Jim wrote:
If this is as good as it gets, there's obviously a significant underlying issue, as the memory itself should provide much higher bandwidth.

And the Qorlq line's focus on network capability, with it deemphasis of serdes connectivity is problematic.
If you were to use the e5500 or e6500 cores in a CPU designed with end user computing devices, you'd probably want to rework some of this to provide better support for other devices.

Still, from the recent benches I have seen Spectre660 post, the overall performance of the X5000 still measures up fairly well.

[Jan 10 14:55:02 UTC] Automatic processor type detection did not
[Jan 10 14:55:02 UTC] OGR-NG: Running micro-bench to select fastest core...
[Jan 10 14:55:06 UTC] OGR-NG: using core #0 (KOGE 3.1 Scalar).
[Jan 10 14:55:25 UTC] OGR-NG: Benchmark for core #0 (KOGE 3.1 Scalar)
0.00:00:16.75 [19,861,876 nodes/sec]
[Jan 10 14:55:25 UTC] RC5-72: Running micro-bench to select fastest core...
[Jan 10 14:55:49 UTC] RC5-72: using core #2 (KKS 604e).
[Jan 10 14:56:08 UTC] RC5-72: Benchmark for core #2 (KKS 604e)
0.00:00:16.06 [5,574,672 keys/sec]

[Aug 23 16:52:41 UTC] Automatic processor type detection found
a PowerPC 970MP (G5) processor.
[Aug 23 16:52:41 UTC] OGR-NG: using core #1 (KOGE 3.1 Hybrid).
[Aug 23 16:53:00 UTC] OGR-NG: Benchmark for core #1 (KOGE 3.1 Hybrid)
0.00:00:16.29 [42,505,737 nodes/sec]
[Aug 23 16:53:00 UTC] RC5-72: using core #4 (KKS 7450).
[Aug 23 16:53:21 UTC] RC5-72: Benchmark for core #4 (KKS 7450)
0.00:00:17.47 [18,852,434 keys/sec]
(benchmark rates are for a single processor core)