не, на аллай lpddr5, 130нс на рандом доступ. Аида не даёт актуальных цифр на steamos, бутиться в Винду лень.
Intel(R) Memory Latency Checker - v3.13
*** Unable to modify prefetchers (try executing 'modprobe msr')
*** So, enabling random access for latency measurements
Measuring idle latencies for random access (in ns)...
Numa node
Numa node 0
0 131.3
Measuring Peak Injection Memory Bandwidths for the system
Bandwidths are in MB/sec (1 MB/sec = 1,000,000 Bytes/sec)
Using all the threads from each core if Hyper-threading is enabled
Using traffic with the following read-write ratios
ALL Reads : 36202.2
3:1 Reads-Writes : 37228.4
2:1 Reads-Writes : 38463.0
1:1 Reads-Writes : 38010.4
Stream-triad like: 41220.7
Measuring Memory Bandwidths between nodes within system
Bandwidths are in MB/sec (1 MB/sec = 1,000,000 Bytes/sec)
Using all the threads from each core if Hyper-threading is enabled
Using Read-only traffic type
Numa node
Numa node 0
0 36098.7
Measuring Loaded Latencies for the system
Using all the threads from each core if Hyper-threading is enabled
Using Read-only traffic type
Inject Latency Bandwidth
Delay (ns) MB/sec
==========================
00000 1053.39 36318.1
00002 1058.31 36279.7
00008 1093.81 36168.6
00015 1040.09 36255.1
00050 1101.49 36130.1
00100 1064.67 35906.9
00200 211.59 25229.9
00300 189.43 17782.3
00400 175.58 13415.9
00500 165.67 10814.4
00700 155.63 8135.2
01000 147.82 5933.7
01300 145.71 4353.4
01700 146.69 3155.2
02500 145.31 2200.7
03500 147.09 1637.4
05000 146.32 1260.0
09000 146.19 886.3
20000 148.39 628.8
Measuring cache-to-cache transfer latency (in ns)...
Local Socket L2->L2 HIT latency 30.0
Local Socket L2->L2 HITM latency 30.1
таких цифр и близко нет.
Поставил настройки как в видосе.
Видимо упор не в проц, а в графику, 50 ФПС есть.