I'm sorry but this super micro thing is awful but parts of it are genuinely hilarious
They literally used a hair dryer to move serial numbers from real servers to dummy servers to throw in a warehouse and got caught on camera
English
Max Weinbach
101.2K posts
Max Weinbach
@mweinbach
Analyst @creativestrat | Analyst and Market Research Firm | Typo ignorer Email: [email protected]
Connecticut, USA 가입일 Haziran 2015
7.3K 팔로잉286.8K 팔로워
@mweinbach @theo and the actual cost is like 2 / 1.5 = 1.33x?
English
Can someone smarter than me explain how these “turbo mode” things work on like an implementation level?
Are they using better GPUs? More GPUs? Reserved allocation during generation?
English
@mweinbach @theo thats not quite the math, its more about the ratio between FLOPS and memory bandwidth
English
@theo i mean if you just use ACP, is that a problem?
are you doing more than that?
English
Is now a bad time to drop our Claude Code integration? 🙃🙃
dax@thdxr
opencode 1.3.0 will no longer autoload the claude max plugin we did our best to convince anthropic to support developer choice but they sent lawyers it's your right to access services however you wish but it is also their right to block whoever they want we can't maintain an official plugin so it's been removed from github and marked deprecated on npm appreciate our partners at openai, github and gitlab who are going the other direction and supporting developer freedom
English
@theo then you have cerebras which is a different model using prefill on GPU and decode on cerebras, GPT 5.3 Spark is like this
but it's mostly the batch size above
English
@theo or if you look at anthropic, it's likely using more/different silicon
English
I think my problem with the Model Y Performance is that it rides so well and it accelerates so fast that sometimes I don’t realize that I’m going 20-30 mph over the speed limit on highways and long open roads
English
Not shocking, makes a ton of sense
First Squawk@FirstSquawk
OPENAI TO MERGE CHATGPT, CODEX APP & BROWSER INTO DESKTOP SUPERAPP TO STREAMLINE RESOURCES & USER EXPERIENCE - WSJ
English
@iHateApplee No it goes to a better chip, display, battery, software, support, cameras, connectivity, storage, features, etc
English
@mweinbach using price as a proxy for "better" is exactly how companies like Apple get away with selling 8GB of RAM in 2026.
If a $500 phone is "really close" to your $2000 flagship, you’re basically admitting that the extra $1500 is just a "logo tax" and a fancy titanium frame.
English
Nothing is done an incredible job making a $500 phone feel very premium and look really nice
This is probably the best feeling cheap phone I've used and feels more premium and faster than a lot of $1,000 phones
It's a good product, which is helpful to have in a weird market
English
Depends on the model, if you're using Claude or something you can burn that in 6 months
If you're using an open model (which you would run locally) it may take a couple years vs. cloud host
If you can license something like Gemini from Google to run locally (which they offer)...
English
@mweinbach Will that really hold true for the entire useful lifetime of the hardware, though?
It might take a long time to burn through $100,000 in tokens.
And during that time, the models should be getting both better and cheaper, so that money will keep stretching further and further.
English
The obvious question is how many engineers can use a local model running on a Dell Pro Max
I think we're looking at one GB300 per engineer, and depending on the engineer, this can be cheaper than buying API tokens even at over $100K per unit!
Ben Bajarin@BenBajarin
We think these token generators that can live on-prem and serve tokens to teams are a new factor to consider in modeling where all this hardware goes and how agentic plays at the edge. Good example from @dell here with Pro Max GB300. thediligencestack.com/p/nvidia-is-ex…
English
@mweinbach and your trackpad will inexplicably fail at just the wrong moment!
English
I really love this ThinkPad T1g Gen 8 but the only part that kills me is the display
I wish I had the tandem OLED version, a 60hz LCD is just not good enough for me. A 120hz LCD would be fine, but 60hz in a $4600 laptop!
English
@KristinaParts I'm the reason why the rule 'no transcripts' exist. 🙃 It's also partly why I don't get invited anymore, I assume (in part). I broke an unwritten rule.
English
@alexkehr @anndrrson Yup. I’ve never been sold on this. I’ve heard the contrasting opinion of in tough times people want special experiences however they can access those, which is metaverse, but that makes it less special. Might as well watch on a TV or your phone
Gonna want to be out and social
English
@mweinbach @anndrrson AI will make people crave more of the exact opposite of the metaverse
English
Wait
Did meta really bring back metaverse after a single fan asked them to? That's crazy
English
Haven't they been talking about this for like a year on this app?
I'm pretty sure the product leads directly have said they're working on this.
Is this supposed to be news?
But also Google can I pls try...
First Squawk@FirstSquawk
GOOGLE HAS BEGUN TESTING A DEDICATED GEMINI APP FOR MAC TO COMPETE WITH CHATGPT AND CLAUDE, OFFERING FEATURES LIKE CONTENT GENERATION, WEB SEARCH, AND PERSONALIZATION.
English