For comparison, a couple weeks ago Oak Ridge National Laboratory's Titan computer topped the Top500 list for fastest supercomputer in the world (although that title is a bit misleading). Here's the specs on Titan:

18,688 AMD Opteron 6274 16-core CPUs

18,688 Nvidia Tesla K20 GPUs

17.59 petaflops

Titan displaced Sequoia (at Lawrence Livermore National Laboratory) from the top spot on Top500. Interestingly enough, Sequoia uses a very different architecture, based on 16-core PowerPC A2 nodes rather than GPUs. Sequoia also has about 1.6PB of memory, while Titan "only" has 1PB.

Both computers have reasonably different use cases. GPUs are great for embarrassingly parallel, non-memory intensive tasks like brute forcing passwords. But all of the rumors about the NSA's massive data analysis needs suggests that they may need a cluster that resembles Sequoia (with fewer cores, but larger caches and available memory) more than Titan.

I'm sure they were able to crack the encryption method before Microsoft started using it.

Put your tin foil hats away. This doesn't "crack" NTLM, it brute forces at a very high rate. The NSA has more money to spend, but are similarly limited by the hardware available at any given point.

The NSA has access to their own chip fabrication facilities. I do not know if they own their own plant, or just have secure fab space at some other company's plant.

So they could have easily fabbed something like this, or a tuned architecture specifically designed for the purpose.

NSA does not own a foundry anymore. Modern top-of-the-line foundries cost too much, even for the NSA.

Instead, they participate in a program to partner with domestic companies to manufacture their chips: http://trustedfoundryprogram.org/

It's very probable that they have and if so, it's almost certain that it would involve specialized hardware implementations (ASIC, FPGA, whatever) rather than commodity graphics hardware which is burdened with expensive and useless stuff like onboard memory and would be power inefficient

considering that the entire purpose of NSA in the first place is to provide SIGINT and encrypt or decrypt signals, it's almost a given that they're trying to the best of their ability to crack stuff.

Not sure

Money buys more commodity hardware faster than the time/money used to develop a chip

It's not hard to make tens, or maybe even hundreds of GPUs beat a specialized chip except for very specific things

And even for something specialized it's probably easier to use an FPGA

I wouldn't call a big cluster of commodity hardware simply commodity hardware because then many of the top supercomputers of the world are commodity hardware.

The combination is exotic enough to be considered non-commodity hardware.