Cascade lake Xeons are vulnerable to Zombie Load... what a good beginning
At least it seems they will bring some improvement over Skylake.
https://www.hpcwire.com/2019/05/13/...nt-gen-on-gen-advantage-on-stac-benchmarking/
Source? That article mentions nothing specifically about Cascade Lake's status.
Cascade Lake SP isn't....
https://arstechnica.com/gadgets/201...leaks-data-from-intel-chips-internal-buffers/
Some Meltdown/Spectre stuff is fixed in hardware and some stuff isn't. (there is a pretty clear table in this article).
https://www.anandtech.com/show/1330...e-intel-clarifies-whiskey-lake-and-amber-lake
These Microarchitectural Data Sampling ( MDS ) variants that popped up in this recent bunch is mainly variations on the same mechanism. The stuff associated with loading which were part of the root cause of Meltdown varianta 3 and 5 doesn't need individual fixes for each exploit variant layered on the same foundation. If they remove the base root cause then numerous of these variants all disappear.
Intel hasn't fixed everything possible in hardware, but there is very substantive fixes in Cascade Lake (and newer Mainstram processors. ). In the context of what is deployed already, that doesn't help much. But in terms of new products through the rest of 2019 the "danger" level is lower.
Searching the Internet found a copy of the paper.
https://arxiv.org/pdf/1905.05726.pdf
Table 2 on page 8.
Most the of "whip it out fast ... instead of correct" tech/rumors press can't wait to whip out the articles.
"ZombieLoad" in Cascade is more salacious clickbait than fact.