CPU Hardware Reviews

2018 CPU and APU Charts (Update)

Much does not always help much - but on the other hand, a system that still looks slightly oversized today may reach its limits tomorrow. You can probably never have enough performance and there is also a certain feeling of future... We start with two synthetic benchmarks, but consciously share this category in DirectX11 and DirectX12. In the 3DMark Fire Strike, the core count dominates the physics calculation, which also means that the older, not very high-clockcore i7-695... Similar to the 3DMark runs, Ashes of the Singularity: Escalation focuses on the core number, followed by the beat. It is a good example of proper scaling over many possible threads.With Civilization VI help zwa... In Warhammer 40K: Dawn of War III, THE CPU clock comes into play when you have more than four well-scaling threads. This throws the current Ryzen CPUs a little backwards, so intel can dominate. Grand Thef... Under Hitman (2016), the AMD world looks pretty okay here. From a certain basic performance of the CPU (Intel Core i5-8400) the graphics card already limits anyway. However, it is a deliberately chosen example, which shows well,... Far Cry Primal is our second example with GPU limit, but slightly different in interpretation. Since the game copes well with up to 8 threads, but does not necessarily rely on real cores, quad-cores with hyper-threading can also be... Since AutoCAD can only scale very poorly (3D) to not (2D) over several threads, the available clock and thus of course also the height of the achievable IPC counts here. Intel is clearly ahead of the way, even if AMD is doing well with the Ryzen CPUs. Here, too, IPC goes beyond sheer core counts, as both programs in the design grid cannot (can) parallelize well across many threads. The two OpenGL classics also benefit from the clock and from a good mid-range CPU, whether from Intel or AMD, the performance is sufficient to very good. All three programs are true core scorners in real-time 3D output, but in turn scale with the beat. Solidworks 2015, like Creo 3.0, not only uses the CPU to realize the graphics output, but also, in contrast to the free SPECviewperf, also uses computing power to simulate or to create perform calculations and create rendering images. The ideal counter-example is pure rendering tasks such as Blender or Luxrender, with the latter using the console version. At Blender both Ryzen Threadripper - Nomen est Omen win, the rest sorts itself well behind it, where here e... With the CPU Composite, the AMD Ryzen 7 are almost closed in front, the thread rippers are right behind them. More than 12 to 14 threads are barely used, so the Intel Core i7-8700(K) can also shine. Without hyper-threading, d... If you only screw the quality settings high enough in Handbrake, the AMD Ryzen Threadripper sshine almost at will and even the cheaper Ryzen 7 still hold well with...... while the hand turns at 7Zip. Here In... We measure the power consumption for the package or the entire CPU directly behind the voltage converters, the losses of which should of course also be planned. The advantage of our direct measurement or Sensor detection is due to the fact that the value... Games can be very different in power consumption, depending on the actual CPU load. Our example shows a rather average scenario, which is most likely to match the average value of all games in our suite. The bad news in advance: THE best CPU does not really exist as a general statement, if you include all facets such as application purpose, actually required performance, overall concept of the PC and your own budget. The one for a ...

Much does not always help much – but on the other hand, a system that still looks slightly oversized today may reach its limits tomorrow. Therefore, one can probably never have enough performance and there is also a certain feeling of future and planning security, if one (justifiably) believes that one is still well prepared for future requirements.

Update from May 3, 2018

We added the new AMD Ryzen 7 2700X, 2700, Ryzen 5 2600X and 2600, as well as repeated some older benchmarks with new BIOS and tested plausibility. All benchmarks are now up to date, but may differ slightly from older publications.

Such an assumption can work, but it does not have to. First, you will have to think about what the PC needs to be prepared for in the present and near future, and what might be added later. That's exactly what you should focus on – and plan a little reserve.

Unfortunately, however, more performance always costs more money – often disproportionately – which is why one has to question the benefits of a (massive) performance reserve extremely critically.

We had just touched on the subject: often enough, demands, desires and financial possibilities are not very close together. But here the miracle medicine "reason" helps, which should lead the user to a willingness to compromise and an insight into the immutable facts. Aspects such as ecology (power consumption and longevity) and economics (effort and benefits) should always be in a healthy relationship. We just wrote it: It's best to buy only what you really need (or. in the near future).

In addition, we also offer instructions on how to check a system for stability and performance limits after assembly:

But it's not just gaming. Whether 2D or 3D output – the requirements in the (semi-) professional area or the home office are very diverse. Depending on the standard software used or the program suites, you will come across very long-lasting software products with an often multi-year usage cycle. This also means that many (older) programs are more likely to (still) rely on a high IPC and hardly scale properly over many threads, or in return also perform calculations in parallel, the load of which is so high that in the case of incorrectly used CPUs, the performance of the graphics output suffers disproportionately.

