-
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
haszek;30490816 said:
pretty much, its all BS, it made me laught when 8-pack started talking about how intel spend a lot of money on TIM research and overclocking. And about how they know better when i was talking about using naked Die again.
Then he starts selling an entire line of delidded and re-TIMed cpu's.
Makes me laugh really, especially when someone replaced the TIM on a kabylake part and got 30 degrees better temps.
StarShock;30490367 said:
This is what I was trying to get at in my previous post.
Non X chipsets will go the the advertised "Precision Boost" limit. That is all.
X chipsets, you can run XFR which I think will be the smart choice as you can keep to TDP limits and your voltage will be constantly adapted etc, or no doubt go full manual.
Hopefully all CPU's are unlocked and it's the mobo that matters, but I guess you might need an X CPU too. That's how I read the slide anyway.
EDIT: Will they allow non X / XFR manual overclocking on the cheap CPU's?
APUs had the newest Bulldozer iterations because they abandoned the mainstream platform. They didn't get those iterations first, they just got them exclusively.CAT-THE-FIFTH;30490764 said:
Perhaps, but we know they're due under 9 months after Summit Ridge launches. It doesn't seem like enough time to bust out Zen+ IMO.CAT-THE-FIFTH;30490764 said:
Apparently that 65 Watt TDP is a misprint...
can you imagine? 8 core 16 thread, 65 watts? 95 watts is already very low for a chip like that, Intel's equivalent has a 140 watt TDP tho as does the 10 core so its probably more like 120 watts....
Half the power of Intel's, wouldn't that be a complete opposite of what went before..... no, i don't think so.
can you imagine? 8 core 16 thread, 65 watts? 95 watts is already very low for a chip like that, Intel's equivalent has a 140 watt TDP tho as does the 10 core so its probably more like 120 watts....
Half the power of Intel's, wouldn't that be a complete opposite of what went before..... no, i don't think so.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Personally I dislike the idea that enthusiasts pay more for higher stock clocks, enthusiasts are more likely to overclock behind stock clocks(and any stock boost levels) anyway.
To me a cheap bare bones model with no heatsink and lower clocks but binned for a nice low TDP screams enthusiast, while a higher TDP slightly higher clocks and a way higher cost screams rich idiot buying an overpriced Dell/Alienware computer, less enthusiast, more show off.
So for me the 1700 fits better, why spend the extra paying for those higher stock clocks, when I want to push higher overclock anyway. As for the chipset, the X looks like the one to get anyway and the one most manufacturers will focus on. Only 4 usb 3.1 or above on the next chipset down and less sata ports, no dual pci-e slots.
The real interesting part is, if it's doing 3.7Ghz at 65W(which I'm not convinced off) it suggests either another new stepping, purposefully sending out chips that basically wouldn't make the grade normally or sending out chips with voltage set high on purpose just to throw Intel off the scent a bit. IE send them all out with 1.2v voltage so they hit 95-105W at 3.7Ghz, in reality they ship at 1v and drops them to 65W and suddenly it's a different ballgame.
If that is the case then the real 95W chips with the sense overclocking on will likely be pushing above 4.2Ghz turbo.
If AMD are actually able to get chips out pushing a decent way over 4Ghz without overclocking, that could absolutely make Intel look bad on <5 thread performance.
IE if AMD were at say 3.8Ghz max, then a 4.5Ghz turbo 7700k was going to win single threaded everything and a lot of games and things only using up to 4 threads. A rushed release to add an extra 100Mhz seems pretty nuts, the 7700k is just going to get destroyed, absolutely completely destroyed in >5 threads, >8 threads will be hilariously far behind, so what does 100Mhz gain? TO me I was basically thinking Intel would only rush out to get every last Mhz possible, if they were scared the 7700k might also lose 5< thread benchmarks as well, because maybe AMD are far far closer on clock speeds than people believed and maybe Intel just realised that.
7640k as well, pushing it to 4Ghz base and adding Ht, because if you don't even have single thread advantage, better damn well give it HT so it gets beaten less badly in other benchmarks?
To me a cheap bare bones model with no heatsink and lower clocks but binned for a nice low TDP screams enthusiast, while a higher TDP slightly higher clocks and a way higher cost screams rich idiot buying an overpriced Dell/Alienware computer, less enthusiast, more show off.
So for me the 1700 fits better, why spend the extra paying for those higher stock clocks, when I want to push higher overclock anyway. As for the chipset, the X looks like the one to get anyway and the one most manufacturers will focus on. Only 4 usb 3.1 or above on the next chipset down and less sata ports, no dual pci-e slots.
The real interesting part is, if it's doing 3.7Ghz at 65W(which I'm not convinced off) it suggests either another new stepping, purposefully sending out chips that basically wouldn't make the grade normally or sending out chips with voltage set high on purpose just to throw Intel off the scent a bit. IE send them all out with 1.2v voltage so they hit 95-105W at 3.7Ghz, in reality they ship at 1v and drops them to 65W and suddenly it's a different ballgame.
