Black Shark, Red Magic vs mainstream flagships

Theoretical performance

To setup the playing field we ran some regular synthetic benchmark runs. Those serve to show that the two gaming phones have the exact same chips as the other Snapdragon 855 flagships, which in turn aren't far off from the Exynos and Kirin flagships.

AnTuTu 7

Higher is better

• OnePlus 7
  367812
• Sony Xperia 1
  356734
• Black Shark 2
  343460
• Samsung Galaxy S10e
  325192
• Huawei P30 Pro (perf. mode)
  316156
• ZTE nubia Red Magic 3
  314975
• Huawei P30 Pro
  290189
• Xiaomi Pocophone F1
  265314
• Xiaomi Mi 9T
  211915

There were no surprises really - the Black Shark 2 and the ZTE nubia Red Magic 3 would post scores virtually identical to other current flagships, showing that the theoretical maximum performance is the same across the board.

GeekBench 4.1 (multi-core)

Higher is better

• ZTE nubia Red Magic 3
  11226
• Black Shark 2
  11192
• OnePlus 7
  11075
• Sony Xperia 1
  10985
• Samsung Galaxy S10e
  10081
• Huawei P30 Pro (perf. mode)
  10014
• Huawei P30 Pro
  9649
• Xiaomi Pocophone F1
  9003
• Xiaomi Mi 9T
  6863

GeekBench 4.1 (single-core)

Higher is better

• Samsung Galaxy S10e
  4518
• Black Shark 2
  3515
• ZTE nubia Red Magic 3
  3493
• OnePlus 7
  3461
• Sony Xperia 1
  3447
• Huawei P30 Pro (perf. mode)
  3323
• Huawei P30 Pro
  3270
• Xiaomi Mi 9T
  2537
• Xiaomi Pocophone F1
  2438

3DMark SSE 3.1 Unlimited

Higher is better

• OnePlus 7
  6388
• ZTE nubia Red Magic 3
  6360
• Black Shark 2
  6330
• Sony Xperia 1
  5792
• Samsung Galaxy S10e
  4545
• Huawei P30 Pro (perf. mode)
  4315
• Huawei P30 Pro
  3522
• Xiaomi Mi 9T
  2329

3DMark SSE 3.1

Higher is better

• Black Shark 2
  5784
• OnePlus 7
  5745
• ZTE nubia Red Magic 3
  5218
• Sony Xperia 1
  5123
• Samsung Galaxy S10e
  4385
• Huawei P30 Pro (perf. mode)
  4215
• Xiaomi Mi 9T
  2182

3DMark SSE Vulkan

Higher is better

• OnePlus 7
  5057
• ZTE nubia Red Magic 3
  4999
• Black Shark 2
  4981
• Sony Xperia 1
  4505
• Samsung Galaxy S10e
  4258
• Huawei P30 Pro (perf. mode)
  4231
• Xiaomi Mi 9T
  2035

The mandatory benchmark disclaimer applies here more than anywhere, though. While these short intensive tests let the phones show what they are capable of without worrying about heat, they aren't always representative about real life performance. After all it's easy for a maker to tune its phone to do well in benchmarks and benchmarks being so popular makers would go out of their way to do so.

Yet there's no cheating the hour-long stress test that would be an intensive gaming session. Running the phone beyond what it can actually sustain will result in very quick heat build up, leaving it no other option but to tune down its chipset so as to prevent irreversable damage.

Heat can be detrimental in a number of ways, ranging from accelerating electrons to an undesirable and uncontrollable high speed and causing computational errors all the way to actually causing physical damage to materials like solder.

This self-destructive behaviour can only be countered by managing heat in some way. Most commonly in electronics that is done through active means, like air or water flow. Since this approach is rarely practical on a smartphone (with the ZTE nubia Red Magic 3 we have here a notable exception) we are typically stuck with passive cooling solutions - heat plates, pipes and the occasional vapour chamber.

Enter thermal throttling - the mechanism to dial back performance by reducing clock speeds and shutting down cores to reduce heat buildup. As we mentioned pretty much every smartphone thermal-throttles eventually when put under pressure. So now we'll check how much pressure exactly is needed before differences start to appear.