hrydgard

15 ms·

91.
by hrydgard 9y ago
Gamma is alive and well on pretty much all displays.
92.
by hrydgard 9y ago
Modern ARM actually has a 16-bit immediate version of MOV, plus MOVT which is the same but moves to the top 16 bits of a register. With a pair of those, you can load any 32-bit value in 2 instructions, without polluting the data cache like
93.
by hrydgard 9y ago
The company was probably founded some time before the service launched...
94.
by hrydgard 9y ago
Swedish does it but only the five before/after half case, not 10. And generally no other number than five actually. Examples: "Fem över halv två (13:35)" or "fem i halv fyra (15:25)".
95.
by hrydgard 9y ago
Well, it should also be noted that the responsible compiler in this case is at least 17 years old.
96.
by hrydgard 9y ago
For big mostly-C++ projects, it's considerably smoother than 2.3. It's still pretty slow to open and "gradle sync".
97.
by hrydgard 9y ago
It is very common on the PSP. Even actual traditional self modifying code is not uncommon.
98.
by hrydgard 9y ago
What? GBA is clearly easier than NES. Just set up a single tile layer in mode 0, and you got what's practically a NES with more colors, sprites and a much more powerful CPU that you can program in C++ if you like.
99.
by hrydgard 9y ago
Though they remain in Vulkan as an optional feature, supported by all the desktop GPUs - so they're aren't going away anytime soon. Of course, how much of that is driver fakery and how much is actual hardware support can be debate
100.
by hrydgard 9y ago
Your second snippet will just zero rdx - don't confuse the registers (rbx is in the mix in the original :)
101.
by hrydgard 9y ago
3D games use a rectangular perspective projection - they project the scene to a 2D rectangle. Problem is that the FOV (field of view, zoom) is not correct unless you put your head very close to the monitor. This leads to distortion. For exa
102.
by hrydgard 9y ago
There's also RenderDoc available for Windows and Linux and maybe more, it's not great with OpenGL but excellent for debugging Vulkan and D3D11.
103.
by hrydgard 9y ago
Chicken is best, pork second and beef the worst. The difference is bigger than you'd expect.
104.
by hrydgard 9y ago
Not true, SSE2 added support for performing all the MMX operations, but at 128-bit width in the SSE registers. AVX2 expanded that out to 256-bit width later.
105.
by hrydgard 9y ago
Switzerland absolutely has factories. Mostly mid/high end mechanical stuff but they are there, and people work in them.
106.
by hrydgard 9y ago
To be accurate, it was the marginal tax rate that ended up 102%, not the total tax rate. Still, of course, that was completely preposterous and the current system is very different, though still high-tax in international comparisons.
107.
by hrydgard 9y ago
Zipped GLSL (if you're using short variable names) is usually smaller than the corresponding DX byte code, so it turns out to actually be an advantage.
108.
by hrydgard 9y ago
It is actually possible to do using a feature called dual source blending, which unfortunately is so rarely used that drivers tend to fail to implement it properly. But it does give you the possibility to use 3+ alpha channels with full har
109.
by hrydgard 10y ago
You can choose to build practical denser cities even in huge states. There'll just be more space between them.
110.
by hrydgard 10y ago
N64 arguably yes, Saturn .. depends, but a bit of a mess. Jaguar and 3DO? No way, PSX was overwhelmingly better hardware than either of those. No idea about PC-FX though.
111.
by hrydgard 10y ago
It's faster to pass small structs in registers, and on x86-64 there's more of them so doing that is more practical than on x86.
112.
by hrydgard 10y ago
Yes, it has its occasional uses and is still supported, but the point is that x86-64 got a noticable FP capability boost through the doubling of the SSE registers compared to 32-bit x86. SSE is not just for SIMD, it has single/double s
113.
by hrydgard 10y ago
The FPU stack is only 8 entries in x86-64, yes, but the 128-bit SSE float/SIMD register set was indeed expanded to 16 entries, just like the GPRs (rax, rcx etc). The FPU stack is legacy and is barely used anymore, instead most floating
114.
by hrydgard 10y ago
I-cache and D-cache means instruction cache and data cache.
115.
by hrydgard 10y ago
I guess the velocity measurement "amount of functionality loaded per second" did improve by 500%, a bit contrived though.
116.
by hrydgard 10y ago
One way is to spread out the too-high value across neighbouring pixels, so-called bloom. This helps in creating a sensation of brightness even if of course no pixel actually has a higher value than 1.0. The easiest way is to make a copy of
117.
by hrydgard 10y ago
I understood that the configurations currently in use usually only power up either the big or little cores at the same time, and that kind of migration has to wipe the caches, right? But that might be inaccurate, and you are of course right
118.
by hrydgard 10y ago
It's fine if core migration happens during the invalidation loop - the core migration itself surely must wipe the non-shared cache levels thoroughly, otherwise nothing would work. EDIT: Actually if the big and little cores are used tog
119.
by hrydgard 10y ago
NULL is actually only a valid memory location from the view of the GPU on this platform, not the CPU, and yeah, this is very likely to simply be an uncaught bug. Traditional game console architectures don't usually have a lot of memory
120.
by hrydgard 11y ago
No processor out there has 64-bit arithmetic immediates - it's not useful enough in practice that you'd want it, and most other architectures have fixed-length 4-byte instructions anyway so it's not practical.

More ›