It's not about I/O or PCIe controller, it's a CPU instruction, an ISA extension of the https://en.wikipedia.org/wiki/Bit_manipulation_instruction_set. In ZEN2 it is apparently emulated in microcode using other instructions which is why it is too slow for any benefit here.
I don't know, the USB bus is pretty terrible for pretty much anything but peripherals. At least in RPi 4 the Ethernet is no longer on USB 2.0...
The proper way to do that would be to get a SoC which natively supports SATA/USB 3.0/Ethernet/PCIe.
But I guess that would increase the price by $20...
Yeah micro-HDMI on something that's gonna be plugged and unplugged often... bad idea. Other missed opportunities: no audio jack (I'm pretty sure there is one on RPi 4), wonky storage (SD cards get corrupted on RPi from time to time which is a pain in the ass) and not using their very own Compute...
I appreciate the thoughtful reply.
To put it in perspective, people were asked to work six extra Saturdays on a seven+ year development cycle. I'm not saying crunch is not an issue in general, but in this case it doesn't seem particularly jarring.
Besides, at the risk of sounding like a...
They will likely keep working on a 0-day patch up until the release.
Still, I think all this reporting about crunch is overblown. People putting in extra work, and being rewarded for it? The horror.
I tried it with my early 3600 bin. Rated bronze by the software. It's running in an ITX-case with a single 120mm AIO so I wanted to optimize for thermals. Worked pretty well:
Voltage down from 1.367 to 1.225V
CCX1/CCX2 down from 4100/4100Mhz to 4000/4000Mhz
CB20 multi score down from 3680 to...
If AMD didn't purchase ATi back then they would be gone today. Their custom SoC business (PS4, Xbox One) is what kept the company afloat in the Bulldozer days.
That's not really true. You can just boot with mitigations=off switch as per kernel-parameters
Nice advertisement. Hint: see who wrote the "article", what is his affiliation, and who publishes the "journal".
It is now in the toolkit of malware writers so why wouldn't they use it.
And trying to shift the blame on researchers is ridiculous, all of these attacks stem from a single decision Intel made about deferring access checks in speculation to chase cheap performance gains and now they are getting...