The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.
← Previous pageNext page →

    AMD's Ryzen CPUs (Ryzen/TR/Epyc) & Vega/Polaris/Navi GPUs

    Discussion in 'Hardware Components and Aftermarket Upgrades' started by Rage Set, Dec 14, 2016.

  1. Talon

    Talon Notebook Virtuoso

    Reputations:
    1,482
    Messages:
    3,519
    Likes Received:
    4,694
    Trophy Points:
    331
    But 4.3Ghz doesn't seem typical at all for most users and reviewers so far. Most top out at 4.2Ghz around 1.4+v. Those voltages are not safe for long term usage, and certainly not good for users on the B series motherboards. In addition if 4.3Ghz was max clock he could get to pass CB alone, nothing says 4.3Ghz is long term stable even at those crazy voltages. Still a far cry from the 4.4 and 4.6Ghz you were trying to say was possible.
     
    Ashtrix, Vistar Shook and Papusan like this.
  2. Reciever

    Reciever D! For Dragon!

    Reputations:
    1,525
    Messages:
    5,340
    Likes Received:
    4,298
    Trophy Points:
    431
    I wouldnt know, I dont spend much time in those forums of late

    @Deks might have a better idea since hes got first hand experience with his machine and I think has done tests for voltage needed for certain clock rates.

    I wish that laptop came with 8GB vRAM, I hope future iterations have that and I will pinch my pennies to get one
     
    Vistar Shook and hmscott like this.
  3. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    RashadQuickTech 's channel seems legit, a history of build and review videos, he answers people in the comments, including the recent 4.4ghz and 4.6ghz videos, so until he wakes up and see's all the butt-hurt Intel fanboy posts in his 4.6ghz video and realize's he made a typo - 4.6ghz instead of 4.4ghz - that's all I can think of other than it's legit 4.6ghz results right now.

    Someone's gotta get the Golden Ticket CPU's, too bad it wasn't you - then they could dump all over your joyful luck. ;)

    If this were an Intel CPU I would look at it the same way, just a lucky guy getting a Golden CPU... relax, this is supposed to be fun.

    That's all I have to say on this until RashadQuickTech posts an update to his video's or posts more video's. If you have a problem with his video's post in his comments, his Facebook, or his Google+ - he lists them on his Channel page.

    I don't imagine Rashad is reading this thread, so you're wasting gas posting about it here. :cool:
     
    Last edited: Apr 19, 2018
  4. Robbo99999

    Robbo99999 Notebook Prophet

    Reputations:
    4,346
    Messages:
    6,824
    Likes Received:
    6,112
    Trophy Points:
    681
    I've just read the Guru3d review of the 2700x:
    http://www.guru3d.com/articles-pages/amd-ryzen-7-2700x-review,1.html

    My general impressions from that review are that this new generation of Ryzen CPUs bring a little bit extra performance to the table in comparison to previous Ryzen, but it's not a big difference that is worthy of an upgrade from one to the other, but that was never to be expected. Intel i7 CPUs are still clearly the best CPUs to get if you're doing high refresh rate gaming at the 120/144Hz+ sector - the fact that the gaming comparison tables there were showing stock Intel vs overclocked Ryzen, and Intel were still winning, you can add another 15% to the 8700K if you overclock it to 5Ghz, and then the performance differences are even larger. But, still a great time to buy a Ryzen based PC for all PC needs including gaming as long as you've got a standard 60Hz/75Hz/(maybe 100Hz) monitor.

    Raw CPU power of the Ryzen CPUs is impressive, and if you do anything where you have the CPU sit at 100% load for long periods of time, then they're probably a better buy than Intel, certainly when factoring price/performance, but that all depends on how precious your time is in that scenario.
     
    Vistar Shook and Talon like this.
  5. Reciever

    Reciever D! For Dragon!

    Reputations:
    1,525
    Messages:
    5,340
    Likes Received:
    4,298
    Trophy Points:
    431
    My 4930mx at 3.5Ghz pushed R6: Siege at 120fps so I am not sure how a Ryzen CPU wouldnt be capable of that @4 Ghz +/- 300Mhz

    Also regarding the 4.6Ghz claim, it could just well be a silicon lottery winner, maybe not something to expect of all chips but we wont know that until consumers get their hands on them. Reviewers have to pump out content daily and overclocking is a vastly labor intensive task.
     
    hmscott likes this.
  6. Talon

    Talon Notebook Virtuoso

    Reputations:
    1,482
    Messages:
    3,519
    Likes Received:
    4,694
    Trophy Points:
    331
    Trust me bud I got a great 8700K, well above average and I am far from upset with my beast. This nothing to do with people being butt hurt but rather people spreading fake stuff. Putting 4.6Ghz in the title of your video and playing techno beat tunes doesn't prove diddly squat. Show real time information, on screen displays, put up or shut up. Until this youtuber shows that information in real time, it's fake news.
     
