Turbo, Power, and Latency

Turbo

As part of our usual test suite, we run a set of code designed to measure the time taken for the processor to ramp up in frequency. Recently both AMD and Intel are promoting features new to their processors about how quickly they can go from an active idle state into a turbo state – where previously we were talking about significant fractions of a second, we are now down to milliseconds or individual frames. Managing how quickly the processor fires up to a turbo frequency is also down to the silicon design, with sufficient frequency domains needing to be initialized up without causing any localised voltage or power issues. Part of this is also down to the OEM implantation of how the system responds to requests for high performance.

Our Ryzen 5 3600 jumped up from a 2.2 GHz high-performance idle all the way to 4.2 GHz in 16 milliseconds, which coincides exactly with a single frame on a 60 Hz display. This is right about where machines need to be in order to remain effective for a good user experience, assuming the rest of the system is up to scratch.

Power

With the Ryzen 5 3600, AMD lists the official TDP of the processor as 65 W. AMD also runs a feature called Package Power Tracking, or PPT, which allows the processor to turbo where possible to a new power value – for 65 W processors that new value is 88 W. This takes into account the power delivery capabilities of the motherboard, as well as the thermal environment. The processor can then manage exactly what frequency to give to the system in 25 MHz increments.

As part of my new test suite, we have a CPU power wrapper across several benchmarks to see the power response for a variety of different workloads.

For an AVX workload, y-Cruncher is somewhat periodic in its power use due to the way the calculation runs, but we see an almost constant 90 W peak power consumption through the whole test. The all-core turbo frequency here was in the 3875-3925 MHz range.

Our 3DPMavx test implements the highest version of AVX it can, for a series of six 10 second on, 10 second off tests, which then repeats. In this case we don’t see the processor going above 75 W in the whole process.

Photoscan is our more ‘regular’ test here, comprising of four stages each changing between single thread, multithread, and variable thread. We see peaks here up to 80 W, but the big variable threaded scenario bounces more around the 60 W mark for over 1000 seconds.

On the per-core power side, using our ray tracing power load, we see a small range of peak power values

When one thread is active, it sits at 12.8 W, but as we ramp up the cores, we get to 11.2 W per core. The non-core part of the processor, such as the IO chip, the DRAM channels and the PCIe lanes, even at idle still consume around 12-18 W in the system.

Latency

Our latency test is a simple core-to-core ping test, to detect any irregularities in the core design.

The results here are as expected.

  • 7.5 nanoseconds for threads within a core
  • 34 nanoseconds for cores within a CCX
  • 87-91 nanoseconds between cores in different CCXes
The AMD Ryzen 5 3600 Review Test Bed and Setup
Comments Locked

114 Comments

View All Comments

  • PeachNCream - Monday, May 18, 2020 - link

    Anandtech spends a lot of time on gaming and on desktop PCs that are not representative of where and how people now accomplish compute tasks. They do spend a little time on mobile phones and that nets part of the market, but only at the pricey end of cellular handsets. Lower cost mobile for the masses and work-a-day PCs and laptops generally get a cursory acknowledgement once in a great while which is disappointing because there is a big chunk of the market that gets disregarded. IIRC, AT didn't even get around to reviewing the lower tiers of discrete GPUs in the past, effectively ignoring that chunk of the market until long after release and only if said lower end hardware happened to be in a system they ended up getting. They do not seem to actively seek out such components, sadly enough.
  • whatthe123 - Monday, May 18, 2020 - link

    AI/tensorflow runs so much faster even on mid tier GPUs that trying to argue CPUs are relevant is completely out of touch. No academic in their right mind is looking for a bang-for-buck CPU to train models, it would be an absurd waste of time.
  • wolfesteinabhi - Tuesday, May 19, 2020 - link

    well ..games also run on GPU ...so why bother benchmarking CPU's with them? ... same reason why anyone would want to look at other workflows .. i said tensor flow as just one of the examples(maybe not the best example) ..but more of such "work" or "development" oriented benchmarks.
  • pashhtk27 - Thursday, May 21, 2020 - link

    Or there should be proper support libraries for the integrated graphics to run tensor calculations. That would make GPU-less AI development machines a lot more cost effective. AMD and Intel are both working on this but it'll be hard to get around Nvidia's monopoly of AI computing. Free cloud compute services like colab have several problems and others are very cost prohibitive for students. And sometimes you just need to have a local system capable of loading and predicting. As a student, I think it would significantly lower the entry threshold if their cost effective laptops could run simple models and get output.

    We can talk about AI benchmarks then.
  • Gigaplex - Monday, May 18, 2020 - link

    As a developer I just use whatever my company gives me. I wouldn't be shopping for consumer CPUs for work purposes.
  • wolfesteinabhi - Tuesday, May 19, 2020 - link

    not all developers are paid by their companies or make money with what they develop ... some are hobbyists and some do it as their "side" activities and with their own money at home apart from what they do at work with big guns!.
  • mikato - Sunday, May 24, 2020 - link

    As a developer, I built my own new computer at work and got to pick everything within budget.
  • Achaios - Monday, May 18, 2020 - link

    "Every so often there comes a processor that captures the market. "

    This used to be Sandy Bridge I5-2500K, all time best seller.

    Oh, how the Mighty Chipzilla has fallen.
  • mikelward - Monday, May 18, 2020 - link

    My current PC is a 2500K. My next one will be a 3600.
  • Spunjji - Tuesday, May 19, 2020 - link

    Sandy was an absolute knockout. Most of the development thereafter was aimed at sticking similarly powerful CPUs in sleeker packages rather than increasing desktop performance, and while I feel like Intel deserve more credit for some things than they get (e.g. the leap in mobile power/performance that can from Haswell) they really shit the bed on 10nm and responding to Ryzen.

Log in

Don't have an account? Sign up now