CPU Tests: Science

In this version of our test suite, all the science focused tests that aren’t ‘simulation’ work are now in our science section. This includes Brownian Motion, calculating digits of Pi, molecular dynamics, and for the first time, we’re trialing an artificial intelligence benchmark, both inference and training, that works under Windows using python and TensorFlow.  Where possible these benchmarks have been optimized with the latest in vector instructions, except for the AI test – we were told that while it uses Intel’s Math Kernel Libraries, they’re optimized more for Linux than for Windows, and so it gives an interesting result when unoptimized software is used.

3D Particle Movement v2.1: Non-AVX and AVX2/AVX512

This is the latest version of the benchmark designed to simulate semi-optimized scientific algorithms taken directly from my doctorate thesis. This involves randomly moving particles in a 3D space using a set of algorithms that define random movement. Version 2.1 improves over 2.0 by passing the main particle structs by reference rather than by value, and decreasing the amount of double->float->double recasts the compiler was adding in.

The initial version of v2.1 is a custom C++ binary of my own code, flags are in place to allow for multiple loops of the code with a custom benchmark length. By default this version runs six times and outputs the average score to the console, which we capture with a redirection operator that writes to file.

For v2.1, we also have a fully optimized AVX2/AVX512 version, which uses intrinsics to get the best performance out of the software. This was done by a former Intel AVX-512 engineer who now works elsewhere. According to Jim Keller, there are only a couple dozen or so people who understand how to extract the best performance out of a CPU, and this guy is one of them. To keep things honest, AMD also has a copy of the code, but has not proposed any changes.

The final result is a table that looks like this:

(2-1) 3D Particle Movement v2.1 (non-AVX)(2-2) 3D Particle Movement v2.1 (Peak AVX)

The 3DPM test is set to output millions of movements per second, rather than time to complete a fixed number of movements. This way the data represented becomes a linear when performance scales and easier to read as a result.

y-Cruncher 0.78.9506: www.numberworld.org/y-cruncher

If you ask anyone what sort of computer holds the world record for calculating the most digits of pi, I can guarantee that a good portion of those answers might point to some colossus super computer built into a mountain by a super-villain. Fortunately nothing could be further from the truth – the computer with the record is a quad socket Ivy Bridge server with 300 TB of storage. The software that was run to get that was y-cruncher.

Built by Alex Yee over the last part of a decade and some more, y-Cruncher is the software of choice for calculating billions and trillions of digits of the most popular mathematical constants. The software has held the world record for Pi since August 2010, and has broken the record a total of 7 times since. It also holds records for e, the Golden Ratio, and others. According to Alex, the program runs around 500,000 lines of code, and he has multiple binaries each optimized for different families of processors, such as Zen, Ice Lake, Sky Lake, all the way back to Nehalem, using the latest SSE/AVX2/AVX512 instructions where they fit in, and then further optimized for how each core is built.

For our purposes, we’re calculating Pi, as it is more compute bound than memory bound. In single thread mode we calculate 250 million digits, while in multithreaded mode we go for 2.5 billion digits. That 2.5 billion digit value requires ~12 GB of DRAM, so for systems that do not have that much, we also have a separate table for slower CPUs and 250 million digits.

(2-3) yCruncher 0.78.9506 ST (250m Pi)(2-4) yCruncher 0.78.9506 MT (2.5b Pi)

y-Cruncher is also affected by memory bandwidth, even in ST mode, which is why we're seeing the Xeons score very highly despite the lower single thread frequency.

Personally I have held a few of the records that y-Cruncher keeps track of, and my latest attempt at a record was to compute 600 billion digits of the Euler-Mascheroni constant, using a Xeon 8280 and 768 GB of DRAM. It took over 100 days (!).

NAMD 2.13 (ApoA1): Molecular Dynamics

One of the popular science fields is modelling the dynamics of proteins. By looking at how the energy of active sites within a large protein structure over time, scientists behind the research can calculate required activation energies for potential interactions. This becomes very important in drug discovery. Molecular dynamics also plays a large role in protein folding, and in understanding what happens when proteins misfold, and what can be done to prevent it. Two of the most popular molecular dynamics packages in use today are NAMD and GROMACS.

NAMD, or Nanoscale Molecular Dynamics, has already been used in extensive Coronavirus research on the Frontier supercomputer. Typical simulations using the package are measured in how many nanoseconds per day can be calculated with the given hardware, and the ApoA1 protein (92,224 atoms) has been the standard model for molecular dynamics simulation.

Luckily the compute can home in on a typical ‘nanoseconds-per-day’ rate after only 60 seconds of simulation, however we stretch that out to 10 minutes to take a more sustained value, as by that time most turbo limits should be surpassed. The simulation itself works with 2 femtosecond timesteps.

(2-5) NAMD ApoA1 Simulation

How NAMD is going to scale in our testing is going to be interesting, as the software has been developed to go across large supercomputers while taking advantage of fast communications and MPI.

