xyx0826

Neat. I’ve been watching a lot of Nvidia board level repair videos; the technician would occasionally probe the SPI signal lines to the flash chip containing “VBIOS” code to see if the GPU shows signs of life. Given no external microcontrollers on the board, I’ve always wondered what is the type of core that runs mundane tasks like executing the boot ROM, controlling PWM output for the fans, setting RGB LED colors etc. My guess was on Cortex-M, but I guess I might be wrong.

I think Nvidia made a good choice to concentrate logic into the main die, despite a CPU usually having a separate EC (Embedded Controller) or Super IO chip to deal with bootstrapping and fluff. It likely helped board vendors avoid having to go through 50 STM32 copycat product lines during COVID, even though this means the GPU die’s signals are somewhat exposed in practice and “power users” keep blowing up their cards when trying to do a cooling mod.

I remember reading an analysis on one of those projectors; the author found a residential proxy running on their device. I would recommend keeping these things off the internet.

Root persistence is more for encouraging you to report all bugs you’ve found instead of holding on to them in fear that the vendor will patch them and make further research difficult. See Apple’s security research device program.

Notably this bill covers Class 2 and 3 ebikes, not Class 1. The key distinction here is Class 2/3 features a throttle that lets you accelerate without having to pedal.

IMO Class 2/3 bikes are more of a bike/moped hybrid than just bikes.

You seemed to have made an account just for this reply. Care to explain the cynicism? I’m out of the loop with the toxicity.

Is it necessary to prevent the water from freezing? If the chamber and water within are subzero while the SoC produces sufficient heat, the ice would simply melt.

* Edit: the article mentioned freezing could crack the seal. Freezing would be a bigger issue than I had thought, then.

I applaud his efforts to document this what must’ve been a nightmare of a case for him. But it felt like a lot of the wording is speculative or hyperbolic in nature and aggressively tries to paint Volvo in a bad light. For example:

“Analysis of Volvo's Final Response: This response … confirms Volvo's complete abandonment of customer responsibility…This is Volvo's definition of ‘customer care’ in 2025.”

“Center Display Failure - Critical Interface Blackout: Main Controls Inaccessible”

“Climate Control Malfunction - Climate System Override: Controls Unresponsive Despite Interface Status”

“Complete Center Screen Malfunction - Total System Breakdown: Hard Reset Failed to Restore Screen”

I know little about Volvo or this case; I’m choosing to offer them some benefits of doubt. Comms and decision making are prone to break down on the corporate ladder. Volvo had no doubt fumbled his case badly but I’m not convinced it is indicative of the company’s overall customer support policy. Sure, the main touchscreen had failed. But how is this an “override” of HVAC or a “total system breakdown”? And what’s the “system” anyways? On top of all that, these subtitle summaries smell like AI.

I don’t deny that Volvo has a lot to answer for. Though the choice of these instigating descriptions might not be the best one giving the author is actively pursuing litigation.

Before I came across this tool, one use case I’d conceived for a tool like this is to help in debugging and reverse engineering. One could hook this up to a microcontroller under test/analysis and essentially monitor its “disk I/O” or supply it with arbitrary data.

Files looks great but it has performance issues and occasional crashes when I tried it out a few months ago. When going into subfolders, there is a very noticeable subsecond lag which I don’t get from native Explorer. For all complaints of lack of features that Windows File Explorer gets, it’s still a very respectable native GUI app for being Windows’ most used program!

> you have a bunch of very smart young people absolutely dripping with theory knowledge and close to zero relevant real world experience

For sure! And that’s perhaps the #1 reason these teams are so valuable: it’s an environment to get hands dirty in. If something sticks, that’s great and goes on the resume. If something awful happens, just walk away with a cool story assuming you didn’t blow up a school building or anything like that. Either way the experience and hopefully learnings stick with these young people like me for a long time.

I think college student teams strike a combo of time, talent and resource that would be surprisingly hard to come by in the larger “civilian world.” In college, you have a bunch of freshly educated, similarly minded people in one place with a whole bunch of free time to put towards one project, highly motivated because it’s both an extracurricular escape and a career prep achievement. And these teams are often financially supported by their school departments or fundraisers. If you fail, there are little if any consequences on your life. All these motivators improve the likelihood of making something truly impressive.

Sure, we can make an arrangement like this out of college. Call up your ex-rocket club teammates, who have all now graduated and making banks at rocket startups. Spend the Thanksgiving week grinding out the CAD, code and circuit boards then test everything out in a desert. But projects like this are a huge time investment and with work and family in the way, they can often be very difficult to coordinate and pull off.

Even if your rocket does end up shooting off and breaking a record, does it truly “beat them”? I find it a bit hard to compare a team of similarly educated college students to a group of adults, usually with relevant professional backgrounds. Maybe the closest we can get are YouTuber collabs. Sometimes I miss my days spent on my college team; it’s pretty hard for me to get an exciting, rewarding, comradely and occasionally traumatizing experience like that ever again.