    Vistar Shook and Papusan like this.
  7. Reciever

    Reciever D! For Dragon!

    Reputations:
    1,525
    Messages:
    5,340
    Likes Received:
    4,298
    Trophy Points:
    431
    You'll just have to wait for the consumers to get their hands on them.
     
    hmscott likes this.
  8. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    My prediction that the 2600x and 2700x will see little to no upgrade money seems to hold true. All it is good for is a slight evolution of the Ryzen line. My fear is since GF was promising 10% better from the silicon itself and is not delivering that the same is true of 7nm. The fear here too is the market sees the same and is loosing faith in AMD, bringing on the short money that is trying to kill AMD. I would hate too see AMD and Ryzen die off because of under delivered promises.
     
    Vistar Shook, Papusan and Talon like this.
  9. Reciever

    Reciever D! For Dragon!

    Reputations:
    1,525
    Messages:
    5,340
    Likes Received:
    4,298
    Trophy Points:
    431
    undelivered promises are basically an industry standard at this point.

    I only ever saw this as a refresh, slight improvements in manufacturing that likely allows slightly more headroom than previous product lineup.

    Though I am also the type that doesnt read into marketing, I remain ignorant until I see what the users on this or other forums I may surf unless I stumble across a heated argument about a future product.
     
    Vistar Shook, hmscott and jaybee83 like this.
  10. Deks

    Deks Notebook Prophet

    Reputations:
    1,272
    Messages:
    5,201
    Likes Received:
    2,073
    Trophy Points:
    331
    The 1700 in my GL702ZC has a TDP rated at 65W.
    As such, the existing cooling in the laptop simply wouldn't be enough for 2700x because it appears that AMD decided to still use a manuf. process suited for lower clocks and mobile parts (the 12nm LP is not based on IBM's high performance process I'm afraid)... hence, overclocked Ryzen 1 with some architectural improvements (tightened latencies and better RAM support) which is what Ryzen2 is, would inevitably result in higher power consumption and heat emissions.
    The cooling in GL702ZC can barely handle the 1700... it is highly unlikely that it could handle the 2700X.

    If Asus makes an uprated version of GL702ZC with 2700x... then they will need to revamp the cooling entirely (which is quite honestly ridiculous, seeing how a 17" chassis like this one should have more than enough cooling to handle a 105W TDP CPU.
     
  11. Reciever

    Reciever D! For Dragon!

    Reputations:
    1,525
    Messages:
    5,340
    Likes Received:
    4,298
    Trophy Points:
    431
    Thanks for sharing

    Did you guys use liquid metal at all? I know you personally were be conservative with your machine because it is under warranty but maybe other members may have tried it out in that owners lounge.
     
    hmscott likes this.
  12. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,701
    Messages:
    29,839
    Likes Received:
    59,615
    Trophy Points:
    931
    Then it's confirmed. A bad overclocker as I have said several times in the thread. Aka as expected. And we shouldn't listening on rumors.... The facts is here today!! We had them already long time ago. Nothing new here :cool:

    https://www.tweaktown.com/reviews/8602/amd-ryzen-7-2700x-5-2600x-review/index10.html
    "Our 2700X was able to maintain 4.25GHz while our 2600X was able to do 4.275GHz since it has a lower TDP and fewer cores (lower temperature)."

    https://www.techpowerup.com/reviews/AMD/Ryzen_7_2700X/19.html
    We set a voltage of 1.4 V, which is in line with what a good air cooler can handle and increased frequencies step by step. Maximum stable overclock ended up at 4.2 GHz. Since the Ryzen 2700X boosts up to 4.3 GHz out of the box, which is higher than our manual overclock, many, especially low-threaded benchmarks show a performance loss after overclocking. This is expected, due to the relatively small OC and AMD's good use of the processor's boost potential.

