Hacker Newsnew | past | comments | ask | show | jobs | submit | Mindless2112's commentslogin

As of b8966, it is still not great.

  | model                 |      size |  params | backend | ngl |   test |            t/s |
  | --------------------- | --------: | ------: | ------- | --: | -----: | -------------: |
  | gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | SYCL    | 999 | pp2048 |  851.81 ± 6.50 |
  | gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | SYCL    | 999 |  tg128 |   42.05 ± 1.99 |
  | gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan  | 999 | pp2048 | 2022.28 ± 4.82 |
  | gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan  | 999 |  tg128 |  114.15 ± 0.23 |
  | qwen35 27B Q6_K       | 23.87 GiB | 26.90 B | SYCL    | 999 | pp2048 |  299.93 ± 0.40 |
  | qwen35 27B Q6_K       | 23.87 GiB | 26.90 B | SYCL    | 999 |  tg128 |   14.58 ± 0.06 |
  | qwen35 27B Q6_K       | 23.87 GiB | 26.90 B | Vulkan  | 999 | pp2048 |  581.99 ± 0.86 |
  | qwen35 27B Q6_K       | 23.87 GiB | 26.90 B | Vulkan  | 999 |  tg128 |   10.64 ± 0.12 |
Edit: I've no idea why one would use gpt-oss-20b at Q8, but the result is basically the same:

  | model                 |      size |  params | backend | ngl |   test |            t/s |
  | --------------------- | --------: | ------: | ------- | --: | -----: | -------------: |
  | gpt-oss 20B Q8_0      | 11.27 GiB | 20.91 B | SYCL    | 999 | pp2048 |  854.16 ± 6.06 |
  | gpt-oss 20B Q8_0      | 11.27 GiB | 20.91 B | SYCL    | 999 |  tg128 |   44.02 ± 0.05 |
  | gpt-oss 20B Q8_0      | 11.27 GiB | 20.91 B | Vulkan  | 999 | pp2048 | 2022.24 ± 6.97 |
  | gpt-oss 20B Q8_0      | 11.27 GiB | 20.91 B | Vulkan  | 999 |  tg128 |  114.02 ± 0.13 |
Hopefully, support for the B70 will continue to improve. In retrospect, I probably should have bought a R9700 instead...

"I've no idea why one would use gpt-oss-20b at Q8" - would you mind expanding on this comment?

In that particular model family, the choices are 20B and 120B, so 20B higher quant fits in VRAM, while you'd be settling for 120B at a lower quant. Is it that 20B MXFP4 is comparable in performance so no need for Q8?

Or is the insight simply that there are better models available now and the emphasis is on gpt-oss-20b, not Q8?


The parameters in the original gpt-oss-20B model are "post-trained with MXFP4 quantization", so there just isn't much to gain by quantizing to Q8. If you look inside the Q8 model, most of the parameters are MXFP4 anyway.

Though, looking inside my "gpt-oss 20B MXFP4 MoE" model, it looks to also be quantized the same way as the Q8, so that was probably an overstatement on my part.

Still, the Q8 is 12.1 GB and the FP16 is 13.8 GB. Not the ~1:2 ratio you might expect.


At this speed, people end up paying more on electricity than api calls. (California electricity)

Compared to a B70, a 5090 is 1x the memory with 3x the bandwidth at 4x the price. Yeah, the 5090 is better, but you're paying for it.

On actual market it’s $1100 vs $3200 now, right? I actually got mine at $2200 at cost in the before days.

Current lowest price for a new card on Newegg: $949.99 vs $3,699.99.

Wow, 5090 prices have exploded. Thanks for looking. I should have known by hardware price intuition is broken.

Never buy another ebook from Amazon, sure.

Never buy another Kindle? I keep mine in airplane mode all the time and sideload all the books/papers I want to read. It works practically just as well as when I bought it. Why wouldn't I buy another? If Amazon makes a Kindle with color at 300 PPI, I will.

Sure, proper EPUB support would be nice, but if I need that I can jailbreak and install Koreader.

If there's another device with comparable hardware/software/battery, I'd consider it. AFAIK, Kindle still has the best standby battery life.


The demonym for France is "French," so it's not wrong (even if it doesn't sound right.)


It's not completely wrong, it will be understood, but it is ungrammatical and a clear marker that the speaker is not native, similar to getting adjectives in the 'wrong' order ('a big tasty sandwich' sounds more natural to a native speaker than 'a tasty big sandwich', even though the latter makes sense and will be understood).

