Associate
- Joined
- 28 Jun 2017
- Posts
- 102
I've got 3 M.2 slots on my board, and I'm using all of 'em. 970 EVO, 960 EVO, 860 EVO. No 2.5" or 3.5" drives in my machine at all.
-
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.
AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***
- Thread starter Jamin280672
- Start date
I've got 3 M.2 slots on my board, and I'm using all of 'em. 970 EVO, 960 EVO, 860 EVO. No 2.5" or 3.5" drives in my machine at all.
Red gaming tech, who has similar sauces to adoredtv (his sauce confirmed radeon vii about 2 months before release), has confirmed the wccftech news. I think this is on the money, personally.
https://m.youtube.com/watch?v=pismIrcucss
https://m.youtube.com/watch?v=pismIrcucss
Because overclocking on ryzen makes you lose single core ipc performance
Soldato
- Joined
- 22 Nov 2018
- Posts
- 2,840
Or because a stock 2700x boosts to 4.35GHz
I thought that's what I said lol.
The out of the box single core boost is 4.35, when you try to overclock you're not likely to match 4.35 all core, which means a loss of single core performance
Soldato
- Joined
- 26 Aug 2004
- Posts
- 5,155
- Location
- South Wales
The leak says it's the quad core part hitting 4.5 right? The lower core count CPUs are rumoured to have less clocks anyway.
Soldato
- Joined
- 22 Nov 2018
- Posts
- 2,840
The IPC part threw me because that doesn't change when you decrease the clock. Only the performance decreases.
Associate
- Joined
- 9 Jan 2019
- Posts
- 885
These are low core count engineering and sample chips being sent out, i wouldn't bet too much money on 4.5Ghz being the ceiling on the much more interesting parts - your guaranteed to loose it.
I'd be surprised if 4.5 was the max as this early sample, lower clocked quad core(if other leaks are correct) is already at its maximum? Probably not. A 10% rise in single core frequency should get us close to 4.9 GHZ and that's after a drop down to 7nm. 5GHz doesn't seem too crazy to me, we'll soon find out
If we are talking clock speeds at the same total package power, and based on an all core overclock then a 105w 2700X goes to ~4GHz, I'd say as a minimum 10% for all core, and single/dual/quad a reasonable amount higher than that. If the rumors become reality and the 2700X equivalent will become the 3600X for core count and threads, I'd expect it to be in the 95w range with clocks slightly below that of the top of the stack 12c/24t part(s).
I'm very interested to find out about the low power performance of these CPU's, e.g what you can get out of a chip in the 35w configuration on a 6/8 core part. PPW is a significant target for what I do, so enabling a leap ahead in that space would be wonderful.
I'm very interested to find out about the low power performance of these CPU's, e.g what you can get out of a chip in the 35w configuration on a 6/8 core part. PPW is a significant target for what I do, so enabling a leap ahead in that space would be wonderful.
I didn't think it was clear that 4.5ghz was base clock or the boost clock. I can't imagine we would have a 2700x boosting to 4.35ghz and the top boost we get from 7nm being 4.5ghz.
It makes much more sense that 4.5ghz would be the base clock, boosting to somewhere close to 5ghz, with an added 15% uplift on ipc based on zen+
It makes much more sense that 4.5ghz would be the base clock, boosting to somewhere close to 5ghz, with an added 15% uplift on ipc based on zen+
That is single core boost only, 4.35Ghz. If you have 8 cores doing the same thing, trying to stay inside the TDP limit is a totally different kettle of fish. Hence why a manual overclock on the 2700X will only normally get you to 4.2GHz, as you are pushing up all 8 cores, not just one so are using way more current.
A single core boost clock of 4.5ghz for the 3700, 3700x or w/e this 2700x replacement 8 core is called makes perfect sense.
Historically, when new process nodes are introduced - the first attempt is immature. A very common trait of this immaturity is to have product that only have a slight, if any improvement in clock speed over the very mature past process nodes.
So a move from 4.3 to 4.5 is perfectly reasonable given the immaturity of this 7nm process for AMD. The bigger clock speed boosts will come from 7nm+ on Zen 3 and that's more likely to be where we see AMD break the 5ghz barrier
Historically, when new process nodes are introduced - the first attempt is immature. A very common trait of this immaturity is to have product that only have a slight, if any improvement in clock speed over the very mature past process nodes.
So a move from 4.3 to 4.5 is perfectly reasonable given the immaturity of this 7nm process for AMD. The bigger clock speed boosts will come from 7nm+ on Zen 3 and that's more likely to be where we see AMD break the 5ghz barrier
Soldato
- Joined
- 26 Aug 2004
- Posts
- 5,155
- Location
- South Wales
Where do you get that idea? Zen 2 is supposed to be a pretty big jump, Zen 3 just a small step.
I doubt that it would be that low for a single core, 4.7GHz+ is more likely.
Associate
- Joined
- 21 Sep 2018
- Posts
- 895
I read somewhere that Zen3 will use 7nm EUV. Whatever that is - it sounds cool. lol
By that time, DDR5 should be in use and M2s will be cheaper. No need for SATA ports.
Associate
- Joined
- 10 Feb 2011
- Posts
- 1,534
Any real release date yet?. I heard June or July 7th is that true?. Cant wait defo buying a new one when it does come out. I'm jumping ship from intel to AMD.
ideally 4.5ghz stock no stupid voltage or memory needed is what i would personally like and hope for then its price. its going to be interesting to see if its all amd hype or they have caught up the 15 percent ipc they were behind. i swear they said 13 but we will soon know.