zimmie

longpath said:

I'm curious how robust such a method is against spoofing/hacking versus physical keys with onboard RFIDs for secondary authentication (example: the now discontinued Mazda RX-8 has a conventional physical key which also carries an embedded RFID which the PCM must authenticate in order to enable to fuel pump).

The big difference is normal RFID tags like the ones used in old car keys for immobilizers just respond with a static number. Their security is purely from the low range from which they can be read. Implantable chips for pet identification are actually very similar. They just contain an ID number, which you then look up in a database to find the owner's contact information.

Remotes for power locks brought this ahead a bit by using a rolling code, but it is still extremely simplistic. Early passive keyless entry systems (such as the one in the C4 Corvette) used the same rolling code system. They were effectively remotes which pressed their own unlock button as you walked around. This also resulted in the remote's battery only lasting a few months at a time.

Newer passive keyless entry systems are significantly improved over that. Instead of the remote detecting you are moving and sending a code, the car (with its much bigger battery) broadcasts a periodic signal to ask if the remote is near it. The signal contains a challenge. When the remote picks it up, it takes the challenge, crafts a response which only that remote could send, then broadcasts it back to the car. If the car sees a response from a remote it trusts, it unlocks.

This has led to a relatively new class of burglary involving a relay system. The attackers bring a radio near the car which receives the signal, then sends it using a much stronger, more directional radio towards where they think the key might be. If the key receives it, it sends its response, and the burglars' radio picks it up, then hands it back to the car. This allows them to unlock the car and take things from inside it, but not necessarily to steal the car itself.

This system is more like the challenge-response system used in modern passive keyless entry. The car will look for phones and send a challenge. If the phone has the proper token, it will be able to generate an acceptable response and the car will unlock. The big question I have is whether this will require cell connectivity. I'm not interested in paying for a cell subscription for my car. It has no business talking to any external computer.

larryjw said:

When first installed, the ChatGPT app highlights the warning that it may return false information, 
The May 10 Lancet published the following article, along with detailed supplemental addendum showing the interaction with ChatGPT which anyone can then reproduce to verify the results. 
The dangers of using large language models for peer review

ChatGPT responded with totally made up material, sounding quite authoritative. 

Banning ChatGPT would seem a good idea. It would give alternative AI systems which actually can tell the truth an opportunity to be developed -- if that can be done. 

Large Language Models like GPT-4 are inherently lying machines. 

GPT isn't strictly lying, you just aren't asking it the question you think you're asking. People think they're asking the question they put in the prompt. What GPT actually does is produce something which looks like what an expert would say in response to the prompt. For example, experts cite stuff when they're asked about legal or medical matters, so the response should have citations. If there are no high-quality real citations to include, make some up which look plausible.

This behavior becomes much more obvious when you realize GPT (and Bard, and so on) will "believe" anything you tell it is at the other end of something which looks like a URL. The domain doesn't have to resolve, and the path doesn't have to be valid. An expert would load the page and potentially respond with new information based on what they read there. GPT can't actually load a URL or read, but it can fake a response which looks like what an expert who did all that might say.

People have been doing things like telling Bard that Google discontinued Bard months ago, and it responds with something like "Oh. I'm sorry, according to the link you provided, you're correct. Bard was discontinued in March."

Edit: Put "believe" in quotes. As Larryjw pointed out, the model doesn't actually know anything, so it doesn't actually believe anything. Ultimately my point is people think of GPT as being built to produce correct output, when it's actually built to produce plausibly formatted output.

144 antennas means 144 amplifiers, and a 144-path phase modulator (unless it clusters them into sectors such that when in one quadrant of the device, the other three quadrants' antennas are inactive) or 144 coherent oscillators. The radiation precision would also depend on the tolerances of several components in each of the amplifiers. It's basically a track-while-scan RADAR.

Based on the last time I dealt with active phased array tech, that base station is likely over $750 in parts alone.

hmlongco said:

Misses the other major benefit of SOC integrated memory. In a traditional system if I want to move an image from memory to a graphics card that image data has to be copied from the CPU's RAM to the RAM on the GPU, byte by byte. Similarly, if I want the GPU to perform some action on that image and return it then the result needs to be copied once more from the GPU back to the CPU.

In Apple's SOC design, you do little more than hand the address of the data in RAM to the GPU, which then can perform the operation in place. You get tremendous gains in throughput when you don't have to copy data back and forth.

This is the main reason why the RAM is soldered. The article's mention of interrupts and related information is technically correct, but irrelevant: Apple's design still has a memory controller and still behaves very similarly to designs with outboard memory controllers (northbridge) or integrated memory controllers (current AMD and Intel processors).

The memory controller has exclusive access to the RAM. Everything else goes through the memory controller to request the contents of an address range. Apple's data sharing is possible because the CPU cores and GPU cores (and Neural Engine cores, etc.) all go through the same memory controller. The performance figures (such as 400 GB/s of memory throughput) are because Apple uses multiple DDR5 channels. Each channel gets 50 GB/s. The base M2 has two. The M2 Pro has four. The M2 Max has eight. The M2 Ultra has 16. Each of these channels could lead to a slot if Apple wanted. Yes, there would be a barely-measurable amount of latency added by the longer traces. They would also consume a barely-measurable amount of extra power. Those are insignificant next to the two main reasons Apple doesn't offer slots, though:

1. Slots take up a lot more physical space than RAM chips soldered directly to the SoC package
2. If you give users slots, they will put RAM in them with different capacities and performance characteristics

For the first one, four memory channels on the M2 Pro would mean four DIMM/SO-DIMM slots. That's not huge, but it's also not nothing. The M2 Max is more significant. Eight DIMM/SO-DIMM slots take up a lot of space. That many sticks of RAM would more than double the size of the MacBook Pro's logic board. 16 DIMM or SO-DIMM slots would roughly double the size of the Mac Studio.

For the second point, with soldered RAM, Apple can guarantee that each channel has the same amount of RAM and each channel's RAM performs the same. This removes a HUGE amount of situational logic to deal with seven sticks at 50 GB/s and one stick at 44.8 GB/s, or five slots populated with 8 GB each and three empty slots. The system can be designed to just assume certain things because Apple can guarantee at a manufacturing level that those assumptions will never be violated.

JamesBrickley said:

The old Apple Xserves had iLO Lights-Out management.  But none of the modern Macs have the necessary BMC chips and remote management functionality of PC servers.  It would be nice to have that.

They actually do have an equivalent to a BMC, though Apple calls it an SMC. Also, all Macs with built in 10g network interfaces except the iMac Pro support lights out management over the 10g interface. This includes the Mac mini if you spend the $100 to get the 10g interface. Would be nice if they had a more standard way of interacting with it (Redfish!), but it's enough to manage the hardware after initial setup. It's also hard to mess up in a way which could leave the system open to attackers.

maltz said:

darkvader said:

maltz said:

At some point you have to ask if a row of Mac Minis are the right tools for the job...  probably a point well before this.  lol

Oh, it isn't.  But since Apple killed the Xserve, it's the only tool we've got.  VERY stupid move on Apple's part.

Honest question:  Why is ANY Mac the right tool for the job?  I'm not sure what server application is better handled by macOS than by Linux these days.

There are certain services which only run on macOS. Most notably, Xcode Server, which allows you to run build and test bots to do work submitted by developer workstations. UI testing in Xcode involves a piece of software operating the system like a blind person would. It moves the cursor around, actually clicks on things, and so on. When running a suite of UI tests, the system running the tests is unavailable for any other UI work, so people really like being able to push that work to a remote server.

Apple also provides a content caching service which runs on macOS and caches software updates, stuff from the store, and so on.

On a purely subjective note, I cannot stand Linux, and I'm of the opinion that any application would be better handled by macOS than by Linux. With the widespread adoption of systemd, pretty much every distribution is wildly unreliable. I would never run anything I care about on it. I'd go with a proper UNIX (BSD, illumos, etc.), then a UNIXy RTOS (Pike|QNX|seL4|VxWorks), then Windows then not running the service at all, then Linux if I had no other option.

Soli said:

While I think it's very likely that Samsung stole this icon from Apple as it seems to be in their nature, to me that icon also seems very obvious.

Obvious to a certain extent, perhaps. Same line weight, same corner radius, same spacing between the corners, same J nose, though? That's a bit much.

They could have included hair.
Or given their icon glasses.
Or shown a whole head, not just a floating face.
Or omitted the frame.
Or even changed the nose to point the other direction!

But they didn't.

If it's using the same GPU cores as the M1, clocked at the same speed, a 64-core GPU could do 20.8 TFLOPS, while a 128-core GPU could do 41.6 TFLOPS. For comparison, a GeForce RTX 3090 (the top consumer card from Nvidia) does up to 35.6 TFLOPS, and a Radeon RX 6900 XT (the top consumer card from AMD) does up to 23 TFLOPS.