AI Benchmark 0.1.2 using TensorFlow: Link

Finding an appropriate artificial intelligence benchmark for Windows has been a holy grail of mine for quite a while. The problem is that AI is such a fast moving, fast paced word that whatever I compute this quarter will no longer be relevant in the next, and one of the key metrics in this benchmarking suite is being able to keep data over a long period of time. We’ve had AI benchmarks on smartphones for a while, given that smartphones are a better target for AI workloads, but it also makes some sense that everything on PC is geared towards Linux as well.

Thankfully however, the good folks over at ETH Zurich in Switzerland have converted their smartphone AI benchmark into something that’s useable in Windows. It uses TensorFlow, and for our benchmark purposes we’ve locked our testing down to TensorFlow 2.10, AI Benchmark 0.1.2, while using Python 3.7.6 – this was the only combination of versions we could get to work, because Python 3.8 has some quirks.

The benchmark runs through 19 different networks including MobileNet-V2, ResNet-V2, VGG-19 Super-Res, NVIDIA-SPADE, PSPNet, DeepLab, Pixel-RNN, and GNMT-Translation. All the tests probe both the inference and the training at various input sizes and batch sizes, except the translation that only does inference. It measures the time taken to do a given amount of work, and spits out a value at the end.

There is one big caveat for all of this, however. Speaking with the folks over at ETH, they use Intel’s Math Kernel Libraries (MKL) for Windows, and they’re seeing some incredible drawbacks. I was told that MKL for Windows doesn’t play well with multiple threads, and as a result any Windows results are going to perform a lot worse than Linux results. On top of that, after a given number of threads (~16), MKL kind of gives up and performance drops of quite substantially.

So why test it at all? Firstly, because we need an AI benchmark, and a bad one is still better than not having one at all. Secondly, if MKL on Windows is the problem, then by publicizing the test, it might just put a boot somewhere for MKL to get fixed. To that end, we’ll stay with the benchmark as long as it remains feasible.

(2-6) AI Benchmark 0.1.2 Total

As you can see, we’re already seeing it perform really badly with the big chips. Somewhere around the Ryzen 7  is probably where the peak is. Our Xeon chips didn't really work at all.

CPU Tests: Office CPU Tests: Simulation
Comments Locked

110 Comments

View All Comments

  • PeachNCream - Tuesday, July 21, 2020 - link

    You don't get what it means to perform a controlled test do you?
  • Aspernari - Wednesday, July 22, 2020 - link

    It's important to note that the environment is not actually well-controlled.

    https://twitter.com/IanCutress/status/128480609693...

    We don't know temperature for the operating conditions for these tests, which matters more and more for boost behavior for CPUs and GPUs. He says 36c when he got into the office, we'll never know what the temperature peaked at, nor how often similar conditions were reached.

    A standard platform is a good choice, but a controlled environment is also important. Unfortunately, the results aren't as reliable as they otherwise might have been.
  • PeterCollier - Wednesday, July 22, 2020 - link

    And that's why this entire test is a complete waste of time. Something like Geekbench or especially Userbench is much, much better because it gives you a range of scores. Instead of trying to create false precision by saying that a AMD 4700U scored, say, a "979" on a benchmark, Userbench will say that all the 4700U's tested scored from 899 to 1008, and break it down into percentiles. This way, you have a range of expected performance in mind instead of being fixated on that "979" number, which could have been obtained in an unrealistic scenario.
  • Rudde - Saturday, July 25, 2020 - link

    Isn't userbench a synthetic together with geekbench? What exactly are they testing? Instead of knowing which of Intel i7 10700k and AMD ryzen 7 3800X is better at rendering, video encoding, number crunching or whatever your use case is, you'll get a distribution based on a largely unknown test. The Intel and AMD processors might end up being within error margins of each other in your use case, but that in itself tells something too. All benchmarks are inherently bad; there is not a single benchmark that captures every use case while not being affected by its environment (ram speeds, temperatures, etc). I prefer tests that I understand, over tests that I do not understand.
  • bananaforscale - Wednesday, July 22, 2020 - link

    One could ask what the point of Userbenchmark is in these days of quadcores being basically entry level while the benchmark has DECREASED its multicore weighting.
  • A5 - Monday, July 20, 2020 - link

    For my own personal test, getting an i7-4770K in the list would be a big help.

    Once you have a compile test, a Xeon E5-1680v3 would be nice to see so that I can sell my corp on newer workstations...
  • Shmee - Wednesday, July 22, 2020 - link

    Those are great Haswell EP CPUs, and they OC too! I have an E5-1660v3 in my X99 rig.
  • Mockingtruth - Monday, July 20, 2020 - link

    I have a 3570k and a E8600 spare with respective motherboards and ram if useful?
  • CampGareth - Monday, July 20, 2020 - link

    Personally I'd like to see a Xeon E5-2670 v1 benchmarked. I'm still running a pair of them as my workstation but these days AMD can beat the performance on a single socket and halve the power consumption.
  • Samus - Tuesday, July 21, 2020 - link

    Do you run them in an HP Z620? I ran the same system with the same CPU’s for years at one of my clients. What a beast.

Log in

Don't have an account? Sign up now