-
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" thread (inc AM4/APU discussion) ***
- Thread starter Boomstick777
- Start date
Caporegime
- Joined
- 18 Sep 2009
- Posts
- 30,288
- Location
- Dormanstown.
Ultimately. No one knows, everything depends on context.
I'm hoping for some type of sandy bridge 8 core type chip ultimately.
I'm hoping for some type of sandy bridge 8 core type chip ultimately.
oh please digitimes has been known for a long time as a clickbait site - its how they get most money....They've been caught several times picking up completely wrong stories and run with them as the rest of the web has.
This is coming from an old AMD slide *from March 2015* that had Zen APUs pushed back to 2017; which was their time frame from pretty much the beginning Summit Ridge hasn't been delayed and was show off......
So where is the delay? what's been Delayed? again this looks like damage control
Yes. Interestingly, that graph might suggest that they think they can improve Zen over the coming years by 15-20% in terms of IPC. It's more likely it's just a graphically nice arrow and they didn't put that much thought into its angle.
At least they give the impression they feel its going up. Nobody would be impressed by a 90 degree right arrow
I thought it was already stated to be 40% from Excavator?
The blue bit, yes.
Well, I wasn't posting the graph for the claimed performance (never trust those), I was just trying to find something quick that Zen's improvement is based from Excavator and not Piledriver 
.
@Orangey pretty sure I remember the architecture names being on the original slides.
.@Orangey pretty sure I remember the architecture names being on the original slides.
Like I said before IPC is a meaningless metric of overall performance without knowing the final clockspeed.
IPC stands for "Instructions per cycle", which means that even if Zen has a 40% higher IPC if there are 40% less cycles then it will execute the same number of instructions overall. I'm guessing AMD are using "instructions per clock" to not give the game away so obviously and feed the hype train.
I'd predict Zen to be clocked in the 3.5ghz region (pure speculation on my part) in which case you're looking at closer to 20% overall improvement per core since Excavator cores would be clocked in the 4-4.2ghz range.
IPC stands for "Instructions per cycle", which means that even if Zen has a 40% higher IPC if there are 40% less cycles then it will execute the same number of instructions overall. I'm guessing AMD are using "instructions per clock" to not give the game away so obviously and feed the hype train.
I'd predict Zen to be clocked in the 3.5ghz region (pure speculation on my part) in which case you're looking at closer to 20% overall improvement per core since Excavator cores would be clocked in the 4-4.2ghz range.
Last edited:
Is that speculation or fact?
Source?
I remember a similar thing being said about Polaris 10.
People claiming to have inside information that AMD couldn't get P10 on 14nm LPP above 850Mhz.
Soon debunked by AMD's response and then proof of the RX 480 running at 1266Mhz at Computex.
There seems to be a lot of people claiming to be in the know with insider information making hyperbolic claims, like Kyle from HardCOP for instance, one of the people perpetuating that myth and many others citing "insider information"
You can't trust any of it.
People claiming to have inside information that AMD couldn't get P10 on 14nm LPP above 850Mhz.
Soon debunked by AMD's response and then proof of the RX 480 running at 1266Mhz at Computex.
There seems to be a lot of people claiming to be in the know with insider information making hyperbolic claims, like Kyle from HardCOP for instance, one of the people perpetuating that myth and many others citing "insider information"
You can't trust any of it.
Last edited:
Associate
- Joined
- 23 Apr 2013
- Posts
- 511
- Location
- Gggg
I imagine these people to be the them who were previously able to get information and leak it.
Fortunately AMD are playing a much better game lately to keep people guessing. Exactly how they should be. When AMD want consumers to know, they will.
Last edited:
Associate
- Joined
- 29 May 2014
- Posts
- 402
FX 9xxx at 4.7GHz get 113 Single Core Score in CINEBENCH R15
Ivybridge at 4.7GHz get 172 Single Core Score in CINEBENCH R15
Skylake at 4.7GHz get 204 Single Core Score in CINEBENCH R15
So if we calculate the ipc gains from the AMD graph (excavator 7.4% + zen 40%) then the IPC will almost match the ivybridge.
zen at 4.7GHz(*) will get 170 Single Core Score in CINEBENCH R15
(*) assuming that it can clock at this frequency
FX 9xxx at 4.7GHz get 113 Single Core Score in CINEBENCH R15
Ivybridge at 4.7GHz get 172 Single Core Score in CINEBENCH R15
Skylake at 4.7GHz get 204 Single Core Score in CINEBENCH R15
So if we calculate the ipc gains from the AMD graph (excavator 7.4% + zen 40%) then the IPC will almost match the ivybridge.
zen at 4.7GHz(*) will get 170 Single Core Score in CINEBENCH R15
(*) assuming that it can clock at this frequency
Isn't excavator quite a bit more than 7.4% over Piledriver?
I thought it was more like a 10% increase from Piledriver to Steamroller, and then a further 8% from Steamroller to Excavator.
So did i.
Over the years AMD have been citing performance increases from Piledriver to Steamroller and most recently +20% IPC Excavator.
So now in this chat those past increases are, well, nothing really.
Where was that specific slide used? i have seen a slide like it but without the 7% Bulldozer to Excavator bit on it.