If that is the case then the real 95W chips with the sense overclocking on will likely be pushing above 4.2Ghz turbo.
If AMD are actually able to get chips out pushing a decent way over 4Ghz without overclocking, that could absolutely make Intel look bad on <5 thread performance.
IE if AMD were at say 3.8Ghz max, then a 4.5Ghz turbo 7700k was going to win single threaded everything and a lot of games and things only using up to 4 threads. A rushed release to add an extra 100Mhz seems pretty nuts, the 7700k is just going to get destroyed, absolutely completely destroyed in >5 threads, >8 threads will be hilariously far behind, so what does 100Mhz gain? TO me I was basically thinking Intel would only rush out to get every last Mhz possible, if they were scared the 7700k might also lose 5< thread benchmarks as well, because maybe AMD are far far closer on clock speeds than people believed and maybe Intel just realised that.
7640k as well, pushing it to 4Ghz base and adding Ht, because if you don't even have single thread advantage, better damn well give it HT so it gets beaten less badly in other benchmarks?
Not all CPU's are equal, i think the idea is pay more for ones that are set aside as having higher quality silicon 'binned' they run cooler, on less volts at higher frequencies.
Basically what EVGA do with the FTW GPU's, what 8Pack does with his 8Pack binned CPU's.
Basically what EVGA do with the FTW GPU's, what 8Pack does with his 8Pack binned CPU's.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
humbug;30491064 said:
Actually AMD have almost always fantastic power compared to Intel, Intel just has always had 1-1.5nodes complete advantage on AMD. So in raw numbers Intel was 'ahead', but when a new node offers half the power per transistor, in reality if AMD made an identical chip but on the same node, AMD would in most years have been whooping Intel in power.
There are still times for instance a FX8350 at 95W on 32nm produces more performance than a 4690k.
Like I said, 65W doesn't seem too likely to me, but Intel is, at least in my opinion, way way behind the power people believe they are. To me Intel has constantly had the worst(for a given node and their technology advantage) power efficiency. At 14nm, the industry has really monumentally caught Intel, they are still behind and Intel will still get to 10nm first, but their 14nm is barely better than GloFo's and 10nm was delayed long enough that they won't beat TSMC/Glofo to it by 2 years as is more normal.
EDIT:- one of the reasons AMD had great power draw compared to Intel when taking in account process nodes is the old story, they had to and Intel didn't. Intel had the process advantage so focused their R&D elsewhere, AMD knew they had a power disadvantage due to process nodes so spent more time finding ways to save power.
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,334
- Location
- Planet Earth
DragonQ;30490967 said:
It happened before that - Llano had the last iteration of the K10 core with the highest IPC,and Piledriver was released with Trinity before Vishera was:
http://www.anandtech.com/show/5831/amd-trinity-review-a10-4600m-a-new-hope
http://www.anandtech.com/show/6396/the-vishera-review-amd-fx8350-fx8320-fx6300-and-fx4300-tested
AMD was demoeing the Piledriver based APUs a few months after the FX8150 was launched.
Now you could be right they stick Zen MK1 with the APUs,but there is precedence for AMD updating the APUs first so it is not that far fetched. It would even make some sense,if AMD is launching them towards the end of the year with Intel starting to move to 10NM designs. A higher IPC Ryzen would mean less pressure on AMD needing to compete through higher clockspeeds alone.
Like I said we will need to wait and see OFC.
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,334
- Location
- Planet Earth
Saw this mentioned on AT forums:
https://forums.anandtech.com/thread...tion-speculation.2498952/page-3#post-38732092
Its from a retailer - interesting how it says how the "guaranteed frequency" is 3.7GHZ,and how the boost frequency is "cooling dependent".
https://forums.anandtech.com/thread...tion-speculation.2498952/page-3#post-38732092
Its from a retailer - interesting how it says how the "guaranteed frequency" is 3.7GHZ,and how the boost frequency is "cooling dependent".
Edit ^^^^^ woooooop getting slow in my old age.
AMD Ryzen 7 1700 Specs Confirmed
This is wccf so digest with lots of salt.
@ CAT
AMD Ryzen 7 1700 Specs Confirmed
This is wccf so digest with lots of salt.
@ CAT
Right, thats the unlimited boost, it will clock its self up and up and.... until it reaches thermal or TDP limits, which ever is first, that could be 4Ghz for one chip with this cooling or 5Ghz with another with that cooling.....
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,334
- Location
- Planet Earth
So it does seem those frequencies are the base clockspeeds??
Edit!!
Lets hope IPC is indeed close enough to BW-E!!
Edit!!
Lets hope IPC is indeed close enough to BW-E!!