Caporegime
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Zen = Sandybridge IPC
- Thread starter RavenXXX2
- Start date
Soldato
Too right it is. If you add up all the microde and OS mitigations so far for Intel, it really makes you wonder if any of the "youtube generation" have actually implemented any of the security updates at all. May well explain Chris's stupid IPC results as well.
Caporegime
lol
Associate
- Joined
- 9 Jan 2019
- Posts
- 885
The youtube chappie prob didnt add all the patches, ok he might... i didnt watch the whole video, but i would guess he didnt.
Seen a lot of the "I wont apply the patches as i just play games and dont need/care about them" on forums around the web. Really dumb idea, someone IS going to weaponize this at some point and i have been on the receiving end of far to many clever bits of malware or viruses as well as having to clean up the crap left by them on poorly configured home users computers (and some of our own IT managed ones
) to not take security seriously.That said i will refrain from turning off hyperthreading on my 5820k for the time being as its going in a couple of months when AMD release Zen2 and i dont want to be running something slower than a slug.
Still means i can bring out this again
Soldato
- Joined
- 22 Nov 2018
- Posts
- 2,715
Lets see how many generations it takes AMD to get from Sandybridge IPC to Coffee Lake IPC. Intel took six generations so AMD will probably do it in one or two.
Honestly though for the average home user there is no reason, unless something changes, to disable HT. For those with public facing systems though another matter entirely.
The people that really need to worry are governments and high end industries, etc. that would be the target for orchestrated hacking attacks by groups with significant resources behind them.
Associate
- Joined
- 9 Jan 2019
- Posts
- 885
It took AMD one generation to do it and they did it 2 years ago, unless folk wanna quibble about a couple of percent here of there which is entirely dependant on type of load ect.
Soldato
Ohhhhhhh, another MR OBVIOUS ? So, obviously your post didn't go via a public facing system of any sort. Get real Rroff. Add all the attacks together, as we know at the moment (and there will be more) It will only go down hill and probably even faster than it has already.
According to RedHat etc even a maliciously crafted website, advert or similar can exploit this. I have HT disabled on my desktop, laptop and router (especially as it runs VPN with a tRNG, and I don't fancy my crypto secrets being breached) but I left it enabled on my NAS as it's underpowered enough.
Caporegime
Here he is, gone into full defensive mode.... basically, no i'm right, you are all wrong and "i may have been using a Different version of Cinebench R15 to the Cinebench R15 that Hardware Unboxed used" can this guy really be that clueless? or is it simply that he thinks we are all stupid?
He then goes on to reinforce Intel better Intel better IPC, we all know that
Now i think he's Intel shilling...
He then goes on to reinforce Intel better Intel better IPC, we all know that
Now i think he's Intel shilling...
Last edited:
Soldato
My 4.6GHz 4820K vs my 4GHz 1920X. Gen 1 zen certainly has ivybridge-E beaten even with a 600MHz clock deficit in Cinebench.
Soldato
He's almost as bad as the Verge in being wrong. 
Associate
- Joined
- 9 Jan 2019
- Posts
- 885
My 4.6GHz 4820K vs my 4GHz 1920X
600Mhz advantage and the Ivy bridge system is scoring practically the same as Zen.
I did this one at 3.0Ghz, the same speed as Chris.
343
vs
265
A difference of 30%, anyone with CPU-Z, Ryzen and Cinebench for that matter can easily replicate this.... Chis is 30% down on what he should be scoring at least in CPU-Z, about 20% down in Cinbench, its utter nonsense, i don't know why.
I score even higher with lower than 3.0 GHz clocks:
I didn't think it needed spelling out but by public facing I'm talking about systems that are hosting services people connect to externally (which especially are very exposed where someone has remote login with any kind of executable access).
So far browser script mitigations are robust against these kind of attacks as interpreted languages like this are slow enough anyhow that a broad spectrum approach to security can be taken which can provide protection against both known and unknown variants of Spectre without having to worry about the penalties that kind of approach has at system level - all Spectre variants rely on the same kind of mechanisms to work which are funnelled through a bottleneck in a browser scripting language unlike a scenario where the exploit is being executed on the OS directly.
These attacks really need to be targetted with some idea as to what you are going after to do anything useful for the attacker - a maliciously crafted website would have to be created with a specific purpose in mind and someone browsing (who was specifically a target) without an updated browser that has mitigations - this isn't like fire and forget malware that can infect any and every exposed system it hits.
Caporegime
I score even higher with lower than 3.0 GHz clocks:
Holly crap so you do...
Caporegime
Perhaps this is the start of Intel trying to cripple AMDs reputation with a few underhand wads of cash
That... it just wouldn't surprise me at all. I think we will pretty quickly find those whose results are not right, "using different versions of Cinebench R15" is a red flag, there is no such thing, that's the sort of desperate excuse you'd make for the obvious discrepancies.