Unless you are referring to FPSVR benchmarks of DCS World with missions containing hundreds to thousands of objects, you missed my point.
Synthetic benchmarks do not reflect poorly optimized application performance.
I am talking game benchmarks. Not synthetic benches. However you already had this discussion long ago and we know your opinion is set on your preference.
https://www.userbenchmark.com/EFps/,,3700X.OC,_,,9600K.OC,_Fortnite,2070S,,
Game benchmarks. Seems I am not the only person still assessing a significant penalty to AMD.
Nah you need to see a wide range of games vs picking ones that favor 1 or the other.
@mirage335 You are correct that disabling HT can enable a higher overclock. Without changing anything else, it now runs nicely at 5.1ghz (up .1). That frequency was a hard fail with HT on.
It’s only passed a few tests, short game sessions and a 1 hour stress test, but it’s promising so far. It’s only a small boost, but it’s another win on top of a win - and DCS needs every little bit! 
1 Like
Afaik not. There are 2 cores and it is possible to execute instructions in parallel. However not everything is duplicated, some things are shared between cores which means not every instruction can be executed in parallel and it also leads to the known vulnerabilities (for example when the cache memory is shared so the data can be potentially accessed by the application running on the other core).
So what is making 1 core virtual? Vs real physical one.
Ok seems you are right that it is actually single physical core, but it seems that it can execute (some) instructions from different threads in parallel (e.g. it is not just multitasking). They mention simultaneous execution on Intel webpage and also wikipedia mentions that:
“it takes advantage of [superscalar] architecture, in which multiple instructions operate on separate data [in parallel].”
But I don’t know any technical details.
2 Likes
Not entirely sure either how they do just a guess likely something similar to process slicing time sharing.
1 Like
On that i7 7700K, can you please share:
CPU at 5000
Cache at ?
cVoltage at ?
AVX offset at ?
I have the same CPU and I am very interested in that. Mine is delidded already, I want to see how it compares
1 Like
It’s more complicated than that. Keeping it simple: A hyperthreaded core has some duplicated parts and some shared parts. 2 arithmetic (or logic, etc.) instructions can indeed be calculated simultaneously. That is, the 2 threads can both be active and running on a single HT core at the same time.
The reason HT can cause slowdowns is because of contention for the shared parts, memory bandwidth limitations, and it can possibly cause thermal throttling.
4 Likes
Thank you that makes it more clear. So items like the quantity & quality of the cache would play into that.
1 Like
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.