Demonyms for historical neighbours of England have irregular forms when speaking of a particular person from there. Scotland has 'Scot' and 'Scotsman'; Wales has 'Welshman'; Spain has 'Spaniard'. Other countries indeed need a second word, such as 'person' or 'citizen' ('a Chinese' sounds offensive to me; I would say 'a Chinese person' in all cases). The only country I can think of where using a bare demonym is grammatical when speaking of a single person from there is Germany with 'a German' - probably because it has the suffix -man.

Edit: A sibling comment pointed out that 'an American' is grammatical, and thinking about it, I think the suffix -an is what makes bare demonyms grammatical - you can say 'an Angolan', 'a Laotian', 'a Peruvian', 'a Moroccan', etc, but wouldn't say 'a Thai', 'a Swedish', 'a Sudanese', etc.


> but it is ungrammatical and a clear marker that the speaker is not native

You mean a native speaker might be ungrammatical when using their non-native language? That makes sense to me.

> Spain has 'Spaniard'.

Even so, you'll hear a ton of native Spanish people saying "As a Spanish person" or "As person from Spain" instead of simply "As a Spaniard", I'm not sure this is very surprising. If anything, that mistake makes it more likely they're a native than not, in the case of Spain, as the level of English outside of metropolitan areas is lacking at best, compared to other European countries.


I'm using the words 'grammatical' and 'ungrammatical' in a linguistic sense; human languages are subtle and fluid, and one doesn't have to be far along the sliding scale between 'doesn't speak a word' and 'well-educated native speaker' to be understood. We speak of 'broken' English when somebody is able to be understood but hasn't fully grasped the language yet; using demonyms incorrectly is a subtler flavor of the same thing. For example 'no come here' -> 'no entering' -> 'no entry'


> but wouldn't say 'a Thai', 'a Swedish', 'a Sudanese'

You also don't say 'a Japanese' but that is an extremely common error with Japanese English speakers when they are first learning.

I am looking for a citation, but I seem to recall the casual rule of thumb is something to do with the ending of the nationality (so '-ish', '-ese','-ch' etc. you can't put 'a' in front). I think the more formal explanation likely centers around rules relating to indefinite articles.


There are some suffixed with "-i" which sound natural to my (American) ears too: "an Israeli", "a Somali", "a Pakistani", "an Omani", etc.


> and a clear marker that the speaker is not native, similar to getting adjectives in the 'wrong' order

I would think that if you say you are French, then everyone know you aren't native anyway. Maybe it's actually a good way, it can distinguish between true natives and false natives


As a Welshman, I’d say North/South Walian are more common among the populace!


When speaking English, the French side of my family refers to themselves like that often, however, they're from Bretagne, so exactly how French they are is up for debate.


nah, "as a french" is perfectly fine english grammar.

english changes over time, and thats been a normal description for at least a decade


Technically yes the demonym is "French", but "I'm a French" just doesn't work in English. The word 'French' is almost exclusively used in English as an adjective or the name of the language. It is never used as a noun for anything else. So in context, it reads as an adjective without a paired noun.

In English, you have to disambiguate be adding a noun: French person, French citizen, or Frenchman if you're old and inconsiderate.

Similarly, we don't call people "a Chinese". That construction is considered derogatory, if not outright racist. Demonyms typically cannot be used as nouns alone without a suffix. "A Brazilian" or "a Spaniard" are acceptable.

As usual for English, the rules are vague and inconsistent.


> "A Brazilian" or "a Spaniard" are acceptable.

Well, context is important on the Brazilian front. ;)

"I had a brazilian at my house" could have other connotations.


How many did you have at your house?


> or Frenchman if you're old and inconsiderate.

Or talking about a man that is French. Neither of which would be considered 'old', or 'inconsiderate".


"Frenchman" (one word) is always... "old and inconsiderate" is a good description. "French man" (two words) is at times still appropriate.


No, Frenchman is fine. It's a neutral and common descriptor, and I've far more often heard it said by French people than about them. Similar to Spaniard.

Not everything needs to be perennially problematised.


It is wrong, in the sense that native speakers do not say that, and it triggers what linguists call a "grammaticality judgment" in a native speaker. Same as "He eat apple" or "I am go to school". These may be comprehensible utterances, but they do not fit into a native speaker's internal grammar of valid English sentence structures.

Yes, languages can change, but there is no evidence that native speakers have started saying "a French". The only context I've ever seen that in is "As a French...", which strongly implies a non-native speaker of English. The evidence suggests that it's a common language interference error from French, not some future development of English.


No.

"French" is adjective or a collective noun, but don't use it as a countable noun.

Trying to say "as a French" makes about as much sense as thinking "as a American" is correct.

