### quanpham0805's blog

By quanpham0805, history, 3 years ago,

In GCJ problems this year, the time limits are stated 20 seconds per test set. I don't know how many calculations could GCJ do in 1 second?

• +47

 » 3 years ago, # |   +14 I think, you can expect ~$10^8-10^9$ operations per second, depending of operations type.
•  » » 3 years ago, # ^ | ← Rev. 2 →   0 I think it should be ~10^8 for each test case, otherwise I could do something like 20*10^8 -> 20*10^9 = 2*10^9 -> 2*10^10 per test case, which is a little bit too much.
 » 3 years ago, # | ← Rev. 2 →   +16 It depends what language you are using. Assuming you use fastIO, the max bigO runtime per testcase you can have per language is approximately:C++: 3*10^8, Java: 10^8, Python: 152
•  » » 3 years ago, # ^ |   0 And idea about What will be the estimated operations if we're using costly operations like / and %?
•  » » » 3 years ago, # ^ |   -14 I have no idea for the /, but it should be equal to +, -, *.For the %, for example, a % b, I think it is implemented like a - (a / b * b) (assume that a and b are integer), so it might be 3 times slower, but I'm not sure about this neither.
•  » » 3 years ago, # ^ |   +7 Is that a typo for Python? It doesn't seem right that it should be 5-6 orders of magnitude slower than Java.
•  » » » 3 years ago, # ^ |   +5 Yeah, I thought Python was even slower than that
•  » » » 3 years ago, # ^ |   +5 Should be closer to 15.2
•  » » 3 years ago, # ^ |   0 I'm using C++ for all submissions with normal cin and cout, so it should be a little bit less I think (maybe around 10^8/test case, I usually assume this number of operations)
 » 3 years ago, # |   0 For a second I thought that we were making fun of the slow servers this year. (Looking at you, shameful KS round 1A)