Monday, October 19th 2020
AMD Ryzen 5 5600X Benchmarked, Conquers Intel Core i5-10600K
Since AMD announced its next-generation Ryzen 5000 series desktop processors based on Zen 3 core, everyone has been wondering how the new processors perform. For a detailed review and performance numbers, you should wait for official reviews. However, today we have the scores of Ryzen 5 5600X CPU. Thanks to the popular hardware leaker @TUM_APISAK, the Ryzen 5 5600X performance numbers in the SiSoftware Sandra benchmark suite have been leaked. When digging under the hood, the new Ryzen CPU contains six of Zen 3 cores with 12 threads, paired with as much as 32 MB of level three (L3) cache. These cores are running at 3.7 GHz base frequency, while the boost speeds are reaching 4.6 GHz.
In the test results, the AMD Ryzen 5 5600X CPU has scored Processor Arithmetic and Processor Multi-Media scores of 255.22 GOPS and 904.38 Mpix/s. These scores are not much on their own until we compare them to some of the Intel offerings. When compared to the Intel Core i5-10600K CPU, which is likely its targeted competing category, it scores 224.07 GOPS and 662.33 Mpix/s for Processor Arithmetic and Processor Multi-Media tests respectively. This puts the AMD CPU ahead 13.9% and 36.5% in these tests, indicating the possibility of Zen 3. Another important note here is the thermal headroom both of these CPUs run. While the Intel model is constrained withing 125 W TDP, the AMD model runs at just 65 W TDP. This could be an indication of the efficiency that these new processors harness.
Sources:
TUM_APISAK, via Tom's Hardware
In the test results, the AMD Ryzen 5 5600X CPU has scored Processor Arithmetic and Processor Multi-Media scores of 255.22 GOPS and 904.38 Mpix/s. These scores are not much on their own until we compare them to some of the Intel offerings. When compared to the Intel Core i5-10600K CPU, which is likely its targeted competing category, it scores 224.07 GOPS and 662.33 Mpix/s for Processor Arithmetic and Processor Multi-Media tests respectively. This puts the AMD CPU ahead 13.9% and 36.5% in these tests, indicating the possibility of Zen 3. Another important note here is the thermal headroom both of these CPUs run. While the Intel model is constrained withing 125 W TDP, the AMD model runs at just 65 W TDP. This could be an indication of the efficiency that these new processors harness.


53 Comments on AMD Ryzen 5 5600X Benchmarked, Conquers Intel Core i5-10600K
I don't know what glue Intel was sniffing back when they came up with about gluing cores together but the same glue they sniffed back then is clearly not working......
On us. :D
On them it seems to be a different matter.
Cause last time I bought a new Intel system it was 2000... A Pentium3 1.0GHz...
ZEN family has grown step by step to a very nice perf/watt ratio.
The humble R5 3600nonX at 65W was really close (less than 10% behind) to all core relative perf to the 10600K and a 2~10% (depending resolution) behind in gaming.
Comparing the R5s 3600 vs 5600X within the same exact power envelope the latter has +20% IPC =(+20 perf on same freq) and another +20% perf/watt improvement =(around +20%** more freq).
**Probably it wouldnt be that much for allcore boost.
The 5600X could actually be +30% up to +35~40% faster across everything vs the 3600nonX without a single watt more... and that can place it +15 up to +30% above 10600K depending the workload, again... with that 65W TDP.
www.techpowerup.com/review/intel-core-i5-10600k/21.html
FYI the 65W parts of ZEN2 (3600, 3700X) have a total max power draw (sustained) of 88W PPT (Package Power Tracking).
The 10600K has a 125W TDP and a total power draw of
PL1: 125W (sustaind)
PL2: 182W (short term, can be altered by BIOS)
(PL=PowerLevel)
Zen3 needs to do one thing to finally take Intel's crown and that's beat it in real-world gaming performance. That's pretty much the only victory Intel has retained over Zen2 in these last 15 months.
Still, leaks and rumours are one thing; We have less than two weeks of waiting and then all this hypothesising and dubious benchmarks from shady sources can stop!