Trinity CPU Performance: The Good and the Bad

We're going to start our performance investigation a little out of order. The big question on everyone's mind is how much single threaded performance has improved over Bulldozer, and whether it's enough to actually make Trinity faster than Llano across the board. We'll use Cinebench 11.5 as it has both single and multithreaded test options:

Cinebench 11.5 - Single Threaded

The good news is that single threaded performance is definitely up compared to Llano. Piledriver likely has some to do with this, but so does the fact that the A10 can run at up to 4.2GHz (~4GHz typically) with one of its cores active compared to the 2.9GHz clock speed of the A8-3850. Compared to the Bulldozer based FX-8150 there's a slight (~6%) increase in single threaded performance. Although I don't expect anyone will be cross shopping a Trinity APU and a FX CPU, it's important to keep an eye on progress here as we'll eventually get a high-end quad-module/eight-core Piledriver CPU.

Note that compared to even previous generation, low-end Intel CPUs without turbo there's a huge gap in single threaded performance. If we look at the gap AMD has to make up vs. Ivy Bridge it's not pretty. Intel's Core i3 3220 manages a 27% performance advantage over the A10-5800K. Even if Steamroller is able to deliver a 15% increase in performance at the same clock speed, there will still be a gap. And we're not even talking about how Haswell will grow this gap. For the foreseeable future I don't see AMD closing the single threaded performance gap. Jim Keller's job is to fix this problem, but it'll probably take 2 - 3 years to get there.

Cinebench 11.5 - Multi-Threaded

The multithreaded test shows the other end of reality: in heavily threaded foating point workloads it's possible that we'll see a regression compared to Llano. Remember the Bulldozer/Piledriver architecture prioritizes integer over floating point performance. Truth be told this regression is pretty rare in our tests, but until we get to Steamroller we will still see these types of situations.

Throw more threads at the problem and even with a floating point workload Intel can't pull ahead however. The A10 offers similar performance to the Core i3 3220 at a lower price. Your decision here would come down to the rest of the factors: single threaded performance, processor graphics performance, overclocking capabilities and power consumption. Intel and AMD both win two of those each, it's really a matter of what matters most to you.

A heavily threaded FP workload doesn't really play to AMD's advantages though, what happens when you get a heavily threaded integer workload however? The 7-zip benchmark gives us just that:

7-zip Benchmark

Here AMD manages a 16% performance advantage over the Core i3 3220. I'd even go as far as to say that Trinity would likely beat any dual-core Intel machine here. The performance advantage is somewhat artificial as Intel purposefully removes turbo from its dual-core desktop CPUs. This should be AMD's best foot forward, but once again it'll likely take Steamroller for this design to start to make sense.

Speaking of artificial product segmentation, one major feature Intel takes away when you get down to the dual-core desktop i3 level is AES-NI support. Hardware accelerated AES support is something that you get only with the more expensive Core i5/i7s. With Trinity, you get AES-NI support for the entire stack. The result is much better performance in those applications that depend on it:

AES-128 Performance - TrueCrypt 7.1 Benchmark

Like most of the advantages we've talked about thus far, there are really very specific use cases where Trinity makes sense over a similarly priced Intel CPU.

Introduction & The Test General Performance - SYSMark 2012
POST A COMMENT

177 Comments

View All Comments

  • mikato - Wednesday, October 03, 2012 - link

    Sounds legit. Reply
  • MySchizoBuddy - Tuesday, October 02, 2012 - link

    Can the GPU be overclocked or is it just for the CPU Reply
  • Medallish - Tuesday, October 02, 2012 - link

    AMD did a preview where they showed off an overclocked APU with the GPU pushed up to 1GHz, so I'm pretty sure it should be possible, but it might be a feature exclusive to some motherboards or chipsets. Reply
  • Doby - Tuesday, October 02, 2012 - link

    I don't understand why the load power is more important than idle. I don't know about most people, but I don't turn off any of the 3 desktops/media center PCs I have in the house, but I do know the majority of the time they sit idle. I haven't done the math, but I would bet quite a bit of money that over all I will spend less on my power bill by having a lower idle than having lower at load power draw.

    I could see the issue on portable machines running off battery, but even then running full out is unlikely, and it becomes a "hurry up and wait" scenario that probably requires better analysis.

