Binkg
@dummerspast39
anti ai lobby
china shill
big hater
276 Following 2.8K Followers
Gemini 4 Carbon - SVG of Cola (I asked for svg yes)
This is actually a great output when taking into account it was in the Arena. For like 5 minutes, shit harness. Even better than Opus
Show more
This is the most misleading chart all year:
-GB300 ultimate rack vs DeepSeek on GB200
-peak prices vs off-peak prices, so half price for DeepSeek
-the tweet calls it 50$ profit but their own chart says 51$ is the revenue
Same prices and DeepSeek makes around 78$ per GPU hour on an older chip. Thats 1.5x the best vLLM setup on the GB300 and they framed it the other way round
Now same hardware, same off-peak price, revenue vs revenue with their own numbers:
-vLLM on GB200: 3.58$ a hour
-SGLang on GB200: 5.38$
-DeepSeek on GB200: 39$
Thats 10.9x vLLM. GB200 rent in their chart is 4$ so vLLM loses 0.42$ every hour at DeepSeeks real price
Show more
@teortaxesTex Oh yeah forgot. Here is it again
H100 equivalent compute by lab - up to current day
Hey Lisan, my dear old @scaling01, could I get a "thank you" for being your MOST VIEWED post in the last 2 months?
Right now I am blocked and uncredited. Must be a mistake. For sure🤷♀️
As you live of X (this is his full time job), you must be enrolled in the ORIGINAL CONTENT REWARD PROGRAM. According to the terms: Original content ≠ making a screenshot and adding 5 meaningless words.
As you deny any wrongdoing I will pursue (already have lol) every action X/support offer me.
Show more
Fable 5.5: SVG of a controller riding a bicyle riding a pelican riding a bicycle riding a pelican
Muse Wonder has that BIG model smell
Like bro I feel sorry for him. Why does he think this is good😭😭😭😭
@teortaxesTex Semianalysis said 92% API margin, and about -360% in subscription when maxing out the whole month on opus 5.5
devday finally killed openai. very happy. this dots stuff seems like codex but inefficient and trash. we got some random other stuff nobody will use like space or whatever. sol 6.1 is mid. lets hope it has a better base than 6 sol. btw it doesnt roll out in chatgpt. >>opus 5.5. and then astra ultrafast. did anybody ask for this? probably they hope some enterprise customer forgets that they selected it. and lets dont forget about the 500$ plan. call me stupid but i would have started by fixing the plus plan. wow this is the worst devday ever. respect to the team. i had some (delusional) wishes: gpt 6o, astra 6,1, 2x limits. openai bros are fucked.
Show more
5.6 sol lost 84% of its value lol
