Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Gemini 3.5 Flash

blog.google

Post reply on HN
Print view View original

491–500 of 692 posts

Re: Gemini 3.5 Flash

#491

Post by agnosticmantis »

reconnecting wrote: post #167
hosel wrote: post #124

Earlier quoted context omitted.

Can you explain what you mean?

LLM pre-training models risk being unable to be updated with data from after 2025, as much of it is corrupted with LLM-generated content. We might be locked into outdated knowledge, where only whitelisted sources decide what to include. Taking into account the sometimes blind belief that 'LLMs know everything', the outcome could be very costly, especially for technologies and businesses unfortunate enough to emerge a…

It may not be mainly or solely due to LLM pollution, but rather the fact that every publisher, (social) media company, newspaper, etc. clammed up and started charging (licensing) fees sometime in the last couple of years.

So maybe there's just not much openly available and new content worth training on that wasn't available prior to 2025.

Re: Gemini 3.5 Flash

#492

Post by gertlabs »

easygenes wrote: post #466

For those who would like to know the total and active parameter count of this model: even though Google doesn't disclose the model technicals, we can infer them within relatively tight margins based on what we do know. We know they serve the model on TPU 8i, which we have plenty of hard specs for (so we know the key constraints: total memory and bandwidth and compute flops). We can also set a ceiling on the compute c…

We've been really impressed with the performance of ~30B parameter class models and how close they are to the frontier from ~6-12 months ago, which begs the question, are the frontier labs really serving 10T parameter models? Seems unlikely.

If these Gemini 3.5 numbers are accurate, then I'd wager GPT 5.5 and Opus 4.7 are a lot smaller than people have speculated, too. It's not that frontier labs can't create a 5T+ parameter model, but they don't have the data to optimize a model of that size.

Gemini 3.5 Flash is really smart in one-shot coding reasoning, btw. Near the frontier. But it doesn't do so well in long horizon agentic tasks with arbitrary tool availability. This is a common theme with Google models, and the opposite of what we see with Chinese models (start dumb, iterate consistently toward a smart solution).

Data at https://gertlabs.com/rankings

Re: Gemini 3.5 Flash

#493

Post by WarmWash »

daemonologist wrote: post #485
easygenes wrote: post #466

For those who would like to know the total and active parameter count of this model: even though Google doesn't disclose the model technicals, we can infer them within relatively tight margins based on what we do know. We know they serve the model on TPU 8i, which we have plenty of hard specs for (so we know the key constraints: total memory and bandwidth and compute flops). We can also set a ceiling on the compute c…

If this is accurate it raises the question: why is this model so expensive? DeepSeek v4 Flash is 284B total/13B active, FP4/FP8 mixed, and only costs $0.14/$0.28 - even less from OpenRouter. Of course Gemini 3.5 Flash is most likely a better product, and therefore it can command a higher price from an economics perspective, but does this imply Google is taking roughly a 90% profit margin on inference? If so they're e…

Rumor is that GCP was happily selling compute to competitors. After all, under the hood, Google is closer to a federation than a corporation. The state of GCP doesn't care about the state of Gemini.

Re: Gemini 3.5 Flash

#494

Post by easygenes »

gertlabs wrote: post #492
easygenes wrote: post #466

For those who would like to know the total and active parameter count of this model: even though Google doesn't disclose the model technicals, we can infer them within relatively tight margins based on what we do know. We know they serve the model on TPU 8i, which we have plenty of hard specs for (so we know the key constraints: total memory and bandwidth and compute flops). We can also set a ceiling on the compute c…

We've been really impressed with the performance of ~30B parameter class models and how close they are to the frontier from ~6-12 months ago, which begs the question, are the frontier labs really serving 10T parameter models? Seems unlikely. If these Gemini 3.5 numbers are accurate, then I'd wager GPT 5.5 and Opus 4.7 are a lot smaller than people have speculated, too. It's not that frontier labs can't create a 5T+ p…

We know from NVIDIA's public Vera Rubin inference engine marketing materials that the frontier lab models are ~1-2T total.

Mythos is an exception that's larger.

Re: Gemini 3.5 Flash

#497

Post by blackoil »

xbmcuser wrote: post #233
hei-lima wrote: post #138

Earlier quoted context omitted.

We need another "Deepseek moment" or else it will become impossible for the regular dude to use AI. It will become something that only big companies can afford.

What we need is a deepseek moment in hardware ie China reaching parity on node size that is the only way latest computers let alone latest ai will be available to us in the future otherwise the profit margins will push most production to AI.

Open Source ASML EUV. But will wipe off trillions from US stocks so 401k may not like that.

Re: Gemini 3.5 Flash

#498

Post by happyopossum »

WarmWash wrote: post #493
daemonologist wrote: post #485

Earlier quoted context omitted.

If this is accurate it raises the question: why is this model so expensive? DeepSeek v4 Flash is 284B total/13B active, FP4/FP8 mixed, and only costs $0.14/$0.28 - even less from OpenRouter. Of course Gemini 3.5 Flash is most likely a better product, and therefore it can command a higher price from an economics perspective, but does this imply Google is taking roughly a 90% profit margin on inference? If so they're e…

Rumor is that GCP was happily selling compute to competitors. After all, under the hood, Google is closer to a federation than a corporation. The state of GCP doesn't care about the state of Gemini.

> Rumor is

It’s not a rumor - there are many public announcements about $B deals around compute for other Ai companies

Re: Gemini 3.5 Flash

#500

Post by winrid »

guelo wrote: post #392
winrid wrote: post #374

Earlier quoted context omitted.

It was probably the right call at the time with low bandwidth. Nowadays I bet flash would execute faster than most js heavy sites :D

It was not the right call, Steve Jobs was just a monopolist killing a competing platform and we're all worse off for it.

I meant that designing Flash to use more CPU to save bandwidth was the right call at the time, unless I misunderstand your reply.
Post reply on HN