For many areas, it no longer necessarily has to be an Intel Xeon, because often the consumer version is enough if you can do without ECC RAM. With the budget thus saved, there is almost always more room for manoeuvre for the graphics hardware.

The test setup

We have already described the test methodology in the basic article "How we test graphics cards, as of February 2017" in great detail and so we now only refer to this detailed description for the sake of simplicity. So if you want to read everything again, you are welcome to do so. 

In this case, only the hardware configuration with CPU, RAM, motherboard, as well as the new cooling is different, so that the summary in table form quickly gives a brief overview of the system used here and today:

Test systems and measuring rooms
AMD Socket AM4 (Ryzen Gen.2 CPUs)
MSI X470 Gaming M7 AC
2x 8GB G.Skill FlareX DDR4-3200

Socket AM4 (Ryzen Gen.1 CPUs)
MSI X370 Tomahawk
2x 8 GB G.Skill TridentZ DDR4-3200 RGB

AMD Socket AM4 (APUs)
2x 8GB G.Skill FlareX DDR4 3200

AMD Socket SP3 (TR4)
Asis X399 ROG Zenith Extreme
4x 8 GB G.Skill TridentZ DDR4-3200 RGB

AMD Socket AM3+
Asus Sabertooth 990FX
2x 8 GB Corsair Dominator Platinum DDR3 2133

Intel Socket 1151 (Z370):

MSI Z370 Gaming Pro Carbon AC
4x 8 GB G.Skill TridentZ DDR4-3600 RGB

Intel Socket 1151 (Z270):
MSI Z270 Gaming 7
2x 8GB Corsair Vengeance DDR4-3200-2666 MHz

Intel Socket 2066
MSI X299 Gaming Pro Carbon AC
4x 8 GB G.Skill TridentZ DDR4-3200 RGB

Intel Socket 2011v3:
Intel Core i7-6900K
MSI X99S XPower Gaming Titanium
4x 4 GB Crucial Ballistix DDR4-2400

All systems:
GeForce GTX 1080 Founders Edition (Gaming)
Nvidia Quadro P6000 (Workstation)

1x 1 TByte Toshiba OCZ RD400 (M.2, System SSD)
4x 1050 GByte Crucial MX 300 (Storage, Images)
Be Quiet Dark Power Pro 11, 850-watt power supply
Windows 10 Pro (all updates)

Alphacool Ice Age 2000 Chiller
Alphacool Ice Block XPX
Thermal Grizzly Kryonaut (for cooler change)
Monitor: Eizo EV3237-BK
Lian Li PC-T70 with expansion kit and modifications
Modes: Open Benchtable, Closed Case
Power consumption:
non-contact DC measurement on the PCIe slot (Riser-Card)
non-contact DC measurement on the external PCIe power supply
direct voltage measurement at the shunts, the respective feeders and the power supply
Reading out the motherboard sensors
2x Rohde & Schwarz HMO 3054, 500 MHz multi-channel oscillograph with memory function
4x Rohde & Schwarz HZO50, current togor adapter (1 mA to 30 A, 100 KHz, DC)
4x Rohde & Schwarz HZ355, touch divider (10:1, 500 MHz)
1x Rohde & Schwarz HMC 8012, digital multimeter with storage function
Optris PI640, infrared camera
PI Connect evaluation software with profiles
NTI Audio M2211 (with calibration file)
Steinberg UR12 (with phantom power for the microphones)
Creative X7, Smaart v.7
own low-reflection measuring room, 3.5 x 1.8 x 2.2 m (LxTxH)
Axial measurements, perpendicular to the center of the sound source(s), measuring distance 50 cm
Noise in dBA (Slow) as RTA measurement
Frequency spectrum as a graph

Danke für die Spende

Du fandest, der Beitrag war interessant und möchtest uns unterstützen? Klasse!

Hier erfährst Du, wie: Hier spenden.

Hier kannst Du per PayPal spenden.

About the author

Igor Wallossek

Editor-in-chief and name-giver of igor'sLAB as the content successor of Tom's Hardware Germany, whose license was returned in June 2019 in order to better meet the qualitative demands of web content and challenges of new media such as YouTube with its own channel.

Computer nerd since 1983, audio freak since 1979 and pretty much open to anything with a plug or battery for over 50 years.

Follow Igor:
YouTube Facebook Instagram Twitter