    http://www.guru3d.com/articles_pages/amd_ryzen_7_2700x_review,29.html
    We set a voltage of 1.4 V, which is in line with what a good air cooler can handle and increased frequencies step by step. Maximum stable overclock ended up at 4.2 GHz. Since the Ryzen 2700X boosts up to 4.3 GHz out of the box, which is higher than our manual overclock, many, especially low-threaded benchmarks show a performance loss after overclocking. This is expected, due to the relatively small OC and AMD's good use of the processor's boost potential.

    https://www.tomshardware.com/reviews/amd-ryzen-7-2700x-review,5571-11.html
    Ryzen 7 2700X can be manually overclocked to 4.3 GHz. But the 1.475V required for this is more aggressive than we want to get long-term. Pushing to 4.35 GHz resulted in a crash no matter how much voltage we applied.

    https://www.sweclockers.com/test/25500-amd-ryzen-7-2700x-och-ryzen-5-2600x-pinnacle-ridge/26#content
    After a little screw we finally managed to nail 4.25 GHz for all cores, which is between 200 and 250 MHz higher than the model's basic clock frequencies with all the cores working. In order to achieve this voltage, a voltage of 1.3 V was screwed from the original 1,275 V.

    https://hothardware.com/reviews/amd-2nd-generation-ryzen-processors-and-x470-chipset-review?page=8
    To see what our chip could do, we manually cranked the core voltage up to 1.4v and shot for a modest 4.2GHz (42x100MHz), which worked without incident. We then moved up to 4.3GHz, but had some instability under load, so we backed things down a bit and settled at 4.25GHz across all cores.

    For the records... I talk mostly about AMD's High-end mainstream chips here... 7-2700x
     
  13. Talon

    Talon Notebook Virtuoso

    Reputations:
    1,482
    Messages:
    3,519
    Likes Received:
    4,694
    Trophy Points:
    331
    Exactly what I'm seeing in almost all reviews.
     
    Vistar Shook and Papusan like this.
  14. Talon

    Talon Notebook Virtuoso

    Reputations:
    1,482
    Messages:
    3,519
    Likes Received:
    4,694
    Trophy Points:
    331
    Basically yes. Gaming Intel. Workstation usage go Ryzen all day. Overclocking fun? Intel. Being able to tune 200mhz and getting a resulting performance loss is boring. Why even advertise unlocked CPUs?
     
  15. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,701
    Messages:
    29,839
    Likes Received:
    59,615
    Trophy Points:
    931
    Yeees. Awful results!! And what with those with the first gen MB and want max overclock from this new chips. Will they get same OC performance as those with the new Motherboard or are they screwed with what they already have? Aka they need to upgrade the MB to get the most out of this new chips (for maximum overclock). @Mr. Fox any comments on this?
     
    Vistar Shook likes this.
  16. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    Well, gaming if you have a 1080 as a minimum. Once at 1070 1080P is about a wash unless you turn down details/resolution and have the monitor that goes beyond 75Hz. As a new system chip looking at is as out of the box faster than before but not much better of an overclocker is about right. Again just not worth upgrade money.

    AMD has such a lackluster silicon enhancement in 12nm from GF they need to show some hard facts fast about 7nm to prove it is just not a joke, and a bad one at that. Otherwise people will look at it like it is Intel's 10nm equivalent and since there is nothing real for them beyond that it could kill AMD.

    Edit; where I thought even 5.0 GHz was too much for 7nm, I was thinking all core 4.8 or 4.9. Now I am more thinking a stretch at 4.5 or more like 4,4 GHz but at much better power. Intel can easily dip into overclock headroom here. (pure speculation)
     
    Last edited: Apr 19, 2018
  17. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,701
    Messages:
    29,839
    Likes Received:
    59,615
    Trophy Points:
    931
    [​IMG]
    Power Consumption... These numbers are measured at the 8-pin connector, which feeds the CPU power, and we can see that AMD has basically tweaked the voltage/frequency curve to favor performance over power efficiency. You can't just compare nanometer sizes, so while Global Foundries might call their process 12nmLP, it doesn't mean it's more power efficient than Intel's 14nm process. With these new CPUs AMD has basically unleashed the full performance of their Zen microarchitecture with leakier transistors, but much better performance gains
    https://www.tweaktown.com/reviews/8602/amd-ryzen-7-2700x-5-2600x-review/index10.html
     
  18. Deks

    Deks Notebook Prophet

    Reputations:
    1,272
    Messages:
    5,201
    Likes Received:
    2,073
    Trophy Points:
    331
    For 7nm Zen 2, I can see the cpu reaching 5GhZ and beyond... but only if AMD uses IBM's or TSMC's high performance process.

