Creator Mode and Game Mode

*This page was updated on 8/17. A subsequent article with new information has been posted.

Due to the difference in memory latency between the two pairs of memory channels, AMD is implementing a ‘mode’ strategy for users to select depending on their workflow. The two modes are called Creator Mode (default), and Game Mode, and control two switches in order to adjust the performance of the system.

The two switches are:

  • Legacy Compatibility Mode, on or off (off by default)
  • Memory Mode: UMA vs NUMA (UMA by default)

The first switch disables the cores in one fo the silicon dies, but retains access to the DRAM channels and PCIe lanes. When the LCM switch is off, each core can handle two threads and the 16-core chip now has a total of 32 threads. When enabled, the system cuts half the cores, leaving 8 cores and 16 threads. This switch is primarily for compatibility purposes, as certain games (like DiRT) cannot work with more than 20 threads in a system. By reducing the total number of threads, these programs will be able to run. Turning the cores in one die off also alleviates some potential pressure in the core microarchitecture for cross communication.

The second switch, Memory Mode, puts the system into a unified memory architecture (UMA) or a non-unified memory architecture (NUMA) mode. Under the default setting, unified, the memory and CPU cores are seen as one massive block to the system, with maximum bandwidth and an average latency between the two. This makes it simple for code to understand, although the actual latency for a single instruction will be a good +20% faster or slower than the average, depending on which memory bank it is coming from.

NUMA still gives the system the full memory, but splits the memory and cores into into two NUMA banks depending on which pair of memory channels is nearest the core that needs the memory. The system will keep the data for a core as near to it as possible, giving the lowest latency. For a single core, that means it will fill up the memory nearest to it first at half the total bandwidth but a low latency, then the other half of the memory at the same half bandwidth at a higher latency. This mode is designed for latency sensitive workloads that rely on the lower latency removing a bottleneck in the workflow. For some code this matters, as well as some games – low latency can affect averages or 99th percentiles for game benchmarks.

The confusing thing about this switch is that AMD is calling it ‘Memory Access Mode’ in their documents, and labeling the two options as Local and Distributed. This is easier to understand than the SMT switch, in that the Local setting focuses on the latency local to the core (NUMA), and the Distributed setting focuses on the bandwidth to the core (UMA), with Distributed being default.

  • When Memory Access Mode is Local, NUMA is enabled (Latency)
  • When Memory Access Mode is Distributed, UMA is enabled (Bandwidth, default)

So with that in mind, there are four ways to arrange these two switches. AMD has given two of these configurations specific names to help users depending on how they use their system: Creator Mode is designed to give as many threads as possible and as much memory bandwidth as possible. Game Mode is designed to optimize for latency and compatibility, to drive game frame rates.

AMD Threadripper Options
  Words That Make Sense   Marketing Spiel
Ryzen
Master
Profile
Two Dies or
One Die
Memory
Mode
  Legacy
Compatibility
Mode
Memory
Access
Mode
Creator Mode Two UMA   Off Distributed
- Two NUMA   Off Local
- One UMA   On Distributed
Game Mode One NUMA   On Local

There are two ways to select these modes, although this is also a confusing element to this situation.

The way I would normally adjust these settings is through the BIOS, however the BIOS settings do not explicitly state ‘Creator Mode’ and ‘Game Mode’. They should give immediate access for the Memory Mode, where ASUS has used the Memory Access naming for Local and Distributed, not NUMA and UMA.  For the Legacy Compatibility Mode, users will have to dive several screens down into the Zen options and manually switch off eight of the cores, if the setting is going to end up being visible to the user. This makes Ryzen Master the easiest way to implement Game Mode.

While we were testing Threadripper, AMD updated Ryzen Master several times to account for the latest updates, so chances are that by the time you are reading this, things might have changed again. But the crux is that Creator Mode and Game Mode are not separate settings here either. Instead, AMD is labelling these as ‘profiles’. Users can select the Creator Mode profile or the Game Mode profile, and within those profiles, the two switches mentioned above (labelled as Legacy Compatibility Mode and Memory Access Mode) will be switched as required.

Cache Performance

As an academic exercise, Creator Mode and Game Mode make sense depending on the workflow. If you don’t need the threads and want the latency bump, Game Mode is for you. The perhaps odd thing about this is that Threadripper is aimed at highly threaded workloads more than gaming, and so losing half the threads in Game Mode might actually be a detriment to a workstation implementation.  That being said, users can leave SMT on and still change the memory access mode on its own, although AMD is really focusing more on the Creator and Game mode specifically.

