports it or not, all that your comment is telling me is that you don't have kids.
10. 130.▲
    BP manager's husband pleads guilty to US insider trading that led to divorce(reuters.com)

    2 points by rthnbgrredf 3y ago|0 comments
11. 131.▲
    by rthnbgrredf 3y ago

    Google Cloud SQL does support quite a few postgres extension, but unfortunately it does not support TimescaleDB, that's quite sad indeed. I hope that at some point in the future TimescaleDB could be merged into postgres to make it a na
12. 132.▲
    by rthnbgrredf 3y ago

    While that is true, there's also no guarantee that we do not see a stagnation of the world stock markets for decades. World indices are also highly concentrated to a few countries, e.g. US is more than 60%.
13. 133.▲
    by rthnbgrredf 3y ago

    You can do something better for 0$, just install the TimescaleDB postgres extension and execute: SELECT create_hypertable('public.challenge_1br', by_range('time')); Now, enjoy your better than Snowflake query performance
14. 134.▲
    by rthnbgrredf 3y ago

    To be fair, companies that merely launch VMs on GCP constitute a small fraction. GCP truly excels when you leverage its object storage, BigQuery, managed postgres database (which starts at 7$/month), and serverless solutions for cloud-
15. 135.▲
    by rthnbgrredf 3y ago

    I already pointed that out multiple times to Rocky Linux staff. They have no answer to this. The whole concept of Rocky is a big bet, that Red Hat won't pull the sources of non-GPL binaries e.g. in cloud instances and other public acce
16. 136.▲
    by rthnbgrredf 3y ago

    As I understand it a firing squad would mean instant death, which would be more humane than struggling for 22 mins. Not sure why this method is not preferable. Full disclosure: I'm against death penalty.
17. 137.▲
    by rthnbgrredf 3y ago

    The TQQQ ETF hasn't experienced a downturn as severe as the 2000 crash. Market crashes recur; it's unrealistic to expect a perpetual bull market. In a major crash, TQQQ could plummet by 99.9%, turning $1 million into $1,000. Being
18. 138.▲
    by rthnbgrredf 3y ago

    I think what is often missed is that many thought it's a good idea to just move stuff from baremetal to cloud VMs as is and then wondering about the bills. IMHO cloud starts to really pay off for cloud native workloads on serverless of
19. 139.▲
    by rthnbgrredf 3y ago

    I'm curious why this blog post is allowed to stay while other very similar postings regarding the topic get flagged to oblivion.
20. 140.▲
    by rthnbgrredf 3y ago

    Certainly, that's a fair assumption. However, advocates of nuclear weapons or, for instance, ordinary gun owners, can also cite statistics in their favor. In the US, there are significantly more fatalities due to car accidents compared
21. 141.▲
    by rthnbgrredf 3y ago

    The problem is that pro-nuclear advocates, like many on HN, always have a convincing explanation for why a particular nuclear contamination isn't as severe as it seems or isn't the fault of X. This was also the case with Fukushima
22. 142.▲
    by rthnbgrredf 3y ago

    This is interesting. While I think this approach may not be practical for a single computer due to poor performance, it could open new possibilities for distributed computing. Here, users across the internet could pool their graphics cards
23. 143.▲
    by rthnbgrredf 3y ago

    > And when Germany cut back on nuclear, renewables fail to absorb that and fossil fuel use went up. I've noticed this claim being made repeatedly, but it's simply not true. Germany is decreasing its fossile energy production lo
24. 144.▲
    by rthnbgrredf 3y ago

    I dislike the term 'hyperscale' as it lacks concrete meaning. We need a metric akin to kilobytes or megabytes for storage. Terms like 'kiloscale' and 'megascale' could better indicate the scaling range of a ser
25. 145.▲
    by rthnbgrredf 3y ago

    Could someone who remember the era of mainframe computing, which of the downsides could apply to the modern version of it (cloud desktop)?
26. 146.▲
    by rthnbgrredf 3y ago

    After traditional diets either failed or provided only temporary success, I've discovered a personalized approach that combines elements from various diets. This allows me to handle the rules of each individual diet more flexibly. My s
27. 147.▲
    by rthnbgrredf 3y ago

    How much longer does `rustc` typically take to compile the same programming tasks compared to `gcc`? Is it just "a bit" e.g. 50% slower or by an order of magnitude?
28. 148.▲
    by rthnbgrredf 3y ago

    But who has the power to remove the board?
29. 149.▲
    by rthnbgrredf 3y ago

    Or one can try to rent a cluster in the clouds.
30. 150.▲
    by rthnbgrredf 3y ago

    I recently compiled a list with more to read on the topic: https://github.com/madnight/awesome-category-theory