    If they stick with a low power process like they have with 12nm LP, then I'd have to agree with you that 4.5 GhZ (maybe 4.6) would be achievable.
    I honestly thought that the 12nm LP was designed for high performance... at least, that's what the technical specs indicated... but AMD apparently decided to go another direction... otherwise, we'd see lower voltages at 4.2/4.3 GhZ along with lower (or same) TDP.
     
    hmscott likes this.
  19. Deks

    Deks Notebook Prophet

    Reputations:
    1,272
    Messages:
    5,201
    Likes Received:
    2,073
    Trophy Points:
    331
    I hadn't used liquid metal... and repasting didn't seem to drop temperatures a lot...
    I think someone else who ordered the GL702ZC from HIDevolution got liquid metal on the CPU and Gelid GC Extreme on the GPU... temps appeared to be about 3 or 5 deg C lower than with stock Asus paste (though I could be wrong).
     
    hmscott likes this.
  20. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    Updates from Rashad, posted in his 4.6ghz video comments, I am downloading his image(s) to include too, I will update here as he posts new info:
    RashadQuickTech 3 minutes ago (edited)
    Here is the overclock voltage and speed http://prntscr.com/j7jetv

    Mr Jeff 14 hours ago
    Fake news
    RashadQuickTech 12 hours ago
    Mr Jeff The ryzen 2000 series was already avilable but we couldn't publish the benchmarks untill 19/4 if u belive its fake then don't waste your time watching my video :)

    Adam Lyon 14 hours ago (edited)
    Deffo fake. My i5 4670k with a 4.3ghz OC gets 105 FPS average using a gtx 1070 on far cry 5
    RashadQuickTech 12 hours ago
    Adam Lyon I bet its on ultra settings

    Camper X 13 hours ago
    You have shown no proof of your supposed overclock. Therefore as other have said, your just lying to get attention. well that's the way I'm putting it anyway. Shame on you for wasting peoples time.
    RashadQuickTech 12 hours ago (edited)
    Camper X Just check other yputubers review and see how well the 2700x can overclock and I didn't say it was easy to get that overclock I had to cranck up the vcore to 1.46v but on cpuz it shows 1.512V

    Scott Stamm 3 hours ago
    Please show CPU-ID showing clock speeds. Please also post settings used to achieve this.
    RashadQuickTech 33 minutes ago
    http://prntscr.com/j7iyqk

    Scott Stamm 12 minutes ago
    Why does it show as a 1700 in specification? No one elses CPU-Z validation is showing that. https://valid.x86.fr/vmjw9n
    RashadQuickTech 8 minutes ago
    when I took out my ryzen 1700 from the motherboard and putted in the 2700x it was showing the old cpu name but after flashing my mob bios and open it back it came back to normal so this is a previous screenshot was cpuid bugging out
    RashadQuickTech 3 minutes ago
    U can see after rebooting my pc it got fixed http://prntscr.com/j7jetv

    De Vil 18 hours ago
    What voltages for 4.6ghz oc?
    RashadQuickTech 12 hours ago
    De Vil vcore @1.46v
    Rashads 4.6ghz CPU-Z snapshot.JPG
     
    Last edited: Apr 19, 2018
    Reciever, Deks and Raiderman like this.
  21. Raiderman

    Raiderman Notebook Deity

    Reputations:
    742
    Messages:
    1,004
    Likes Received:
    2,434
    Trophy Points:
    181
    Lol, I use 1.4875 volts on my 1700x now to get 4.1ghz. What a bunch of weenies! Push the damn thing! 1.4 was as much as we felt comfortable with? Really? Mines in the mail, so I will see soon.

    Sent from my SM-G960U using Tapatalk
     
    hmscott, Vistar Shook and jclausius like this.
  22. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,701
    Messages:
    29,839
    Likes Received:
    59,615
    Trophy Points:
    931
    From what I see from this voltage chart... This doesn't look pretty once passing 1.425v.
    [​IMG]
     
    Vistar Shook and Falkentyne like this.
  23. Vistar Shook

    Vistar Shook Notebook Deity

    Reputations:
    2,761
    Messages:
    1,256
    Likes Received:
    1,362
    Trophy Points:
    181
  24. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    If these can do 4.6 that makes a huge difference. I just somehow doubt it. I can see there being golden chips but hat is a huge difference between 4.2 and 4.6 GHz. But it very well could be once a specific chips hits it's wall then that is it no matter what the vcore!

