AMD Carrizo Part 2: A Generational Deep Dive into the Athlon X4 845 at $70
by Ian Cutress on July 14, 2016 9:00 AM ESTReal World Performance at 3 GHz
For our generational testing, we took each of the four main processors in this test and adjusted their CPU frequencies in the BIOS to 3 GHz. This was achieved through a 30x multiplier and 100 MHz base frequency, which for each processor is a reduction from the stock speeds. We set each CPU to perform at 3 GHz only to fix the frequency, and ran the memory in each case at the maximum supported frequency by the processor. Some benchmarks in the generational tests will probe the memory, and an upgrade in the memory controller to support higher frequencies (officially) than an older processor is, a generational upgrade, as important as the core or cache performance.
| AMD CPUs | ||||||||||||
| µArch / Core |
Cores | Base Turbo |
TDP | DDR3 | L1 (I) Cache |
L1 (D) Cache |
L2 Cache |
|||||
| Athlon X4 845 |
Excavator Carrizo |
4 | 3500 3800 |
65 W | 2133 | 192KB 3-way |
128KB 8-way |
2 MB 16-way |
||||
| Athlon X4 860K |
Steamroller Kaveri |
4 | 3700 4000 |
95 W | 1866 | 192KB 3-way |
64KB 4-way |
4 MB 16-way |
||||
| Athlon X4 760K |
Piledriver.v2 Richland |
4 | 3800 4100 |
100 W | 1866 | 128KB 2-way |
64KB 4-way |
4 MB 16-way |
||||
| Athlon X4 750K |
Piledriver Trinity |
4 | 3400 4000 |
100 W | 1866 | 128KB 2-way |
64KB 4-way |
4 MB 16-way |
||||
Speaking of cache, as mentioned at the beginning of this review, the Athlon X4 845 has a significant advantage in the L1 cache layout, affording a 2x size L1 data cache along with a move from 4-way to 8-way associativity. Each of these methods, as a broad rule of thumb, typically decreases the cache miss rate by a factor of 1.414 (square root of 2x). Combined should see a factor two decrease in cache misses overall, and this will affect a number of benchmarks when we compare each processor at a fixed frequency. On the other side of the equation, the L2 cache for the X4 845 is half that of the X4 860K, meaning that if the data is not in the L1, it is less likely to be in the L2, which will add additional latency.
Dolphin Benchmark: link
Many emulators are often bound by single thread CPU performance, and general reports tended to suggest that Haswell provided a significant boost to emulator performance. This benchmark runs a Wii program that raytraces a complex 3D scene inside the Dolphin Wii emulator. Performance on this benchmark is a good proxy of the speed of Dolphin CPU emulation, which is an intensive single core task using most aspects of a CPU. Results are given in minutes, where the Wii itself scores 17.53 minutes.
Emulation takes cues from a high IPC and base frequency, however for our generational testing it is all about the microarchitecture. The Carrizo has a 9% advantage here over the Kaveri.
WinRAR 5.0.1: link
Our WinRAR test from 2013 is updated to the latest version of WinRAR at the start of 2014. We compress a set of 2867 files across 320 folders totaling 1.52 GB in size – 95% of these files are small typical website files, and the rest (90% of the size) are small 30 second 720p videos.
WinRAR enjoys memory bandwidth with its variable workload, and seemingly the Kaveri has a strong showing here. The Carrizo only has 2MB of L2 cache, which most likely puts it at a disadvantage.
3D Particle Movement v2
The second version of this benchmark is similar to the first, however it has been re-written in VS2012 with one major difference: the code has been written to address the issue of false sharing. If data required by multiple threads, say four, is in the same cache line, the software cannot read the cache line once and split the data to each thread - instead it will read four times in a serial fashion. The new software splits the data to new cache lines so reads can be parallelized and stalls minimized. As v2 is fairly new, we are still gathering data and results are currently limited.
We saw this in our laptop Carrizo testing: if we adjust the software to avoid false sharing (which decreases performance), the Excavator microarchitecture pulls a significant lead in 3DPMv2. Part of this is most likely down to the larger L1 data cache as well.
Web Benchmarks
On the lower end processors, general usability is a big factor of experience, especially as we move into the HTML5 era of web browsing.
