The latest series of Ryzen CPUs has been on the market for six weeks and yet we only got our hands on the Ryzen 7 3800X about a week ago. Due to the delay, we received numerous comments asking us to review and compare it with the 3900X, 3700X, and 3600, which we have all reviewed by now.
So what's the deal? Why was the 3800X so difficult to get, how is it different from the 3700X and why did the TDP increase by over 60% when the boost frequency was increased by 100 MHz? Apparently it's all about binning and returns that may not be as good as AMD hoped, or maybe demand has played a role.
Silicon Lottery recently released some Ryzen 3000 binning data. This indicates that better quality silicon has been reserved for the 3800X. The top 20% of all 3800X processors tested passed their 4.3 GHz AVX2 stress test, while the top 21% of all 3700X processors were only stable at 4.15 GHz. In addition, all 3800X processors passed the test at 4.2 GHz, while 3700X processors were only good at 4.05 GHz, which means that the 3800X has about 150 MHz more headroom when overclocking.
Silicon Lottery AMD Ryzen 3000 binning results
In other words, the average 3800X should overclock better than the best 3700X processors, but it's still a slight 6% frequency difference between the absolute worst 3700X and the absolute best 3800X per test. For more casual overclockers like us, the difference will likely be even less. Our 3700X appears to be stable in our internal stress test and has never crashed at 4.3 GHz. This is the same frequency limit for the 3800X in retail that we have received. As far as the TDP is concerned it is confusing to say the least, but we'll go through some performance metrics first and then discuss what we think is going on.
The benchmarks shown below were tested on the Gigabyte X570 Aorus Xtreme with 16 GB DDR4-3200 CL14 memory. The graphics card of choice for CPU tests is of course the RTX 2080 Ti. So let's get to the numbers.
First we have Cinebench R20 and hold on to your hats. We expect a 3% increase in multi-core performance on the 3800X, but hey, at least it's now faster than the 9900K.
We see an increase of around 2% in single-core performance, so that the 3800X matches the 3900X and 9900K.
The coding time in Premiere is reduced by 3%, which shortens our test by 12 seconds. Given what we saw in Cinebench R20, this result is not surprising.
We also see a 3% reduction in render time when testing with Blender Open Data, so not much more to say. Let's take a quick look at a few games before we look at power consumption and thermals.
In Assassin's Creed Odyssey between 3700X, 3800X and 3900X we really look at the margin of the types of errors. Needless to say, they all offer a similar gaming experience in this title.
The same applies to tests with Battlefield V: The 3800X is 1-2 fps faster than the 3700X, thanks to a very small increase in operating frequency. You cannot notice this difference in performance here either.
A practically identical performance is also found when testing with The Division 2. We look at the performance of the Core i7-8700K in this title.
Finally we have Shadow of the Tomb Raider and here the 3800X was 2 fps faster than the 3700X. So let's continue with power consumption.
Here we see a 9% increase in power consumption, which is surprising since we are only talking about a 2-3% increase in operating frequency. However, this frequency surge will be accompanied by a surge in voltage, and that's probably why the 3800X was a bit more power hungry than you might have expected.
As a result of this increased power consumption, the 3800X runs 3 degrees hotter with the box cooler and 4 degrees hotter with the Corsair H115i Pro. Interestingly, the 3800X clocked 100 MHz higher with the box cooler, but only 75 MHz higher with the Corsair AIO.
When activating PBO, the 3800X was still 2-3 degrees hotter, but now it is only 25-50 MHz higher than the 3700X. So if you want to turn your 3700X into a 3800X, just activate PBO.
Because of the power consumption, the 3800X consumes significantly more power than the 3700X, at least compared to the additional power it offers. We saw a performance improvement of up to 3%, which cost us an increase in power consumption of around 12%.
The margins remain largely the same when PBO is activated. The 3800X still uses around 12% more power than the 3700X.
It wasn't the most exciting benchmark session ever, but it answered the question: what's the difference?
As it turns out, not much. At high workloads, the 3800X clocks between 100 and 150 MHz higher, which corresponds to a frequency increase of 2.5 to 4%. This increased CPU power consumption by around 12%, which meant it was running a few degrees hotter and possibly getting a bit louder.
For this slight performance boost, AMD raised the MSRP by 21% from $ 330 to $ 400. The biggest percentage increase, if we ignore the TDP, comes from the price. And we believe that's all you need to hear. At best, you get 3% more performance if you spend 21% more money.
If you are interested in this offer we have a bunch of old Xeon systems … you can have them for a really good price.
Let's talk briefly about the 105 W TDP, which has been increased by 62% compared to the 65 W TDP power of the 3700X. It seems that AMD basically says this: with a cooler that is designed for heat dissipation of 65 watts, the 3700X does not run lower than its base clock. The 3800X, which is clocked 300 MHz higher for the base, may not be able to hold 3.9 GHz with a 65 watt cooler.
The confusion creeps in when AMD skips its 95-watt output for 105 watts with the 3800X. We accept that the 3800X may not be able to maintain 3.9 GHz with a 65-watt cooler, but certainly with a 95-watt cooler.
As far as we can tell, the TDP is a metric for OEMs who usually try to cut as many corners as possible. When an OEM puts a 65W cooler on a 95W part and a buyer says "I don't hit 3.9 GHz," AMD can say, "Well, the OEM doesn't meet the basic specification for the cooler."
Ultimately, we don't understand why TDP is advertised when it is less useful to consumers. AMD may be better off just posting the rating of the cooler they provide for each processor and telling customers what type of cooler performance they need if they want to upgrade. For example, the Wraith prism is a 105 W cooler. Therefore, you want something above the value when you want to upgrade, e.g. B. 150 watts.
Short buying advice: We strongly recommend avoiding the 3800X and taking the 3700X instead. If you think it is necessary, update the box cooler with the saved money. The Ryzen 5 3600 remains the king of values, and the 3900X offers more cores for productivity. Games may not benefit, as we found in our GPU scaling benchmark.