The Sandy Bridge Core i7 3960X Benchmarked Against Today's Six-Core / 12 Thread AMD/Intel CPUs

[coder]

Btw: that nice command is inverted. That's giving the process a priority of +12 (lower priority) instead of -12 (higher priority) which I'm sure is what you wanted.

I find "nice -n -12" to be easier to remember and more intuitive than the double hyphen (nice --12) and prevents accidentically using the positive number instead of negative.

With normal user privileges, you can't usually use a negative niceness.

[coder]

I agree that mitigations=off is good for notebook&desktop machines (assuming the user ensures that the machine never runs malicious code).

Yeah, as long as you don't run any web browsers on the box. Otherwise, you can certainly run malicious code, in the form of javascript + web assembly.

[Michael]

Thanks. Does PTS log raw dmidecode output?

Yes it parses dmidecode to report more detailed RAM information, when run as root.... Sadly dmidecode is of no use when running as normal user, sure wish there could be improvements to that area for Linux.

[coder]

Quad channel, don't remember the speed offhand but would be whatever the optimal rated configuration was.

Thanks. Does PTS log raw dmidecode output?

[coder]

• If a machine feels unresponsive then the user should primarily take a look at how many tasks are competing for the available hardware threads and should check whether the peak number of running tasks exceeds the number of HW threads. A desktop environment indicator such as xfce4-systemload-plugin can help identify cases of high CPU utilization. The scheduling priority of build jobs (Makefiles, etc) should be kept lower than the scheduling priority of GUI applications.

One issue I've run into is libraries that create their own thread pools. It's becoming more common, too. It'd be nice if there was some unified way for all of the libraries in an app to negotiate how many threads each should launch. Or, better yet, to have each be capable of submitting work to a shared thread pool.

[Michael]

What's the memory configuration of the 3960X - did you use dual-channel or quad-channel? What speed?

Also, the article doesn't mention that the 3 newer systems used a NVMe SSD, while the Sandy Bridge used SATA. That said, the NVMe drive was one of the slower models out there, but IMO it should be noted.

Thanks.

Quad channel, don't remember the speed offhand but would be whatever the optimal rated configuration was.

[Michael]

What's the memory configuration of the 3960X - did you use dual-channel or quad-channel? What speed?

Also, the article doesn't mention that the 3 newer systems used a NVMe SSD, while the Sandy Bridge used SATA. That said, the NVMe drive was one of the slower models out there, but IMO it should be noted.

Thanks.

The drives are mentioned in the system article, but anyhow the workloads tested sans like the Linux kernel build test were all CPU focused so shouldn't be touching the drives in any meaningful manner.

[coder]

Phoronix: The Sandy Bridge Core i7 3960X Benchmarked Against Today's Six-Core / 12 Thread AMD/Intel CPUs

What's the memory configuration of the 3960X - did you use dual-channel or quad-channel? What speed?

Also, the article doesn't mention that the 3 newer systems used a NVMe SSD, while the Sandy Bridge used SATA. That said, the NVMe drive was one of the slower models out there, but IMO it should be noted.

Thanks.

[perpetually high]

It isn't inverted. Build tasks aren't interactive applications and so giving them higher priority seems contradictory to me. I don't want build tasks to run with higher priority than Chrome/Firefox for example.

Gotcha, my bad. Yeah that makes sense. I thought in the context of what we were talking about you were trying to squeeze latency/priority but I misread.

[atomsymbol]

Btw: that nice command is inverted. That's giving the process a priority of +12 (lower priority) instead of -12 (higher priority) which I'm sure is what you wanted.

It isn't inverted. Build tasks aren't interactive applications and so giving them higher priority seems contradictory to me. I don't want build tasks to run with higher priority than Chrome/Firefox for example.