sliken

16 ms·

181.
by sliken 2y ago
Why? Not like there's a single memory channel. Keeping the memory controller busy with tons of pending requests is a great way to make use of a large fraction of the total memory bandwidth. The M2 Ultra has 32 or 64 memory channels,
182.
by sliken 2y ago
Indeed, much like the pending AMD Strix Halo and already chipping AMD MI300A. Much like dual socket servers, where each socket can address all memory, these new servers have two memory systems, one optimized for CPU and another optimized fo
183.
by sliken 2y ago
Not sure why that would be true. Slow UMA (like the vast majority of Intel and AMD desktop chips with 128 bit wide memory) hurts GPU performance. However the M4 Pro has 256 bits wide, M4 max 512 bits wide, and M2 Ultra has 1024 bits wide.
184.
by sliken 2y ago
Not to mention nearly every Intel and AMD desktop that has a small iGPU on board, even for the chips not marketed that way.
185.
by sliken 2y ago
Indeed. Very similar to the grace + hopper, grace + dual hopper, and grace + dual blackwell.
186.
by sliken 2y ago
Maybe, but it's pretty lame when you buy a new $500 CPU (7800 XT) and the docs say it's unsupported, and makes you feel like even if you reported a bug they would just say "Sorry, not supported". Did make me wish I bough
187.
by sliken 2y ago
PCs are all about customization/flexibility, control, performance, and value (perf/$). On your watch, phone, or Mac desktop you generally have no choice on OS, not much control on ram, storage, GPU performance, etc. You can'
188.
by sliken 2y ago
Very interesting. I'll keep an eye out. This is one I've been tracking as well: https://news.ycombinator.com/item?id=40744293
189.
by sliken 2y ago
Paywalled. Neil deGrasse Tyson claimed that several telescopes contributed to the dataset of orbits that lead to the conclusion of an 9th planet is impacting orbits. He also claimed there was a misalignment of one of the telescopes and that
190.
by sliken 2y ago
I've been looking for a replacement for Google Assistant/Nest/Nest Mini that is cloudless and could support things like verbally setting alarms, inquiring about the weather, and random queries. This looks like a decent fit.
191.
by sliken 2y ago
I tinkered with Wasabi, but their minimum bucket lifetime made it not work out for me.
192.
by sliken 2y ago
QNAP has the QSW-M2116P-2T2S-US. Has 16 ports of 2.5G, 2x10G (copper), and 2x10G SFP+. The 10G ports support 10G/5G/2.5G unlike the older 10G chipsets. Also has 280 watts of PoE. Goes for $700 or so.
193.
by sliken 2y ago
That's my preference as well. Keep the router simple, don't put a transmitter near the sensitive bits to cause interference with the CPU, ethernet, or USB bits. I want my AP to be dumb and do nothing besides connect ethernet to
194.
by sliken 2y ago
Seems kinda weird, why one GigE and one 2.5GbE? Maybe they will make a one plus with double the ram and 2x2.5Gbe?
195.
by sliken 2y ago
Sad that they don't support any of their more popular cards, like say the Radeon 7800xt, which is their current generation card. The 7800xt has 16GB vram and looks pretty good at $400 these days ($500 at launch).
196.
by sliken 2y ago
Well there's a fair bit of chemistry related to the explosions to bring the sub-critical bits together. Time scales are in the nanosecond range. Then as the subcritical bits get closer obviously the nuclear effects start to dominate.
197.
by sliken 2y ago
Another fun fact, the priority for detecting nuclear testing led to seisometers all over the planet. So the detection the exact position and nature of any disturbance on the planet became radically better. This was quite the boon to anyon
198.
by sliken 2y ago
I've seen nothing showing that slingshot has any particular advantage over IB for HPC. Sure HPE pushes slingshot (an HPE interconnect) over giving bags of money to Nvidia, but that's a business decisions. Eagle (the #4 cluster o
199.
by sliken 2y ago
I'd expect linpack to be much closer to a user research application than training LLMs. My understanding of LLMs is that it's more about throughput and has a very predictable communication patterns, not latency sensitive, and ban
200.
by sliken 2y ago
Not a secret, from IEEE: The NNSA—which oversees Lawrence Livermore as well as Los Alamos National Laboratory and Sandia National Laboratories—plans to use El Capitan to “model and predict nuclear weapon performance, aging effects, and safe
201.
by sliken 2y ago
Microsoft has the #4 cluster on the top 500 list. Sure not everyone reports, still seems like a useful list to watch the trends in computing and in particular HPC. Keep in mind the average hyperscalers cloud is not a particularly good setu
202.
by sliken 2y ago
Nvidia GPUs, that support ECC, do this. I believe it's called inline ECC and does cost latency, bandwidth, and memory capacity. This helps, but ideally the entire path from CPU to Dimms is wider and covers not just what is being read
203.
by sliken 2y ago
Worked yesterday.
204.
by sliken 2y ago
If you just did a FFT on microphone and acceleration for every link in a subway system and recorded the past travel times. Seems pretty straight forward to estimate your speed vs the historical records and then predict your arrival time. S
205.
by sliken 2y ago
I have pondered adding 4 microphones to the corners of my house. I suspect with some processing and filtering you could likely tell quite a bit about the house. When the front door or garage opens, what rooms are occupied, when it's
206.
by sliken 2y ago
I was pleasantly surprised to note my pixel 9 has a barometer, 81,500 or so right now, I'm at 6,000 feet of elevation.
207.
by sliken 2y ago
Sure, assuming you are not in a building, balloon, helicopter, plane, drone, etc. Not to mention interesting topology like a cliff where an error in long/last of even a single foot could mean your altitude would vary by 1000 feet.
208.
by sliken 2y ago
That would be selfish and terrible. The most productive farming in the USA is largely because of previous volcanic eruptions. Similarly the water cycle, subduction, earthquakes, floods, fires, and carbon cycle are what keeps soil healthy
209.
by sliken 2y ago
I believe they are using m2 ultra studios to provide cloud based AI services.
210.
by sliken 2y ago
Ah, interesting and disappointing. I've been looking for more memory bandwidth. The M4 max is tempting, even if only half the bandwidth is available to the CPUs. I was also looking at the low end epyc, like the Epyc Turin 9115 (12 c

More ›