topspin

24 ms·

331.
by topspin 1y ago
I haven't had to feed a building full of GPUs, but I do wonder: does Ceph ever appear on your radar for this sort of work? And if not, why? Seems like it would fit rather well. Yet it never seems to get mentioned.
332.
by topspin 1y ago
Yeah, it's not really about jamming more DIMMs into more sockets.
333.
by topspin 1y ago
> Why has it taken so long to get the Windows ecosystem fully on ARM? It is possible to ignore Windows on ARM and remain viable. That wasn't feasible with Apple. If Microsoft could decide tomorrow that the next Windows would be AR
334.
by topspin 1y ago
> It is non-trivial additional work to add since these are high frequency signals. And there it is. Yes, PCI-E 3.0 from 2010, 15 years ago, involves 4 GHz wire level signals. A 4x PCI-E connector has four differential pairs of this, no
335.
by topspin 1y ago
Ask yourself this: what is the correct markup for delivering this nearly four years before everyone else? Because that's what Apple did, and why customers have been eagerly paying the cost. Let us all know when you've computed
336.
by topspin 1y ago
> Is the problem truly down to physics Yes. The "conservative attitudes" of JEDEC et al. are a consequence of physics and the capabilities of every party involved in dealing with it, from the RAM chip fabricators and PCB manuf
337.
by topspin 1y ago
Maybe, but due to the physics of signal integrity, socketed RAM will always be slower than RAM integrated onto the same PCB as whatever processing element is using it, so by the time CAMM / LPCAMM catches up, some newer integrated RAM
338.
by topspin 1y ago
> Perhaps it is not possible to simulate higher-level intelligence using a stochastic model for predicting text. I think you're on to it. Performance is clustering because a plateau is emerging. Hyper-dimensional search engines ar
339.
by topspin 1y ago
Guess we should count our blessings that 7-bit bytes didn't become the de facto standard. Given that 7 bits is sufficient for ASCII and BCD, and the popularity of the IBM 1401 in the 1960's, that's not at all implausible. T
340.
by topspin 1y ago
> Why are people still installing security cameras that are monitored by them? The point of Frigate et. al. is to not have to do the monitoring. The false positives of small wildlife, known persons/vehicles, etc. do not consume att
341.
by topspin 1y ago
Also, whataboutery.
342.
by topspin 1y ago
I'm not an artist, so perhaps perspective is the wrong term. Depth could be what I have in mind. In the first drawing, on the left, there are parts of the two cat's both above and below parts of the stag. The tail on the cat on
343.
by topspin 1y ago
Those drawings show better expression of perspective, motion and proportion than what one sees in medieval drawing. And this is on skin, around limbs, as opposed to parchment. Genuine art.
344.
by topspin 1y ago
Yes. The same insight periodically appears. Also, often it's done with high contrast text, which is great.
345.
by topspin 1y ago
This might be what you have in mind, or similar[1]: RISC-V Taipei Day May 21, 2025. There are some hard numbers on Ascalon performance in there, anyhow. One figure from that: "36 SPECINT2K17 Rate" for the "8 core Athena Chip
346.
by topspin 1y ago
> You'll be able to buy RISC-V at that level sometime next year. Will you? What exactly are Tenstorrent's plans for Ascalon? They are chasing the AI market, and I haven't seen any clear indication that they intend to sel
347.
by topspin 1y ago
Standard low rate production effects. Equipment like fire trucks are full of weird, low volume things made by a few, or only one shop somewhere, built exclusively to order, with long lead times. A fire truck is simultaneously a mobile pow
348.
by topspin 1y ago
"a single, larger layer of hidden units" Does this not mean that the entire model must cycle to operate any given part? Division into concurrent "modules" (the term appearing in this paper,) affords optimizing frequency
349.
by topspin 1y ago
> "After completing the T steps, the H-module incorporates the sub-computation’s outcome (the final state L) and performs its own update. This H update establishes a fresh context for the L-module, essentially “restarting” its compu
350.
by topspin 1y ago
The Springfield XD line has this. At least one S&W Shield model as well. Both are excellent pistols. Not that you're wrong: grip safeties aren't common. It's amazing to me this P320 problem exists. Making semi-autos s
351.
by topspin 1y ago
> they're using CO2 as a working fluid is an attempt to sound carbon-sequestration-adjacent Um no, that's unfair. CO2 is an easy engineering choice here. It's easy to compress and decompress, easy to contain, non-flamabl
352.
by topspin 1y ago
The core idea: "Instead of instructions flowing through a centralized pipeline, the E1 pins instructions to specific compute nodes called tiles and then lets the data flow between them. A node, such as a multiply, processes its operand
353.
by topspin 1y ago
> At high impulses deformations and elasticity throw the kinematics Sure. Yet evolution has achieved astonishing kinematics with all manner of deformation and elasticity inherent to the materials, and also constantly changing physical p
354.
by topspin 1y ago
"It added an additional three-kilometer-long ring between the main interferometer and the detector to circulate the light before it exited the interferometer’s arms." Isn't that a delay line? The benefit being that when the
355.
by topspin 1y ago
The big players in the Wi-Fi SOC market are all using ARM in recent designs, Intel and its multi-core Atom being an exception. RISC-V isn't supplanting these at the moment. The chipsets in this market are highly specialized: most of
356.
by topspin 1y ago
Replying to myself. io_uring is only a more ambitious manifestation of a pattern that has been commonplace for a long time. High performance network, storage and other hardware use circular buffers internally in their drivers to process co
357.
by topspin 1y ago
I can't speak to the degree to which Raku VM ideas are being backported. I'm not that deep into Perl. I just know that the Perl 5 language itself is still regularly enhanced, and some amount (can't give you a percentage or
358.
by topspin 1y ago
> and perl hasn't evolved much since then Perl (5) has been evolving, quietly. The language has had many nice additions over the years. Some of it is backporting from 6.
359.
by topspin 1y ago
It's tragic, and it bothers me that Jens Axboe's work is suffering due to this. Obviously, with the clarity of hindsight, this might have been avoided. Now the cost of the damage it high. My idea: consider formal verification.
360.
by topspin 1y ago
As part of that, there is a strong trend toward consolidation, and little competition emerging to challenge the oligopolies that have formed.

More ›