2017's Intel i7 will become 2020's i3 - Comet Lake brings Hyperthreading to Core-i3

Only should have been done even 5 years earlier...

Yeah, Intel certainly missed the boat by a few years on this one.

Would I be right in saying their product stack/s are still affected by numerous security vulnerabilities which are having to be addressed/mitigated by both firmware and software?

(I haven't really been paying much attention to Intel since progressing through all three generations of Ryzen to date)
 
Last edited:
Yes you'd be correct. AMD still does have some insecurities.

That said most ones require physical access to the system. So not the biggest deal. Though the others shouldn't be forgotten. This security stuff only started popping up when AMD released Ryzen since more people started paying attention and wound up finding security holes in everything basically.
 
Good that they are finally waking up although with nothing new to the table is too little to late imho
 
The security holes have nothing to do with AMD's Ryzen release, security has overall become a bigger concern and some researchers decided to look into branch prediction, etc.
Though Intel hasn't had a major overhaul in architecture since Nehalem (and to lesser extent, Sandy) and they've mostly just done minor optimisations, like rely heavily on making branch prediction more intelligent. So as such they got hit harder - not even Ryzen was developed late enough that they would've been aware of Spectre's school of vulnerabilities early enough to specifically design around them.
Physical access isn't relevant, ability to run your code on the machine is. This, at some point, was doable with JavaScript, until browsers mitigated against such attacks with reporting timers less accurately.
And overall the effect of these vulnerabilities on any desktop oriented task is rather minor, database operations got hit the hardest.
 
The security holes have nothing to do with AMD's Ryzen release, security has overall become a bigger concern and some researchers decided to look into branch prediction, etc.

The only point I was making was prior to Ryzen we rarely heard anything. Since Ryzen every couple of months we hear about a new vulnerability. I did not say it's because of Ryzen, just since then we hear more and more about it. It does make sense as researchers will want to test Ryzen. Then that curiosity extends to testing Intel as well. Whether that's true or not is hypothetical and not worth arguing about.
 
Pretty interesting! These new i3s have the potential to become the new standard for gamers who don't plan to be streaming! The pricing will be interesting. No need to get the i5-XX600 or XX600k anymore.
 
i3's used to be HTed right? I'm sure I put a 4th gen i3 into my mums PC and it was dual core HT. They only stopped as they got stuck on their core count and didn't want their customers buying i3's instead of i5's.
Whilst we are on that subject, i'd say the i5 range is the area Intel really need to work on. There's very little reason to buy an i5 over a R5 right now, they are just not priced right. You might get an extra 2% fps in some games but that's it but the way that the market is priced right now, you want more than just gaming performance from your PC.
 
Intel have been milking the market with quad cores for nearly a decade with no real innovation (because of little real competition). I'm astounded that now adding HT to quad cores is even mentionable as part of a strategy to best AMD. Let's be honest, Intel have been left flat on their backs by AMD and appear to have little in the way of a technical response. What we appear to be seeing is Intel leaning on existing market share, marketing, borderline anti-competitive practices and a large war chest (from previously said milking). I honestly think that Intel are just going to buy their way through the next 3 years, hoping to maintain their market share for when they might have some chips that can actually respond to AMD.
 
Back
Top