Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

[CatMerc]

https://twitter.com/CPCHardware/status/836214212301053952

 

[cytg111]

redacted

 

[antihelten]

Who said price cuts ??

https://hardware.slashdot.org/story...y-cutting-prices-on-core-i7-and-i5-processors

Thats some hefty numbers is true

There's nothing particularly new about those prices when talking about Microcenter, these aren't even the lowest they've been, back in november they had a sale for the 6800K at only $330 ($30 lower than the prices in the article).

If these prices start showing up at other retailers and if they actually stick, then we might be looking at something notable.

 

[lolfail9001]

Thats some hefty numbers is true

Uhm, what hefty numbers? I see a mind-boggling $7 cut on Amazon, and that's about it.

Microcenter prices are nothing new either.

 

[KTE]

Like I said a month back. Price cuts would be excellent, and they will definitely come fast if Ryzen is challenging.

And do stay aware that Intel will certainly know of Ryzen's performance and what to fully expect by now.

It's just the way business works. Vulture Culture.

 

[inf64]

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.

How much 1700X scores in Fritz according to this Chinese leak? Is it 21386pts? I found 6800K results on hardware.fr and it gets 19576pts.

 

[PhonakV30]

21386 vs 20151 = 6.13%
Fritz , It's above Cinebench R15 ST score.

 

[Det0x]

http://valid.x86.fr/bench/rjmzdu/1

CPU-Z benchmark found it here for 1700X: https://www.reddit.com/r/Amd/comments/5wg020/1700x_single_multithreaded_cpuz_benchmark_valid/

Of course there's zero way to tell if its overclocked (probably is) and with what. Interesting nevertheless.

For comparison: i am getting 2004/17819 with a 5960x @ 4.4ghz core, 32gigs 2667mhz CL15 and 4ghz uncore. (24/7 clocks)

vs this 1700x rysen @ 2327/19874

 

[asendra]

Has this been posted already? Just saw this over at Neogaf..
https://videocardz.com/66451/amd-ryzen-rumors-part-3

Edit: Yep, sorry. Already posted in the page before!

 

[lolfail9001]

Has this been posted already? Just saw this over Neogaf..

Simple look at power consumption numbers should increase salt levels in blood to lethal levels.

 

[vissarix]

Like I said a month back. Price cuts would be excellent, and they will definitely come fast if Ryzen is challenging.

And do stay aware that Intel will certainly know of Ryzen's performance and what to fully expect by now.

It's just the way business works. Vulture Culture.

Intel definitely knows exactly how Ryzen performs...

Just remember when Nvidia launched the gtx 980ti 1 month before the fury x release...Amd was showing to us fury x beating a titan x...yet Nvidia knew exactly how fury x was going to perform so they released a cut down gtx 980ti just to beat the fury x and dictating the price...

Intel doesnt need to lower prices right now since Ryzen is not out yet...obviously when it launches and if it is fast then they will have to lower prices...

 

[lolfail9001]

Yep, i just have seen a confirmation that XFR's boost over stock boost bins... is 100Mhz on water cooling. Color me unimpressed.

Spoiler

Catch it while you can.
https://www.youtube.com/watch?v=1h9-MH9pfkw&t=1020s

 

[french toast]

Yep, i just have seen a confirmation that XFR's boost over stock boost bins... is 100Mhz on water cooling. Color me unimpressed.

Spoiler

Catch it while you can.
https://www.youtube.com/watch?v=1h9-MH9pfkw&t=1020s

Er, didn't we know this already? Not much of a disappointment if its not much of a surprise.

 

[lolfail9001]

Er, didn't we know this already? Not much of a disappointment if its not much of a surprise.

There was some speculation that 100Mhz was only air related. That slaps the final nail in the coffin.

Though i won't deny, thinking about reaction people there had when AMD went like: "We put like 100 sensors and stuff all over the chip, all to reward the consumer for better cooling". Then they proceed to slap a water cooling on 95W TDP soldered 200mm^2 die... and get 100 more Mhz. That had to be anti-climatic. Not to mention the live feed crashing out when they ran all cores at 4.1Ghz.

 

[inf64]

Yep, i just have seen a confirmation that XFR's boost over stock boost bins... is 100Mhz on water cooling. Color me unimpressed.

Spoiler

Catch it while you can.
https://www.youtube.com/watch?v=1h9-MH9pfkw&t=1020s

Yeah but that means that if ACT for 1800X is 3.7Ghz, XFR makes it 3.8Ghz, right? That is ~5% more performance when all cores are loaded , albeit not in *all* workloads. Also that makes ST max. Turbo 4.1Ghz, which is some 10% less than ST Turbo of 7770K.

 

[lolfail9001]

Yeah but that means that if ACT for 1800X is 3.7Ghz, XFR makes it 3.8Ghz, right? That is ~5% more performance when all cores are loaded , albeit not in *all* workloads. Also that makes ST max. Turbo 4.1Ghz, which is some 10% less than ST Turbo of 7770K.

Nope, ACT for 1800X is 3.6Ghz and XFR makes it 3.7Ghz. And sure, it is a nice "free" boost, but i would rather get all 8 cores stable at 4.1Ghz they tried to do, if i am to slap a water cooling on "95W" CPU.

 

[unseenmorbidity]

Nope, ACT for 1800X is 3.6Ghz and XFR makes it 3.7Ghz. And sure, it is a nice "free" boost, but i would rather get all 8 cores stable at 4.1Ghz they tried to do, if i am to slap a water cooling on "95W" CPU.

source of act?

 

[french toast]

There was some speculation that 100Mhz was only air related. That slaps the final nail in the coffin.

Though i won't deny, thinking about reaction people there had when AMD went like: "We put like 100 sensors and stuff all over the chip, all to reward the consumer for better cooling". Then they proceed to slap a water cooling on 95W TDP soldered 200mm^2 die... and get 100 more Mhz. That had to be anti-climatic. Not to mention the live feed crashing out when they ran all cores at 4.1Ghz.

Yea around christmas time, its a little disappointing from back then yes, but that the immature 14nmlpp process for you.

 

[LTC8K6]

Nope, ACT for 1800X is 3.6Ghz and XFR makes it 3.7Ghz. And sure, it is a nice "free" boost, but i would rather get all 8 cores stable at 4.1Ghz they tried to do, if i am to slap a water cooling on "95W" CPU.

Sounds like sandbagging with only 4.1Ghz...even on one core.

 

[tamz_msc]

I thought 3.7 was ACT with 4.0 being the SCT with XFR adding 100MHz on top of SCT, with potentially more depending on better cooling? Never heard any mention of XFR being available for ACT as well.

 

[LTC8K6]

I guess I was expecting more. I predicted 100-200mhz for XFR, but I thought manual would go higher, but they only tried 4.1 manually.

 

[tamz_msc]

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.

In-game fps shows wildly varying scaling with clockspeeds.

 

[french toast]

In-game fps shows wildly varying scaling with clockspeeds.

The one game we have seen that pits 1700 vs 7700k shows same average fps.

 

[thigobr]

If XFR is indeed unlimited and only based on sensors data this means that the silicon is at its limits when clocking to 4.1GHz. In this case would be better if XFR were limited to 100MHz boost so we could hope for better manual overclocking...

 

[french toast]

If XFR is indeed unlimited and only based on sensors data this means that the silicon is at its limits when clocking to 4.1GHz. In this case would be better if XFR were limited to 100MHz boost so we could hope for better manual overclocking...

we already have pre overclocked towers @ 4.2ghz