New GPU Testing Methodology Puts Multi-GPU Solutions In Question

Vigile writes "A big shift in the way graphics cards and gaming performance are tested has been occurring over the last few months, with many review sites now using frame times rather than just average frame rates to compare products. Another unique testing methodology called Frame Rating has been started by PC Perspective that uses video capture equipment capable of recording uncompressed high resolution output direct from the graphics card, a colored bar overlay system and post-processing on that recorded video to evaluate performance as it is seen by the end user. The benefit is that there is literally no software interference between the data points and what the user sees, making it is as close to an 'experience metric' as any developed. Interestingly, multi-GPU solutions like SLI and CrossFire have very different results when viewed in this light, with AMD's offering clearly presenting a poorer, and more stuttery, animation."

Comments

You use GPUs for video games?

[amanicdroid]

My AMD is cranking out Bitcoin hashes 15 times faster than an equivalently priced Nvidia so I'm okay with the results of this article.

Re:You use GPUs for video games?

[Anonymous Coward]

People that "mine" bitcoins don't pay for their own electricity. Most people don't have the basement circuits metered separately from the rest of the house.

Re:You use GPUs for video games?

[lordofthechia]

Don't forget electrical costs. At $0.10 a kWh you are paying $0.24 a day (24 hours) per 100 watts of continuous average power consumption. This is $7.20 per month per 100W @ $0.10 /kWh or $87.60 a year. Adjust up/down for your cost of electricity and power usage (120W and $0.12/kWh = 1.2 * 1.2 = 1.44x adjustment)

Now add to this the waste heat vented into your house on the months you cool your house + the depreciated costs (and wear and tear) of the computer assets you tied up processing Bitcoins, then you'll have your true cost and you can calculate your break even point based on initial investment + ongoing costs - product (bitcoins) produced.

Re:You use GPUs for video games?

[amanicdroid]

Haha, I'm at less than 1:1 electricity to bitcoin ratio after ~5 months.
Kill-A-Watt says I've used approx $68.23 of electricity at 11.5 cents per kWh. Bitcoins currently trade at 1 to $30 and I've got 2.2 bitcoins. The Radeon 6770 was (and still is) ~$110.

Additional factors to consider:
-The bitcoin machine is also my daily workstation so if it were running headless and otherwise unused it would have probably done better in the electricity used category.
-It makes a damn fine space heater and I've enjoyed it immensely this winter.
-My focus in this project was to learn hands-on about scientific computing applications and it's been great for that.

In conclusion: as a business it would have been a flop, partially because I haven't optimized the setup for that application. As a learning opportunity and 200 watt heater it's been phenomenal.

Re:You use GPUs for video games?

[amanicdroid]

This is the explanation I've been given for the disparity between Nvidia and AMD:
https://en.bitcoin.it/wiki/Why_a_GPU_mines_faster_than_a_CPU#Why_are_AMD_GPUs_faster_than_Nvidia_GPUs.3F

Specifically:

Secondly, another difference favoring Bitcoin mining on AMD GPUs instead of Nvidia's is that the mining algorithm is based on SHA-256, which makes heavy use of the 32-bit integer right rotate operation. This operation can be implemented as a single hardware instruction on AMD GPUs (BIT_ALIGN_INT), but requires three separate hardware instructions to be emulated on Nvidia GPUs (2 shifts + 1 add). This alone gives AMD another 1.7x performance advantage (~1900 instructions instead of ~3250 to execute the SHA-256 compression function).

For GPU programming I've enjoyed Nvidia's CUDA package greatly over wrangling OpenCL that Radeon relies on.

Re:You use GPUs for video games?

[pclminion]

Out of curiosity, what's your break even point?

I don't know where the break even point is, but once you pass it, you can be very profitable. One of my friends built a custom "supercomputer" out of cheap motherboards and graphics cards for about $80k -- along with completely custom software to automatically tune clock speeds and fan rates in real time (all of which was written in bash script). At peak performance, his machine generated about $20k worth of bitcoin every month, which easily paid for the $12k monthly electric bill.

After a couple of difficulty-doublings, and the imminent arrival of the ASIC miners, this lost its profitability, and he went back to being a DBA... The machine is still out at the farm, cranking away. I think he'll disassemble it and part it out for cash in a month or two.

Regardless...

[Cinder6]

As an owner of a Crossfire setup, it's obviously not a 2x improvement over a single card; however, it's also a marked improvement over a single card. When I first set up this rig (August), I had problems with micro-stutter.* Now, though, after AMD's newer drivers and manually limiting games to 59 FPS, I don't see it anymore; games appear smooth as silk.

At a mathematical level, it may not be a perfect solution, but at a perceptual level, I am perfectly satisfied with my purchase. With that said, buying two mid-line cards instead of one high-end card isn't a good choice. Only buy two (or more) cards if you're going high-end.

*I was initially very disappointed with the Radeons. That's no longer the case, but I will probably still go nVidia the next time I upgrade, which hopefully won't be for years.

Another test I'm seeing more of

[gman003]

99th percentile frame times. That gives you a realistic minimum framerate, discarding most outliers (many games, particularly those using UE3, tend to have a few very choppy frames right on level load, that don't really affect performance).

Developers

[LBt1st]

This is interesting from a developer standpoint as well. This means we are wasting processing time rendering frames and are only displayed for a handful of milliseconds. These frames could be dropped entirely and that processing time could go to use elsewhere.