I am actually quit interested in this.

送交者: steven 于 2012-03-26, 12:34:05:

回答: 史提文，GPU编程会不会变成将来科学计算的一个方式方法？由 Nixrreg 于 2012-03-26, 12:14:10:

Many years ago, I programmed on Connection Machine CM2, which is also a SIMD massive parallel processors architecture. It was quit useful. However, not all problems can be solve with parallel speed up,many problems are inherently serial.

所有跟贴:

我更关心的是编程的难以度，话很大时间学值不值得，想听听专家意见 - Nixrreg (252 bytes) 2012-03-26, 12:46:24 (655565)
- 不值得, you just need to understand the principle. - steven (204 bytes) 2012-03-26, 14:31:23 (655596)
- 我学过，大数组处理。不用学，用到自然会。 (无内容) - bullrider (0 bytes) 2012-03-26, 13:08:55 (655572)

加跟贴