Top network posts

• 88 Emulate "double" using 2 "float"s
• 65 Why was the caret used for XOR instead of exponentiation?
• 52 128 bit integer on cuda?
• 47 Efficient floating-point division with constant integer divisors
• 31 What did Pete Stewart think he knew about efficient implementation of floating point denormals?
• 30 Optimized low-accuracy approximation to `rootn(x, n)`
• 27 Numerically stable way to compute sqrt((b²*c²) / (1-c²)) for c in [-1, 1]
• View more network posts →