Overclock.net banner
1841 - 1857 of 1857 Posts

·
Premium Member
Joined
·
20,851 Posts
Is this program still a good way to detect if my computer is safe from all these cpu bugs?
Only time you really need to be concerned is if you have any sensitive data in your pc. If just for gaming, make sure you have the latest update from MS and you are set.
 

·
Registered
Joined
·
4,964 Posts
They are patched for current bugs and had redesign to restore the lost performance. However, they are still based on the core architecture (but heavily modified in some areas). There are bound to be other holes popping up.

My feeling is intel won't be safe till Ocean Cove (desktop) or laptop variant comes out. Safest bet is AMD for now though these security researchers will find exploits given an couple of years of being able to dig around the code.
 

·
professional curmudgeon
Joined
·
10,585 Posts
I guess almost anyone would do online e-banking with computer....
hold my beer.



i have a gaming machine, not a banking machine.

maybe i'm too old fashion but i go to the bank.

get off my lawn. :p
 

·
Not a linux lobbyist
Joined
·
2,849 Posts
I do my banking on a different computer.
My gaming computer is so far from a microcode update it's microcode isn't even detected :p
 

Attachments

·
Premium Member
Joined
·
6,933 Posts
That could be extremely bad for the financial trading sector, as they rely on RDMA to do their work. They also require internet connections to their computer clusters, so no option to turn it off or secure their network from the outside. Not a single company is going to turn the feature off and take a guaranteed loss in millions of dollars profit against the threat of a possible loss of profit.
I didnt see whether this required local access or not though. They talk about the attack, but no mention of whether the attack can be executed remotely against the target. Hopefully not as that would solve much of the problem.
 

·
Banned
Joined
·
8,589 Posts
That's the same exploit:

https://arstechnica.com/information...s-researchers-steal-encrypted-ssh-keystrokes/

The researchers have named their attack NetCAT, short for Network Cache ATtack. Their research is prompting an advisory for Intel that effectively recommends turning off either DDIO or RDMA in untrusted networks. The researchers say future attacks may be able to steal other types of data, possibly even when RDMA isn't enabled. They are also advising hardware makers do a better job of securing microarchitectural enhancements before putting them into billions of real-world servers.
 

·
Old to Overclock.net
Joined
·
216 Posts
Another issue with these vulnerabilities especially with those with private clouds is the risk of data breach from one container to another. One system could be exposed to external networks and can access data from a different container that would be for internal use only. IDK why people say that they are basically unaffected by this. If only you know how many CIOs and CTOs are scratching their heads thinking how many more vulnerabilities would be coming up. Especially with fixes with performance impact, client facing companies rely on their capacity planning for purchases which are done in advance. You can't just order a server like a laptop and get it delivered next day. With Epyc being cheaper and better performing, IDK how the long term sales projections are.
 

·
Registered
Joined
·
108 Posts
Well since half these bugs are due to the same security check flaw in the cache, yes Ryzen is safer than the current and older gen Intels. And Intel is fixing it, and they they will be the same.
I doubt Intel will fix it in hardware design in years since it would hurt its performance a lot and that would hurt sales number a lot.
 

·
Amatuer Specialist
Joined
·
186 Posts
I doubt Intel will fix it in hardware design in years since it would hurt its performance a lot and that would hurt sales number a lot.
They already have implemented several hardware fixes. The whole point of them is to mitigate the performance loss of software workarounds. For instance, on the 9900K it brings back the theoretical 4KQ1T1 read of an Optane 900 series SSD back up to ~300 MB/s from ~200 MB/s on patched Skylake-X.

(Note these are performance differences you see in benchmarks. Such large gains are rarely mirrored in actual application performance.)
 

·
Banned
Joined
·
8,589 Posts
I doubt Intel will fix it in hardware design in years since it would hurt its performance a lot and that would hurt sales number a lot.
They already have implemented several hardware fixes. The whole point of them is to mitigate the performance loss of software workarounds. For instance, on the 9900K it brings back the theoretical 4KQ1T1 read of an Optane 900 series SSD back up to ~300 MB/s from ~200 MB/s on patched Skylake-X.

(Note these are performance differences you see in benchmarks. Such large gains are rarely mirrored in actual application performance.)
They already implemented the hardware fixes in Ice Lake which is 18% faster than Skylake and 9% faster than Zen 2. My next notebook will have Ice Lake.

https://www.anandtech.com/show/14664/testing-intel-ice-lake-10nm/3
 
1841 - 1857 of 1857 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top