Few days ago I bought this Asus Radeon X1950 Pro. Funny fact - I owned the same card 10 years ago when it was still modern. At the time is was considered to be very silent and even for today's standards it is good.
Just a quick testing - it can survive several loops of 3DM06, so all good.
And this is not all...
I'm preparing for another round of GPU benchmarking. There wil be two parts - the first one you can see in the table below. It is more or less a reboot of this test here
http://hw-museum.cz/article/3/benchmark-vga-2004---2008--2012-edition-/1 That one ended up like half-way finished. I started it in 2011 and worked on it till 2012... but at the time other things got in my way and this project was set aside. Now, when I considered to resume the work... it turned out the methodology is way too obsolete and thus unacceptable for me
So this time I will try to do it better, also better HW will be used - instead of i5-2500k - overclocked Sandy Bridge-E Xeon will do the job this time. It means full 16+16 SLI/CF and for the second part also 3-way SLI/CF. I'm still not sure whether go with 4.8 GHz SB-E or try to get i7 5930k (I already have X99 board). I don't think gaming performance of Haswell-e (at 4.5 - 4.6 GHz) is that much better to be worth $200-300 extra spend on this project. And given the fact I can go up to 2560x1440 resolution (which wasn't possible for me in 2011), CPU limitation shouldn't be an issue.
Also since I will be testing multi-GPU, there will be some kind of frametimes analysis of course. Not really the "pro" level which can be achieved by FCAT, but fraps or similar SW measuring should be good enough. The tricky part is to choose the right method of presenting the measured results (there will be a lot of them). I'm sure will figure something out.
Last thing - the last ATi GPU is hidden, this will be something special, stay tuned