For this review, we tested both Creator (default) and Game modes on the 16-core Threadripper 1950X. As an academic exercise we looked into memory latency in both modes, as well as at higher DRAM frequencies. These latency numbers take the results for the core selected (we chose core 2 in each case) and then stride through to hit L1, L2, L3 and main memory. For UMA systems like in Creator Mode, main memory will be an average between the near and far memory results. We’ve also added in here a Ryzen 5 1600X as an example of a single Zeppelin die, and a 6950X Broadwell for comparison. All CPUs were run at DDR4-2400, which is the maximum supported at two DIMMs per channel. 

For the 1950X in the two modes, the results are essentially equal until we hit 8MB, which is the L3 cache limit per CCX. After this, the core bounces out to main memory, where the Game mode sits around 79ns while the Creator mode is at 108 ns. By comparison the Ryzen 5 1600X seems to have a lower latency at 8MB (20ns vs 41 ns), and then sits between the Creator and Game modes at 87 ns. It would appear that the bigger downside of Creator mode in this way is the fact that main memory accesses are much slower than normal Ryzen or in Game mode.

If we crank up the DRAM frequency to DDR4-3200 for the Threadripper 1950X, the numbers change a fair bit:


Click for larger image

Up until the 8MB boundary where L3 hits main memory, everything is pretty much equal. At 8MB however, the latency at DDR4-2400 is 41ns compared to 18ns at DDR4-3200. Then out into full main memory sees a pattern: Creator mode at DDR4-3200 is close to Game Mode at DDR4-2400 (87ns vs 79ns), but taking Game mode to DDR4-3200 drops the latency down to 65ns.

Another element we tested while in Game Mode was the latency for near memory and far memory as seen from a single core. Remember this slide from AMD’s deck?

In our testing, we achieved the following:

  • At DDR4-2400, 79ns near memory and 136ns far memory (108ns average)
  • At DDR4-3200, 65ns near memory and 108ns far memory (87ns average)

Those average numbers are what we get for Creator mode by default, indicating that the UMA mode in Creator mode will just use memory at random between the two.

Silicon, Glue, & NUMA Too Test Bed and Setup
Comments Locked

347 Comments

View All Comments

  • nitin213 - Thursday, August 10, 2017 - link

    Thanks for your reply. Hopefully the test suite can be expanded as Intel's CPUs probably also move to higher core count and IO ranges in future.
    and i completely understand the frustration trying to get a 3rd party to change their defaults. Cheers
  • deathBOB - Thursday, August 10, 2017 - link

    It's clear to me . . . Ian is playing both sides and making out like a bandit! /s
  • FreckledTrout - Thursday, August 10, 2017 - link

    Ian can we get an updated comments section so we can +/- people and after x number of minuses they wont show by default. I'm saying this because some of these comments(the one in this chain included) are not meaningful responces. The comments section is by far the weakest link on Anantech.

    Nice review btw.
  • mapesdhs - Thursday, August 10, 2017 - link

    toms has that, indeed it's kinda handy for blanking out the trolls. Whether it's any useful indicator of "valid" opinion though, well, that kinda varies. :D (there's nowt to stop the trolls from voting everything under the sun, though one option would be to auto-suspend someone's ability to vote if their own posts get hidden from down voting too often, a hands-off way of slapping the trolls)

    Given the choice, I'd much rather just be able to *edit* what I've posted than up/down-vote what others have written. I still smile recalling a guy who posted a followup to apologise for the typos in his o.p., but the followup had typos aswell, after which he posted aaaaagh. :D

    Ian.
  • Johan Steyn - Thursday, August 10, 2017 - link

    Ian thanks for at least responding, I appreciate it. Please compare your review to sites like PCPer and many others. They have no problem to also point out the weak points of TR, yet clearly understand for what TR was mostly designed and focus properly on it and even though they did not test the 64 PCI lanes as an example, mention that they are planning a follow-up to do it, since it is an important point. You do mention these as well, but could have said more than just mention it by the way.

    Look at your review, most of it is about games. Are you serious?

    I have to give you credit to at least mention the problems with Sysmark.

