ribit

29 ms·

271.
by ribit 5y ago
If I am not mistaken the ports on recent Apple machines are on a separate, easily replaceable board.
272.
by ribit 5y ago
That's the thing, who is everyone? Such technology would only be relevant to premium-level ultraportable notebooks. Desktops, gaming laptops as well as budget laptops would stick with the cheaper and more ubiquitous DDR. Smartphones an
273.
by ribit 5y ago
Once you get to a certain stage, yes, modularity becomes very expensive. I do believe that things like SSDs or WiFi cards should be modular, because there is no reason for them not to, but RAM? We are moving towards high-bandwidth on-packag
274.
by ribit 5y ago
By the way, check out reviews of the Framework Laptop for a first-hand illustration of this effect. They seem to be getting 40-50% lower battery runtimes than laptops with comparable specs. That's example the problem — of you design fo
275.
by ribit 5y ago
RAM energy efficiency is extremely important because it is a significant part of the baseline power consumption. Power-hungry components like CPU and GPU can be run in the lowest power state most of the time, so it really comes down to RAM
276.
by ribit 5y ago
The beauty and success of DIMM modules is that they are a simple slide-in to install technology. It's easy to operate, cheap, robust, and it is very space efficient. Imagine RAM modules in an LGA socket instead... how much space would
277.
by ribit 5y ago
As far as I understand, a number of things make a difference: - Socketed RAM is limited by the number of pins, DDR SO-DIMM for example has 2-3x fewer pins than your average LPDDR4X module. Having more pins helps running the chips on lower p
278.
by ribit 5y ago
LPDDR is low power because it is a more sophisticated device with more complex protocol compared to regular DDR, because it has a higher pinout allowing it to have a more efficient electrical connection and because the computer manufacturer
279.
by ribit 5y ago
Unfortunately, I am not aware of any document like that. There are some papers comparing everything efficiency of different types of RAM, and there is of course indirect evidence of premium notebooks having significantly improved the batter
280.
by ribit 5y ago
> In what universe is that normal? In a universe where you want high-performance energy efficient RAM. Premium ultracompact laptops have long moved to LPDDR which is both faster and consumes significantly less energy than regular modular
281.
by ribit 5y ago
> "Another class of hybrid architecture is one that is often referred to as tile-based immediate-mode rendering. As dissected in this article, this hybrid architecture is used since NVIDIA’s Maxwell GPUs." Yes, they do mentio
282.
by ribit 5y ago
I don't think anyone has ever suggested that Maxwell or Navi were doing tiled deferred rendering, but all evidence (as well as technical documentation) point to them doing tiled immediate rendering. The tests in question rely on an ato
283.
by ribit 5y ago
This has been empirically tested on multiple occasions. There is an article on realwordtechnologies discussing this, and the results have been related for newer AMD GPUs as well. I have a little tool for macOS that tests these things out, a
284.
by ribit 5y ago
I think that the article focuses too much on the academic distinction between immediate renders and tilers but fails short to discuss how these techniques relate to real-world GPUs. For example, the fact that all contemporary AMD and Nvidia
285.
by ribit 5y ago
> Installing beta OSes side by side with your main OS is standard practice for app developers. This is not some arcane niche use. The problem is not a multi-OS environment, but multi-admin setups on secondary partition (which are extreme
286.
by ribit 5y ago
The most likely explanation for the behavior is that creating multiple admin users on a secondary partition is such a niche use case that Apple simply didn't test and/or consider it. This is a bug, report it and it will be fixed e
287.
by ribit 5y ago
Apple Maps is quite good if you use it one of the major US cities… but it’s still weak in most other places. As a European customer, Google maps have more up to date information, better traffic data, public construction information etc.
288.
by ribit 5y ago
> The M1 is probably the worst hardware you could have picked in 2020-2021 if computational power was your main concern. For highly > parallelizable work tasks, the top end GPU alone has 10x the computation power than the M1, and a t
289.
by ribit 5y ago
The previous beta 2.4 branch would also utilize AMX coprocessor when appropriate. Does anybody know whether the plugin is also based on ML Compute?
290.
by ribit 5y ago
This is all true, but I don't think that they are positioning this "Xcode Cloud" as an all-purpose continuous integration platform. It's really only for folks building iOS/macOS apps using standard templates and wor
291.
by ribit 6y ago
Macs didn't have any user-upgradeable parts since 2016..
292.
by ribit 6y ago
> It is, but I am afraid the next thing Apple will do now would > be to DRM the memory chips. Why "afraid"? I hope you are not considering resoldering BGA chips on tightly integrated machine a viable upgrade path?
293.
by ribit 6y ago
Apple M1, can do four fused multiply-adds per cycle with latency of 4 cycles. Interestingly enough it seems that the latency on the vector FMA is even lower. So it’s 16 float FMA per cycle. Source: https://dougallj.github.io&#x2F
294.
by ribit 6y ago
Maybe you are right. What I am reporting here are the impressions from my early PhD around 10 years ago when I was still actively participating in conferences and workshops on generative grammar. Most contributions I have seen were along th
295.
by ribit 6y ago
I don't think that you will find my answer satisfactory, but does there have to be an alternative? I think that many syntactic mysteries are mysterious only because syntacticians assume that they work in particular ways. Once you start
296.
by ribit 6y ago
This is very true. There is no denying that Chomsky and colleagues have contributed significantly to our understanding of the language. My comment is more in regards to the current day state of the field, which in my experience is overly dr
297.
by ribit 6y ago
I fully agree with you. I started studying linguistics since I come from a math background and Chomskian promise of mathematical linguistics was something I found very enticing. I quickly became disillusioned. Most of the Chomsky-inspired r
298.
by ribit 6y ago
Very true, but the problem is in paying these models to the actual study of linguistics and human language. For example, you can approach computation using something like a Turing Machine, but this won't help you understanding how to u
299.
by ribit 6y ago
I think this is more a testament to the quality of Adobe's GPU pipeline (or lack thereof). M1's GPU is surprisingly capable for a 10W part, and more importantly, it supports low-latency CPU/GPU communication and opt-in memory