i9 observations

[henryzz]

Originally Posted by axn

Unlikely that 560K result has anything to do with memory. The i9 is a 20MB L3 part so the whole FFT should be running entirely from L3 and shouldn't be hitting RAM at all. Probably just seeing multi-threading overhead (particularly since it is such a small FFT size).

Upto 20/8 = 2.5M FFT (or slight smaller to account for other overheads) should run entirely within L3.

Is it possible the L3 bandwidth limit is being hit?

[axn]

Originally Posted by henryzz

Is it possible the L3 bandwidth limit is being hit?

Quite possible, now that you mention it.

[Xyzzy]

Here is a chart of timings for a 6144K FFT on the i5-10600K.

The orange is 3200 14-14-14-34 memory and the yellow is 4000 18-19-19-39 memory.

We also tested the 560K FFT but as expected it was the same for both memory configurations.

We have already sold the i9 so we are unable to test it with the faster memory.

[henryzz]

Originally Posted by Xyzzy

Here is a chart of timings for a 6144K FFT on the i5-10600K.

The orange is 3200 14-14-14-34 memory and the yellow is 4000 18-19-19-39 memory.

We also tested the 560K FFT but as expected it was the same for both memory configurations.

We have already sold the i9 so we are unable to test it with the faster memory.

It would be interesting to see how many cores and at what speed provide the maximum efficiency(lowest TDP limit) at the minimum clock speed that reaches 130 iter/sec

[Xyzzy]

We have the 10600K from this thread for sale. $225 shipped (US) seems like a fair price. PM us for more info!