As has already been said ... "a French (wo)man","a French person","a French citizen" is the correct way to go.

The reason you can say "an American" is because America starts with a vowel.

Same reason why you would not say "a British" but you could say "a Brit".


Demonyms don’t use the same rules as countable nouns. Both “French” and “British” are acceptable demonyms, they’re just not particularly idiomatic in American English (which likes to overcorrect with “person” like you’ve noted).

(There’s no particularly consistency with this, it’s just what sounds “good” to American ears. We’re perfectly fine with “as a German” or “as a Lithuanian.”)


> Both “French” and “British” are acceptable demonyms

No they are not.

The Oxford English Dictionary, for example makes it quite clear re. 'French':

    "With plural agreement, and frequently with 'the' French people regarded collectively ..."
I draw your attention to the first three words ... "with plural agreement".

It is explicitly telling you that "French" is a collective plural noun and hence cannot be used as a singular countable noun.


I think we’re past OED being a normative arbiter of what does or doesn’t pass for acceptable English usage.


Precisely so: the OED's role is descriptive, in that it is to describe English as it is used, not how it ought to be used. It provides evidence of a grammatical rule - it is not the rule itself.

You can speak however you like, there is no language police, but the fact is the average English speaker will perceive certain constructs to be grammatically incorrect. "He eat", "I driving", "a French" etc.


a French; an American; a Brit, or a British

sounds casual but correct to me


> sounds casual but correct to me

I don't care if it "sounds ok to me".

If you're going to make statements like that to go against what I've written then at least come up with some viable citations to grammar literature.

Honestly, in all my years on this earth I have never, ever heard anybody in any English speaking country I've spent time in say "a French" "a American" "a British".

And that amounts to a lot of time surrounded by people speaking VERY "casual" English.

P.S. I said "an American" was ok if you re-read.. an NOT a


The reason you can say "an American" has nothing to do with a vowel or not, there are just some demonyms that for some reason can be used like this, and some that can't.

For example:

* German is countable: https://www.oxfordlearnersdictionaries.com/definition/englis... * French is uncountable: https://www.oxfordlearnersdictionaries.com/definition/englis... * American is countable: https://www.oxfordlearnersdictionaries.com/definition/englis... * Spanish is uncountable: https://www.oxfordlearnersdictionaries.com/definition/englis...

But your explanation about why it is correct is bullshit, has nothing to do with "an" vs "a", the English language is just inconsistent as fuck and some demonyms can be used like this and some can't.


Funny they didn't include any CJK languages on their list.


I heard an anecdote that Qwen Coder works better when prompted in Korean - haven't tested it for myself though.


Deepseek will regularly spit out Chinese (汉字)during English sessions. They generally seem to be syntactically related but it makes me think that there's some overhead of using English with an engine that's primarily trained in Chinese.




It supports linking two Android devices.


This is huge. There's been 3rd party Signal library for this for years -- and for some reason I can't determine, the developers have opted NOT to do this.


Yeah, the Signal team's roadmap seems very strange to me as an outsider. There are some low hanging fruits which they just seem to refuse to fix.

And given how in this case Molly could fix it it cannot have been that hard to fix.


this is why molly.im was a lifesaver for me.. trying to move a family member from VIBER to SIGNAL and ran into the annoying roadblock of not being able to link an Android tablet to an Android phone like Viber can - but molly does it fine.


If you use Signal on your Android phone, can you link with Molly on another Android device (tablet) without Signal complaining?


I just tried it on a 2nd phone and it seems to work.


I got logged out of Signal after setting up Molly


Now reading this on the GitHub page:

"If you wish to use the same phone number for both Molly and Signal, you must register Molly as a linked device. Registering the same number independently on both apps will result in only the most recently registered app staying active, while the other will go offline."


Yeah, pretty sure that's what me and the other comment meant. Linked device, like using Signal on Desktop. Or Signal on iPad. Linking wasn't available on Signal for Android for some reason.

Specifically I'm using Signal as the main device, with Molly as the linked device on 2nd phone.


There was a time when we could say "our greenhouse gas emissions are nothing compared to regular biological processes," and yet here we are.


> “The outbound and cross-bound DDoS attacks can be just as disruptive as the inbound stuff,” Dobbin said. “We’re now in a situation where ISPs are routinely seeing terabit-per-second plus outbound attacks from their networks that can cause operational problems.”

ISPs are starting to feel the pain, so perhaps in the near future they will do something about it.


Perhaps, or perhaps not. Maybe if we held them accountable they would?


Consider applying for YC's Summer 2026 batch! Applications are open till May 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: