aparashk

"Starting from $3K" likely means that the maxed out memory configuration will cost more than $3K.

I can't see how this will work in terms of a TDP. 2/3 of the 4090 power would be several times more power than can be effectively cooled in the physical form factor of an Apple Mini. Either there is severe downclocking happening under full throttle, or NVIDIA have come up with more low power design mojo than Apple has been able to muster for the M4 Max.

Delta strain reproduces at ~1,200x faster rate than the already contagious original COVID19 strain. This is an absolutely massive difference and works on quasi-exponential scale! The body cannot detect and react immediately in force to a virii intrusion, hence a vaccinated person can be infectious, but with everything else being equal, he or she will have lower viral load (hence level of infectiousness) and shorter period of infectiousness compared to an unvaccinated person.

Why is that not clear to most people opposing vaccination on these grounds, is beyond me.

In a mine -- constricted space of very small volume with forced ventilation -- the conditions would be ideal for getting infected with a very high initial viral load. Under these conditions, it is not unreasonable that immunity of any kind (from infection or vaccination) would not be sufficient to prevent symptomatic COVID19 infections.

Did not know this, thanks for the info, makes me even happier with my TIDAL subscription!

I have Plex lifetime sub and TIDAL sub, but never explored them in combination. Where would I start to learn more? Have you compared to Roon's or Audirvana's integration of TIDAL?

There is some subjectivity to it, but yes, I agree from experience. Have both Apple Music, for the kids, and TIDAL for myself, BTW.

Master quality is king these days, when "transparent" compression codecs are used. However, there are still many people who believe they can hear/feel the difference. You will need a combination of a golden ear and extra good equipment, starting from about 1K and above for a headphone setup.

My experience has been that TIDAL CD quality streams are of great quality, and some of the MQA ones are brilliant. Of course, a lot depends on the master quality, that seems to be the key these days as a lot of the sound compressionist's work has been automated by tools.

I fear there is a consolidation coming up for the music streaming services. Smaller companies like TIDAL will find it difficult to compete with Apple, Spotify, Amazon and Google.

As an avid TIDAL user I wish them well in their new "home".

Fascinating, so you are saying if a Rosetta2 like software can be made for Graviton or X1, that’d be an instant hit?

What is the x64 on ARM opportunity in a bit more detail? What would drive a billion dollar valuation for such a business?

I would have thought that anything of lasting interest would be available in source code and thus ported to run natively on ARM...

TLDR

iMac Pro vs Apple M1 Macbook Pro FCPX summary:

Exporting H.264 Sony 10 bit 422 footage with just one rec709 lut: 11"30' vs 10"20'

30 seconds long H.265 canon 10-bit 422 footage at 100fps: 1"20' vs 45'

10bit 422 canon h265 8k ipb footage for 30 seconds with one rec709lut: 2"47' vs 1"02'

You are correct, static binary translation is what Rosetta does first. That, however, is what I called niche technology in another post, most of the research so far had focused on dynamic binary translation.

Furthermore, SBT, even for user mode binaries, can rarely reach the performance levels that we see with Rosetta2. There are many issues in determining what is code, where are the branch destinations in case of indirect branches, etc. What we have here is certainly a feat of engineering on its own.

"If an app marks a page executable at runtime (such as a JIT) it will interpret that."

Was that from a presentation on Rosetta2 that I have missed? It certainly makes sense, but also you'd need to watch for writes to executable pages that have been already AOT translated.

Understood, I called this "ahead of time binary translation" in my post above, and I know it would use JIT-based emulation when necessary.

Do you have a reference for this? If true, this could have some interesting implications...

It is more than likely that Geekbench will use CPUID to establish what it is running on, and then adjust its kernels to suit the host CPU architecture. Rosetta2 would be using a CPUID that excludes the AVX and any other Intel ISA extensions that it does not support, therefore, Geekbench should be running without issues on the fastest vector extension supported, most likely SSE3 or SSE4.x.

I am similarly curious about Rosetta2, but there seems to be very little marketing let alone technical information being made available. All I can figure is that it performs user mode emulation, similar to what QEmu can do, and does not cover some of the newer instructions.

Off topic, my job has been in virtualisation for the last 12 years, thus I am very familiar with the publicly available body of research on this topic. Ahead of time binary translation has been a niche area at best.

Killing someone under the influence is not an act which would attract lenience from the law in any developed country. The media, especially in Australia, has not been great at reflecting the incident in all of its subtlety.

This is what I did, keyboard is fine, I like the short travel. Maxed out CPU and memory and saved nearly 40% while buying brand new old stock. Very happy overall upgrading from a maxed out late 2013 Macbook Pro 15”.

Which patent is that? Genuinely curious.

This appears to be a super-optimizer (https://en.wikipedia.org/wiki/Superoptimization) implementation, one of the smaller ones I have seen. Very cool!