Considering the RTX 3090 and RX 6900 XT still universally sell for more than double their MSRP, I wonder if Apple will have scalping problems. Their system of allowing backorders mitigates scalping, but doesn't eliminate it. With the added demand from blockbros, it may be difficult to get one for a year or so.

mayfly said:

"Apple said at WWDC that the battery is not casually removable from the headset. There is a USB-C port on the battery for charging and directly powering the Apple Vision Pro."

Just my uninformed opinion, but I think Apple missed an opportunity by attaching the battery this way. Better to put two USB-C ports on the headset, so users could just swap battery packs without it powering down. Weight gain would be marginal, and it would be easy to add a latch to prevent accidental detachment. Unless it's possible to attach a backup battery to the USB port on the battery. At this time, there's been no mention of that from Apple.

USB power delivery is extremely janky. It isn’t suitable for anything without a built-in battery. Devices can abruptly lose their internal state for any number of reasons. The only possible safe response to this failure is to drop to the lowest supported voltage and current, which definitely isn’t enough to keep the Vision Pro running. It also flatly does not support seamless handoff between two power sources.

They could have done it with a sizable internal battery and USB power delivery, but then all this upcoming legislation mandating easily replaceable batteries would bite them. They would probably also need to change away from glass for the outer face to hit their target weight. Apple uses plastics for AirPods, the Magic Mouse, and all of their keycaps, but they haven’t done plastic in front of a display in a long time.

hmlongco said:

newisneverenough said:

Probably this is common knowledge, but why has Apple made so much Not user upgradable? Is it to sell more new machines ? Is it to eliminate technical problems when users upgrade? Conceptually, I want more control of the thing I pay thousands for. 

One thing is that all of the RAM in onboard the SOC and shared between the CPU, GPU, and various other systems like the NE. THat's where a lot of the performance comes from in that memory is on the chip and it doesn't have to go to an external bus to get it. Further, moving something like a shader or bitmap from the CPU to the GPU is basically just passing a pointer, as opposed to copying a bunch of bytes back and forth.

Adding external RAM kills off aspect of the performance.

This is incorrect. Apple could get the same capabilities with slotted RAM. They're just using multiple DDR channels. The base M2 uses two DDR5 channels at 50 GB/s each, M2 Pro uses four channels, M2 Mac uses eight channels, and M2 Ultra uses 16 channels. For the laptops, iMac, Mac mini, and Mac Studio, the reason for soldered RAM is space. You need a separate slot per channel. Reserving space for eight SO-DIMM slots for the MacBook Pro would be impractical. Eight or 16 SO-DIMM (let alone full-size DIMM) slots on the Mac Studio would be hilarious to see, but also impractical. It would probably double the volume.

For the Mac Pro, there's no technical reason they could not have used slotted RAM. It definitely has the space and cooling capacity to spare. They simply chose not to.

newisneverenough said:

Probably this is common knowledge, but why has Apple made so much Not user upgradable? Is it to sell more new machines ? Is it to eliminate technical problems when users upgrade? Conceptually, I want more control of the thing I pay thousands for. 

The SSD in the Mac Studio actually is upgradable. Apple released the repair manual last December, and it includes details about the flash carts which make up the SSD's storage. 512 GB and 1 TB storage options use a single cart (which limits performance), while the higher capacity options use two carts. The upgrade process involves removing all installed flash carts, installing a new set, then using Apple Configurator on a separate Mac to restore the SSD controller.

There are a few reasons for integrating the SSD controller and only swapping the flash the controller uses. One big one is that it allows Apple to guarantee the security characteristics of the SSD controller. Several SSD vendors have been caught lying to computers about their encryption capabilities. The machine tells the drive to encrypt itself with a particular key, the drive says "On it!", then the drive just ignores the key and leaves the flash encrypted with the default key (typically all-zeros).

It also allows Apple to target their own performance goals rather than simply accepting other vendors' market segmentation strategies (Oh, you want a lot of random IOPS? You'll need to go with an "enterprise" drive at 3x the price and 8x the power draw.)

The RAM isn't upgradable for space reasons, as I explained above.

DT36MT said:

amar99 said:

How will he survive on such a low salary? What a guy.

Will you do it if you were in his place just because someone or some group of people say you should?

If I were in such a position, I would have my salary lowered without somebody else pushing for it because I can't imagine ever needing more than about $100M in assets. That level of wealth is self-sustaining for even an unreasonably extravagant lifestyle. It's enough for separate his and hers Ferraris for every day of the month with plenty left over to still get about two million dollars of income per year on bank interest. What would you ever spend it on?