WebXPRT 2013
This benchmark can be memory intensive, as it draws various graphs and applies filters to pictures, among other things. The lower L2 cache hurts here.
Google Octane v2
In contrast, Octane attempts to stay as close to the execution ports as possible, and the Carrizo cores take an 18% lead over Kaveri.
Facebook
Twitter
RSS
131 Comments
View All Comments
Peichen - Thursday, July 14, 2016 - link
I stopped reading at the messy chart where Athlon X4 750 is available from 2 µArch and there is no science to naming. AMD's CPU division was a bigger failure than its GPU division and it still it. I used to have 3 AMD systems at the same time back in 2005 but have steadily moved away and was considering replacing the last AMD system (Phenom II 840) using Prime Day sales.Ian Cutress - Thursday, July 14, 2016 - link
The 750 and 750K are two different SKUs. As mentioned, trying to find the 750 at retail (and in stock) is *really difficult*, and I'd probably point to it being an OEM-only SKU for a certain design. If anyone has a 750/can find one, let me know.artk2219 - Wednesday, July 20, 2016 - link
I found a link for one, but I think its just an OEM one, and it's from Aliexpress, so it may take a while to get to you. I hope I could help!http://www.aliexpress.com/item/AMD-Athlon-X4-750-A...
http://www.cpu-world.com/CPUs/Bulldozer/AMD-Athlon...
jefeweiss - Thursday, July 14, 2016 - link
I don't think poignant is the word you are looking for here. That would be something that makes you very sad, or emotionally touches you deeply.wallysb01 - Thursday, July 14, 2016 - link
I guess I don't totally understand the comparison to the G3258. I know its a fairly popular processor (I have its non-overclockable little brother, the G3250, and love it in a simple home office set up that I use occasionally for light gaming too). However, its 2 years old. Where is the G4500 or G4520? Supposedly the skylake Pentium would be about 10% faster than haswell, no? And with some perf/W gains too?Speaking of which, where's the comparison to Intel in the power consumption? Do I just not want to see that? If in these multithreaded tasks the 845 is chugging along using 80-85 W, while the Intel parts are still near their stated TDP, it more or less invalidates the small performance gain of the AMD chip in those tasks. Looking back at the skylake review, it seems at load the Intel chips might be using anything from 0-20% more than their TDP, this AMD chip is going 30% above.
I'd love it if AMD could push intel into offering more cores at lower price points, but this doesn't seem good enough to do that....
stardude82 - Thursday, July 14, 2016 - link
Tacking on a G4500 is not terribly pertinent to the task at hand and they are just using archived results here. The comparison from the 2 i3 parts should give you an idea of relative performance. Otherwise, I'm surprised we get such an in depth review from an hollowed out Anandtech.You are confusing TDP with total system draw.
You have a "four core" CPU here for $70, what more do you want? They soundly thrash the dual-core Pentiums and i3s in some well parallelized applications. The problem is the world isn't well parallelized and the CPUs don't have 4 real cores which is why you have i3s still competitive 8 core FX chips.
yannigr2 - Thursday, July 14, 2016 - link
WOW. This is an amazing work. It does explain too much about AMD's Bulldozer versions. It does explain why AMD doesn't bother to bring the full Carrizo line in desktop. It does show, at least from my perspective, that Bulldozer architecture was not a bad architecture to begin with, but a dead end from the beginning.artk2219 - Thursday, July 14, 2016 - link
Sigh, poor Llano and your FM1 package, always forgotten. Granted it was the last of the stars cores, and not a bulldozer derivative, but it would still be nice to see it included with its brethren.TheinsanegamerN - Thursday, July 14, 2016 - link
The last of AMD's good cores. Mobile Llano was fantastic for OCing and undervolting.artk2219 - Thursday, July 14, 2016 - link
Indeed it was/is, my fiance is still using my old Asus K53TA. I overclocked the A6 (back when A6's came with 4 cores) in that laptop to run at 2.4 instead of 1.4 for the base, and the turbo to 3Ghz. while overclocking i was able to undervolt at base clocks, and slightly overvolt the turbo clocks. In effect I got better battery life and waaayyyy better performance by just spending a little time playing with the P-States. I still think thats one of the best laptops I've ever owned in terms of reliability and capability, if only it weren't so clunky. Sigh, well you cant have it all.http://www.newegg.com/Product/Product.aspx?Item=N8...