I've been working on a micro Linux distro for a couple of months now. User mode is just a single static binary (and a few files to support confidential microVM containers). The initramfs is such an interesting feature, the kernel knows how to unpack cpio archives and just magically drops you into a tmpfs and executes /init. You can have multiple concatenated cpio archives, and each can be compressed, and they all just overlay in sequence. There's an elegance to the design. I also had to write some code to unpack them and learned more than I wanted to.
Compare the size of the new air-cooled 40XX cards and the iChill 4090, which is tiny by comparison. The simple answer is just to use liquid cooling if you have a card using 400w. Then all the absurdity just goes away.
ProxyJump was added in OpenSSH 7.3 so systems from a few years ago might not support it. It does the same thing as ProxyCommand with -W %h:%p but you can’t set custom options for the jump connection with ProxyJump.
Yes, the kernel will map virtual to physical however it sees fit, no guarantees there (at least not in user-space via glibc). Realloc will always return a virtually contiguous slice or NULL if it can’t. And you’re right, obviously all those potential subordinate mallocs would be inefficient. :) In this contrived FizzBuzz example case, you could pretty easily do the math and just malloc it all in one go at the start of the function. Fizz and Buzz are the same size, you would just need to add up the iota lengths and the trailing \0. If you take an arg for N (1..N) then stack allocations are not going to work. You need to statically declare their size.
You can use realloc. It will try to use the existing contiguous span of memory, but if it cannot, it will malloc a new span and copy the previous bytes in and return a pointer to it.
Someone has to make a business case to get the chargers on the street. There are questions like who owns the sidewalk, how to route sufficient power there, capital outlay for digging up the street and getting the charger there. All those costs and up and the revenue for a given charger is capped. People might pay a premium over the the cost of electricity, but how much? And they have to compete with free charging in many cases.
The throughput is amazing. The screenshot shows 28,375 MB/s — the capital B typically indicates Bytes whereas a lowercase b indicates bits, so your Gb/s translation is a little confusing. The latency is still much higher than main memory or Optane, but that might be fine for many applications.
This is from mid 2017. "First notified Tesla on 31/08/2017 ..
Tesla vehicles produced from June onwards use a new key fob". Not sure if the new fob is significantly better as the presentation is not clear there.
..by the Republican majority House of Representatives. It can be roughly equated to being charged with a crime. The Senate held a trial and voted for acquittal and therefore no action was taken. Two thirds of the Senate must vote to remove a President from office.
This is promising, but for dedicated high-performance I/O I am much more interested in how the SPDK is progressing. Avoiding syscalls is great, but without any benchmarks I wonder about the benefits over mmap and vectorizing. You still have to contend with locking, interrupts, copying, and ring 0 abstraction impedence.
Curious to know if they block Google’s DNS servers as well. That 1.x space was a RIPE research segment, so it’s possible that some internal AT&T group was using it with the assumption it would not be publicly routable and got bit. I was enjoying the shorthand ping of 1.1 for my router at home until Cloudflare took it over. Needless to say, if that was the case for AT&T, their ‘fix’ is not at all acceptable.