dagmx

7 ms·

1.
by dagmx 5d ago
Geekbench when run on Asahi Linux is practically the same as when run on macOS, just like Linux and Windows in the same x86 chip don’t give significant differences.
2.
by dagmx 5d ago
Because it ruffles x86 feathers, despite geekbench corresponding very closely to SPEC as well and there being no indication that the results aren’t comparable. What people willfully conflate this with often is that you shouldn’t compare FLO
3.
by dagmx 5d ago
You have that exactly backwards. The MacBook Neo released with the previous years A series Pro chip. It’s currently got an A18 Pro.
4.
by dagmx 10d ago
Your intuition here is wrong. Processors, not just Apple’s, change constantly between generations. The core instruction set doesn’t change, but they get new capabilities, not just with new instruction sets. Remember the SoC isn’t just the C
5.
by dagmx 17d ago
Sure, you can also have iOS binaries forward relevant iOS SDK symbols to macOS. That doesn’t mean they’re not different SDKs with different limitations, and not everything that is on the store can be patched to run in the simulator because
6.
by dagmx 19d ago
The simulator is a different SDK target than iOS itself. You have to compile completely independently for it, and depending on your dependencies they may not compile for the simulator. Additionally the simulator runs a really ancient and fe
7.
Nvidia Starts Pac as AI Chip Maker Builds DC Influence Force(bloomberg.com)
2 points by dagmx 19d ago|0 comments
8.
by dagmx 22d ago
HBM itself is very expensive but it’s not really fair to compare to LPDDR or GDDR They’re very different things. The more logical argument to me is that Apple uses its upgrade price points as more than just direct BOM and rather as a proxy
9.
by dagmx 22d ago
I’m not trying to say you shouldn’t be excited by it. But there’s a significant number of caveats to every single number you posted but you glossed over the points I made. And in being defensive, you jumped straight from the topic you were
10.
by dagmx 23d ago
I’m not sure I follow why the performance is “Insane” This is a M5 level chip that runs slightly within a ballpark of M5 wattage (afaik M5 is between 15W max on an iPad and 30W max CPU+GPU on a cooled Mac) This is running in a thermally unc
11.
by dagmx 1mo ago
Oh yes, Claude seems to be loving the word vacuous recently. My test bed side project is full of vacuous this and that now. I even try and get it to define what it classifies as vacuous and it can’t do so without getting stuck in some kind
12.
by dagmx 1mo ago
Ah fair, I forgot Panasonic owned Lumix. Though the lawsuits are more complex with multiple parties suing RED, of which Nikon also fought RED and ended up buying them out to win that side of the battle. On the panasonic side it was more of
13.
by dagmx 1mo ago
Unless I’ve missed it; Lumix has nothing to do with this long standing patent battle. Could you elaborate on the supposed relationship?
14.
by dagmx 1mo ago
I disagree with your assessment. My twitter feed pre and post Elon’s takeover changes dramatically. Pre his takeover , it would more reliably show me my niches of design, technology, gamedev etc After his take over I’d start seeing more and
15.
by dagmx 2mo ago
Yes because SwiftUI is way more self contained and minimally verbose, it’s a better fit for LLM use than AppKit/UIKit
16.
by dagmx 2mo ago
Thank you very much for the corrections here. As someone else who has professionally worked in the film space, it always irks me when people talk about film in digital resolutions for the reason you mention. I also agree with many of your o
17.
by dagmx 2mo ago
A post mortem is a very common term in many industries on how something went after you shipped it. You’re taking it very literally here, it does not involve anything being dead or killed. We’ve used this in the software, games, film and con
18.
by dagmx 2mo ago
Corresponding Apple press release that isn’t paywalled https://www.apple.com/newsroom/2026/07/apple-maps-to-power-n...
19.
by dagmx 2mo ago
The commercial Autodesk apps still dominate, as do Maxon apps like C4D for motion graphics and zbrush for sculpting. The big problems are that license cost doesn’t outweigh the pain points. In production I can get support from Autodesk for
20.
by dagmx 2mo ago
Do you have comparisons for Apple’s GPU to NVIDIA’s at the same power and thermals? Right now the M5 Max is one of the best laptop compute GPUs for blender rendering and doesn’t do too badly on raster tests either. But it operates at a frac
21.
by dagmx 2mo ago
A lot of unreal engine games also fork the engine. If your bar is being forked then unreal doesn’t meet the bar for a lot of people either. 3D Engines of that fidelity are rarely genre specific. Editors might be but the engines are often fa
22.
by dagmx 2mo ago
Your first statement is wrong. The new Indiana jones game is not an FPS and it uses the same engine
23.
by dagmx 2mo ago
Oh I definitely agree the C-suite at Microsoft needs blame. Microsoft need everything to be framed as avenue into services. But Phil was the one who dug two incompatible holes and kept digging hoping they’d meet in the middle with buried tr
24.
by dagmx 2mo ago
Okay so your entire argument hinges on her having the ear of Satya , and out arguing Phil? But even then you wouldn’t be able to deny that Xbox was floundering under Phil. And for all of this you’d still be left with conjecture?
25.
by dagmx 2mo ago
How would Asha pull him out? How do you think she even had that kind of authority or seniority?
26.
by dagmx 2mo ago
Based on what? This just feels like baseless conjecture given she hasn’t been at Microsoft for that long.
27.
by dagmx 2mo ago
This is incredibly sad for a lot of my friends who are finding themselves out of work despite delivering well received products. But at the same time I appreciate the candor of Asha saying that the corporate management are to blame and lett
28.
by dagmx 2mo ago
I found this paper quite disappointing. It requires a fairly expensive precomputation pass and can only work for static scenes. Meanwhile interactive path tracing is fast enough that the scenes they showed would only be minorly slower to be
29.
by dagmx 3mo ago
Anything AI focused in silicon is also valuable for a ton of other use cases. If LLMs and GenAI don’t pan out, that silicon just gets used for other processing. Then they scale back on the dedicated die space in subsequent generations. It’s
30.
by dagmx 3mo ago
Where would they make that which isn’t constrained already?

More ›