    Let me give you an example of slanted journalism, When you do the rendering benchmarks, where AMD is known to shine, you only mention at each benchmark what they do etc, and fail to mention that AMD clearly beats Intel, even though other sites focus more ons these benchmarks. In the one benchmark where Intel get a descent score, you take time to mention that:

    "Though it's interesting just how close the 10-core Core i9-7900X gets in the CPU (C++) test despite a significant core count disadvantage, likely due to a combination of higher IPC and clockspeeds."

    Not in one of the rendering benchmarks do you give credit to AMD, yet you found it fitting to end the section of with:

    "Intel recently announced that its new 18-core chip scores 3200 on Cinebench R15. That would be an extra 6.7% performance over the Threadripper 1950X for 2x the cost."

    Not slanted journalism? At least you mention "2x the cost," but for most this will not defer them in buying the monopoly.

    After focussing so much time on game performance, I am not sure you understand TR at all. AMD still has a long way to go in many areas. Why? Because corrupt Intel basically drove them to bankruptcy, but that is a discussion for another day. I lived through those days and experienced it myself.

    Maybe I missed it, but where did you discuss the issue of memory speed? You mention in the beginning of memory overclock. Did you test the system running at 3200 or 2666? It is important to note. If you ran at 2666, then you are missing a very important point. Ryzen is known to gain a huge amount with memory speed. You should not regard 3200 as an overclock, since that is what that memory is made for, even if 2666 is standard spec. Most other sites I checked, used it like that. If you did use 3200, don't you think you should mention it?

    Why is it that your review ends up meh about TR and leave you rather wanting an i9 an almost all respects, yet most of the other site gives admiration where deserved, even though they have criticism as well. Ian I see that you clearly are disappointed with TR, which is OK, maybe you just like playing games and that is why you are so.

    It was clear how much you admire Intel in your previous article. You say that I gave no examples of slanted journalism, maybe you should read my post again. "Most Powerful, Most scalable." It is well known that people don't read the fine print. This was intentional. If not, you are a very unlucky guys for having so many unintended mishaps. Then I truly need to say I am sorry.

    For once, please be a bit excited that there is some competition against the monopoly of Intel, or maybe you are also deluded that they became so without any underhanded ways.

    By the way, sorry that I called you Anand. I actually wanted to type Anandtech, but left it like it. This site still carries his name and he should still take responsibility. After I posted, I realised I should have just checked the author, so sorry about that.
  • vanilla_gorilla - Thursday, August 10, 2017 - link

    "Intel recently announced that its new 18-core chip scores 3200 on Cinebench R15. That would be an extra 6.7% performance over the Threadripper 1950X for 2x the cost."

    How do you not understand that is a dig at Intel? He's saying you have to pay twice as much for only a 6.7% improvement.
  • smilingcrow - Thursday, August 10, 2017 - link

    The memory speed approach taken was clearly explained in the test and was stated as being consistent with how they always test.
    I don't take issue with testing at stock speeds at launch day as running memory out of spec for the system can be evaluated in depth later on.
  • Johan Steyn - Friday, August 11, 2017 - link

    That is just rubbish. Threadripper has no problem with 3200 memory and other sites has no problem running it at that speed. 3200 memory is designed to run 3200, why run it at 2666? There is just no excuse except being paid by Intel.

    Maybe then you can accuse other sites of being unscientific?
  • fanofanand - Tuesday, August 15, 2017 - link

    Anandtech always tests at JDEC, regardless of the brand.
  • Manch - Friday, August 11, 2017 - link

    ""Intel recently announced that its new 18-core chip scores 3200 on Cinebench R15. That would be an extra 6.7% performance over the Threadripper 1950X for 2x the cost."

    Not slanted journalism? At least you mention "2x the cost," but for most this will not defer them in buying the monopoly."

    You call Intel the monopoly and call him out for not wording the sentence to dissuade people from buying Intel. Who has the bias here? If he was actively promoting Intel over AMD or vice versa, you'd be OK with the latter, but to do neither. He's an Intel shill? Come on. That's unfair. HOW should he have wrote it so it would satisfy you?

    FYI Anand is gone. He's NOT responsible for anything at Anandtech. Are you going to hold Wozniak's feet to the fire for the lack of ports on a Mac too?

Log in

Don't have an account? Sign up now