• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

Page 52 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
it is impressive that AMD got to broadwell-E.

I wonder how much ipc they can squeeze out of zen+ because intel will now start trying again.
Well, they have obvious (or seemingly obvious?) memory latency issue to address, then there are probably some general refinements to be made to make it clock better, because it's present scaling is outrageous.
From there on it is kind of confusing, because it is fairly wide core already.

Damn, that is a sad perspective.

Clearly he isn't the only one clueless about what a totalitarian regime is.
Khe-Khe.
 
Well, they have obvious (or seemingly obvious?) memory latency issue to address, then there are probably some general refinements to be made to make it clock better, because it's present scaling is outrageous.
From there on it is kind of confusing, because it is fairly wide core already.

Damn, that is a sad perspective.


Khe-Khe.
how on earth you conclude that?
 
That power consumption Platform Power - CPU Full Load 1700x 33.77% better
Prime95?
how on earth you conclude that?
Conclude what? That there is little room to grow left IPC-wise? It is sort of obvious. You can only make a core so wide, and pipeline so long, before diminishing returns make the gains not worth the effort. And well, both Zen and Skylake are a pushing that limit. For frequency, i would not be surprised if AMD managed to find the secret sauce that would make Zen+ clock to 5Ghz on high end water if they really wanted to. But that's about as far physics will allow it to go imho.
 
Anyways, i worry that AMD's attempt to compare 1700 and 7700k will backfire badly, because unlike 1800X/1700X where they are in worst case on par with CPUs AMD compares them to, 1700 may just get roflstomped in badly threaded benchmarks. Though maybe it is their intent to move more X chips instead.
AMD is offering better performance than an i7 5960X at less than half the TDP with per-core overclocking for situations when you need more single-threaded performance at a lower overall platform cost than the i7 7700K. It's a no-brainer.

Don't worry about badly-threaded benchmarks. CSGO at 300fps is plenty enough for me.
 
AMD is offering better performance than an i7 5960X at less than half the TDP with per-core overclocking for situations when you need more single-threaded performance at a lower overall platform cost than the i7 7700K. It's a no-brainer.
But that's the damn point: it (the 1700) won't beat stock 7700k in every bench that uses less than 5 threads [actively]. And that's what AMD positions it against, ultimately. Comparisons with 6800k and 6900k work brilliantly because AMD basically ties Intel there in the worst case. 7700k? Not so much. Not that it will change my decision, because my stuff can thread reasonably well.
 
Btw, either translation fails in that table, or numbers are straight BS. 480+"95W TDP" CPU producing a 90W power consumption delta? BS.
 
Last edited:
From what I translated from that chinese post: http://pic.yupoo.com/ztwss/Gg1Rl9iq/12exqY.png

the 1700X which is 8/16 is drawing less power than the i7 6800K...

For gaming for example, it says 1700X is doing 154.66 watts while the i7 is doing 194.2 watts.

EDIT: For Office related work it says 81.55 while i7 is doing 113.5. WTF

EDIT2 Idle is 62.77 for 1700x, 98.74 for the i7
Full Load is 123 AMD, 126.87 i7
 
Last edited:
From what I translated from that chinese post: http://pic.yupoo.com/ztwss/Gg1Rl9iq/12exqY.png

the 1700X which is 8/16 is drawing less power than the i7 6800K...

For gaming for example, it says 1700X is doing 154.66 watts while the i7 is doing 194.2 watts.

EDIT: For Office related work it says 81.55 while i7 is doing 113.5. WTF
These numbers have to be BS. Like, to the point where they cast doubt on legitimacy of the other numbers. Also, 1700X test was done on QS, apparently, because it happened on Feb 10th.


EDIT: Another fun part, apparently one of benches is fritz and this QS gets beat by 5960X on lower frequency in it. Another thing to work in future Zen iterations, ha.
 
But that's the damn point: it (the 1700) won't beat stock 7700k in every bench that uses less than 5 threads [actively]. And that's what AMD positions it against, ultimately. Comparisons with 6800k and 6900k work brilliantly because AMD basically ties Intel there in the worst case. 7700k? Not so much. Not that it will change my decision, because my stuff can thread reasonably well.
As long as it beats it in overall application performance and gives 90% of the performance of the i7 7700K in the less multithreaded games, it will still be OK.

As long as it gives 300fps in CSGO compared to 350 for the i7 7700K, the average person won't care if the overall performance is higher, which it will be.
 
according to Data https://ibb.co/bCxkwF
I translated :
平台功耗-CPU满载(越小越好) : Platform power consumption - CPU full load (the smaller the better)
平台功耗-待机(越小越好) : Platform power consumption - standby (the smaller the better)
平台功耗-游戏(越小越好) : Platform Power - Game (the smaller the better)
平台功耗-办公(越小越好) : Platform power consumption - office (the smaller the better)

so summary :
123 vs 126.87
62.77 vs 98.74
154.66 vs 194.2
81.55 vs 113.5
 
All I can tell you is what I read from that chinese post, not drawing any conclusions from it yet. Also again, DDR4 2133mhz RAM for the 1700X. So those scores might go up a tad when the bios get fixed, if these are legit.
 
As long as it beats it in overall application performance and gives 90% of the performance of the i7 7700K in the less multithreaded games, it will still be OK.
The 90% part is the one i worry about. Because it is a stock clock of 3-3.7 against 4.4-4.5. 4.5/3.7=20% faster single thread + whatever IPC advantage gives and that only gets wider with more cores because of dropping boost until 1700's core advantage kicks in.

so summary :
123 vs 126.87
62.77 vs 98.74
154.66 vs 194.2
81.55 vs 113.5
Actually, looks like the OP of that thread doubts that data himself (probably not his to begin with).
 
Yeah, he claimed it was sent to him by someone if my chinese is right. Could always google translate it, but it messes up the sentences horribly.
Eh, Google translate is so popular for chinese translations that they have actually learned to do it somewhat decently with their machine learning stuff.
 
Eh, no. It's monopolistic exploitation of a dominant market position. You understand that the next step after this is massive price hikes (due to there not being any competition any longer), right? Let me guess, you're a "free market economy" guy? I'd love to see you read up on some of the thinkers that laid the foundation for modern capitalism (you know, the people near deified by the American right). They were extremely in favor of protecting markets from monopolies, as unregulated markets always tend toward unfair competition.

Of course, you're free to believe that corporations should be free to exploit customers and force competitors out of the market. I just think that's bonkers, and a recipe for a truly f*cked up society.

Milton Friedman was of the view that all monopolies were caused by government's actions, except for De Beers.

Truly unregulated markets are nonexistent in the West currently. Do you have a unregulated market that has a monopoly where consumers are hurt?

PM me if you have anything.
 
http://valid.x86.fr/bench/rjmzdu/1
http://valid.x86.fr/bench/rjmzdu/16

What kind of clocks would you need for this? 4.7-4.8Ghz-ish? Looks like a dry ice / LN2 run, yes?

For reference, Kaby on water 5.0.

Xlj10zA.png
 
Last edited:
If this is true, Intel must be worried.

- Too much room for interpretation, it might just be guidelines for how to review the Intel products.. On the other hand *If* they're instructing how to setup and bench the AMD rig then its hammer and clubbing time. (that would be like ... russia toppling the us elections... It would never stand. j/k.)
 
Back
Top