    I feel like benchmark performance is a bit over rated, we need to see value the consumer can leverage. I want to know which one "feels" faster, likelihood of running into application issues such as graphics drives, large displays, or even stability problems. I want to know what I can recommend to my parents for a basic internet PC. Even if I was doing video transcoding I'm not sitting around waiting for it, if it takes more than 30 seconds I'm up not waiting for it, it might as well be 5 min.

    I know its easier to just post a bunch of benchmarks, and I do still like to see them. But lets progress computing to the next level, user experience. I'm fine with a "doesn't make a difference" answer too, but that's still better than "12% faster for a process that you won't wait for anyway".

    Don't get me wrong, I enjoy the article, and appreciate the write up. Just looking for a bit different education.
    Reply
  • Visual - Tuesday, October 02, 2012 - link

    I don't worry about power bills from running 24/7 idle. I have enough other devices at home that a PC wouldn't make a dent either way. But I do prefer if the computer cooling was not heard throughout the apartment.

    Power use under max load is important in selecting the case and cooling system, etc. You may load it that much only for a few minutes, but your setup still needs to handle that "worst case" situation. And if you are going for a quiet setup, it would be no comfort for you if it is quiet when not used, but racks up the noise and begins to bother you when you start using it.
    Reply
  • halbhh2 - Tuesday, October 02, 2012 - link

    Great questions, and we can see that the idle power draw of the A10 is good, and so it would make a good choice for typical use (which is 80-85% idle) for a typical use laptop, and it would do well obviously in playing many games, and can potentially do this at a reasonable cost (depending on the OEM maker, like HP). So, just like you, I wonder how long the typical battery run time is. That's how I bought my yr-old HP laptop -- just typical battery run time and a good display, and good price. I knew enough to know those were the parameters that would matter for our laptop, and it would be okay at occasional demanding use just as you describe -- when it takes more than 20-30 seconds, you are off doing something else anyway. Reply
  • Hubb1e - Tuesday, October 02, 2012 - link

    Doby, I agree with you. I think it really comes down to a yes or no question on each use case. Benchmarks are nice, but i3s and A10s are really not enthusiast level CPUs so 12% single threaded advantage doesn't matter in the long run. The question to ask is "Can it run my applications"

    And in the case of Trinity vs i3 the main difference is that it can play games while the i3 with HD2500 graphics can't. If you look at Diablo 3, that game was played by a lot of people that are not traditionally PC gamers. An i3 with HD2500 is barely playable while Trinity is a pleasure to use on Diablo 3. For most PC buyers I think Trinity makes a lot of sense where idle power matters most, CPU performance is competitive, and gaming is possible. People who buy off the shelf PCs are not that comfortable putting a GPU in their rigs so Trinity is a good option for a general use PC. Enthusiasts who aren't AMD fanatics will stay away from this chip and that's fine. It will be a sales success for AMD if they can make enough profit on it. It is a pretty big chip...
    Reply
  • Roland00Address - Tuesday, October 02, 2012 - link

    The reason why idle is unimportant for desktop is that both companies processors idle at such an insignificant amount.

    The Amd a10 idles 7 watts lower than the i3 and i5.

    1 watt used 24 hours a day 7 days a week is 8.76 kWh thus we are talking about 61.32 kWh (7*8.76) a year. The cost per kWh is different in differnt places in the US but it is about 10 cents a kWh so we are talking about 6 dollars more a less in energy savings a year.

    Aka talking about idle power is a virtually insignificant number for desktops. Now for laptops on the other hand it is a big deal for laptop idle power use affects battery life.
    Reply
  • phoenix_rizzen - Tuesday, October 02, 2012 - link

    If it's 8.76 kWh per week, should you be multiplying by 52 weeks in a year? Thus 455.52 kWh per year? Reply
  • Roland00Address - Tuesday, October 02, 2012 - link

    1 watt *1000 hours equals 1 Kilowatt hour which is abbreviated 1 kWh

    24 hours a day *365 days per year equals 8760 hours or 8.76 kWh

    I should have said 24 hours a day, 7 days a week, 52 weeks a year. I apologize for leaving off the 52 weeks a year part it was a slip of the tongue.
    Reply

Log in

Don't have an account? Sign up now