hajile

126 ms·

1591.
by hajile 5y ago
I used a pixelbook with Android apps as a daily driver for a couple years. I wouldn't use the touchscreen unless I absolutely had to. It gunks up the screen and a mouse is just more precise. I had the option of "tablet mode"
1592.
by hajile 5y ago
There’s actually three major version on N3 (sorry, I don’t have time to look up the similar articles on N5/7, but I think wikichip has a couple). https://www.techdesignforums.com/blog/2021/06/03/thre
1593.
by hajile 5y ago
That figure is compute workloads. Also, those measurements are “at the wall” and include RAM, SSD, and everything else. At the very least, you must subtract most idle power, but you also need to subtract more when DRAM/SSD kicks up fro
1594.
by hajile 5y ago
They have 5.2 and 10.4 TFLOPS of GPU power to feed in addition to 10 very wide cores.
1595.
by hajile 5y ago
The A15 efficiency cores will be in the next model. They are A76-level performance (flagship-level for Android from 2019-2020), but use only a tiny bit more power than the current efficiency cores. At that point, their E-cores will have som
1596.
by hajile 5y ago
M1 only allows one external monitor (you can run 2 with some docks if you turn off the laptop screen). This isn't such a problem for me as I have an ultra-wide, but lots of people with dual/triple monitor setups haven't been
1597.
by hajile 5y ago
When I start up kubernetes, my usage for that alone goes up to 6+GB. Swapping to SSD is terrible for it's lifecycle. 32GB should have you going for quite a while unless you need/want lots of vRAM in which case I'd go with 64G
1598.
by hajile 5y ago
The M1 Max is slightly off-square and 432mm^2. 19.1 x 22.6 seems like a decent fit. TSMC has stated that N7 defect density was 0.09 a year ago or so. They have also since stated that N5 defect density was lower than N7. Let's plug that
1599.
by hajile 5y ago
At worst, wait until they release these in the iMac Pro and maybe even the Mac Mini. Both easily have the headroom for these chips.
1600.
by hajile 5y ago
But if you run TWO single-threaded workloads at once, the clocks dial WAY back. Meanwhile, the M1 can keep all cores at 3.2GHz pretty much indefinitely.
1601.
by hajile 5y ago
M1 only uses around 15-18w with all 8 cores active according to Anandtech's review (with E-cores using 1/10 the power, that's equivalent to just a little less than 4.5 big cores. I'd guess 30w for all cores would be the
1602.
by hajile 5y ago
1/3 the performance, but 1/10 the power. Not adding more was a mistake IMO. Maybe next time...
1603.
by hajile 5y ago
M1 uses TSMC high-density rather than high-performance. They get 40-60% better transistor density and less leakage (power consumption) at the expense of lower clockspeeds. Also, a core is not necessarily just limited by power. There are oft
1604.
by hajile 5y ago
> Lisp like syntax is very peculiar. It puts off novices and pros alike. It would be more proper to say that it has no syntax. It's like everything is just a function call. `add(1, 2, 3, max(4, 5))` vs `(+ 1 2 3 (max 4 5))` or `_.re
1605.
by hajile 5y ago
Apple is finally taking that more seriously. The 6.1" pro and the 6.7" max have the same hardware. Likewise, the 5.4" mini has no compromises over the 6.1" iphone except battery life. I used Android for about a decade, b
1606.
by hajile 5y ago
I feel we should take a moment to appreciate the USB-C end on the magsafe charger so we can replace the cable when it breaks instead of paying for the entire brick.
1607.
by hajile 5y ago
If I don't add a brace, I'll get an immediate error that something doesn't add up. If I mess up indentation, the program will carry on as if nothing had happened.
1608.
by hajile 5y ago
I think that was more a function of being the only browser language. The target of JS wasn't pro developers in those days anyway, so familiarity wasn't really a consideration. Most importantly, all the "new" features of
1609.
by hajile 5y ago
Compose macros have existed for decades to handle that issue.
1610.
by hajile 5y ago
If he'd been allowed to implement Scheme instead, none of these arguments would have ever been needed in the first place.
1611.
by hajile 5y ago
The 3080M has something around 20TFLOPS of theoretical f32 performance. That's double that of Apple's biggest offerings. In theoretical compute, it's closer to the 3060M which is nothing to sneeze at.
1612.
by hajile 5y ago
We solved that almost a decade ago now with HBM. Sure, the latencies aren't amazing, but the power consumption numbers are and large caches can hide the higher access latencies pretty well in almost all cases.
1613.
by hajile 5y ago
AMD showed with their Infinity Cache that you can get away with much less bandwidth if you have large caches. It has the side effect of radically reducing power consumption. Apple put 32MB of cache in their latest iPhone. 128 or even 256MB
1614.
by hajile 5y ago
In A15, Anandtech claims the Efficiency cores are 1/3 the performance, but 1/10 the power. They should be looking at (effectively) doubling the power consumption over M1 with just the CPUs and assuming they don't increase clo
1615.
by hajile 5y ago
It looks like an interesting project. How does ISLisp compare to Common Lisp?
1616.
by hajile 5y ago
Sweden’s vaccination rates (75% with another 5-10% having the first shot) are not really lower than surrounding countries. No reason to lock down at that point unless the vaccine is ineffective.
1617.
by hajile 5y ago
`const` means a constant pointer, but doesn't guarantee that the data it points to will remain constant. In practice, functions, numbers, bigint, symbols, booleans, strings, regex literals, null, and undefined are all immutable. Since
1618.
by hajile 5y ago
Those are pretty impressive results for what didn't amount to a huge amount of changes (mostly just adding some types).
1619.
by hajile 5y ago
From a few years ago until we perfect the QFET and can put it to use.
1620.
by hajile 5y ago
I'm really starting to believe that mini-LED displays are generally better for long-term use.

More ›