    As far as 1.4, I think this was AMD's recommendation to only push it that far. We can not blame the reviewers for following AMD's guidelines.

    Now a final note would be I hope there is even just a single golden chip that does 4.6 Hz. It gives hope that down the line as the process matures we may see even more capable CCX units especially in the next TR.
     
    Vistar Shook likes this.
  25. Reciever

    Reciever D! For Dragon!

    Reputations:
    1,525
    Messages:
    5,340
    Likes Received:
    4,298
    Trophy Points:
    431
    You run daily at that voltage?

    Ive been out of the desktop game for a while so I dont know what the typical tolerance levels are
     
    hmscott likes this.
  26. jaybee83

    jaybee83 Biotech-Doc

    Reputations:
    4,125
    Messages:
    11,571
    Likes Received:
    9,149
    Trophy Points:
    931
    also watched der8auer's video on 2700x overclocking. seeing as he went through a lot of chips firsthand for prebinning on air before moving to LN2, the chart he showed for a chip maxing out at 4.3 ghz is actually data for a "top chip" in his own words. so on average one should expect 100 mhz less at identical voltages.

    summing everything up weve seen in reviews thus far and adding der8auer's experience id say we got this:

    ryzen 1 averages around 3.9-4.0 ghz stable oc inside "safe voltages", whereas ryzen 2 gets to 4.2-4.3 ghz inside "safe voltages". thats a plus of 200-400 mhz, so lets say around 300 mhz more overall.

    thats really not too bad guys. sure, its not as much fun as seeing 5ghz+ on intel chips, but still nothing to sneer at. especially considering the 3% IPC bump at identical clocks when comparing ryzen 1 and 2 (according to der8auer).

    consider when we last saw an actual IPC bump from intel! that was going from haswell/broadwell to skylake, since then NO IPC bumps, just clocks and cores

    btw, if that 4.6 ghz chip holds true its not only a golden but a diamond chip basically, one in a million! again, consider der8auer going through like 50 chips to get a golden one that does 4.3 ghz on air to break the frequency world record at 6009 mhz with LN2

    Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk
     
    Last edited: Apr 20, 2018
    Vistar Shook and hmscott like this.
  27. Talon

    Talon Notebook Virtuoso

    Reputations:
    1,482
    Messages:
    3,519
    Likes Received:
    4,694
    Trophy Points:
    331

    When is he uploading a video of his 4.6Ghz overclock in real time, showing a few benchmarks and gaming would be great. I mean he's at 4.6Ghz according to his claim so it should be super easy to render that video out.
     
    Vistar Shook likes this.
  28. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    I do not even care if 4.6 is on a diamond chip. It may be only one in a million but it would show it can happen. It might be some pathways or circuits that were better etched but something physical. As the process is refined and improved this may become more consistent. Again improving the CCX's output. Again this would be just a hope and dream.
     
  29. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    Nice to see someone else noticed, the Ryzen 2.0 in some reviews posted including the just before release ones and Anandtech show Ryzen 2.0 outperforming the 8700K in more things this go around, including games. The video below briefly covers this subject.

    As I've said before, don't get overly excited, let things unfold and roll out over time, there are tweaks and tunings to find - some may have accidentally already found them and it shows in their scores. The 1700 vs 2700x tests @ 4.0ghz show 5%-13% improvements in results, that's another avenue to explore.

    In any case, if you have a Ryzen 1.0 that only OC's to 3.8/3.9ghz and want some entertainment between now and the next Zen 2 release (7nm) you can probably sell your Ryzen 1.0 CPU and for a small cost differential have fun with a 2700x, maybe even do water cooling to make it more interesting.

    And, it looks like so far the BIOS updates on x370 put them on par with the x470.

    XFR2 lets the X CPU's run almost as well as if overclocked, but not quite - 8c/16t OC still wins in many things. Still to be determined if the non-X CPU's OC'd can beat the X version stock / OC'd, RGB Coolers aside is it worth saving the $30?

    I'll post a bunch more review videos over time, for now this one speaks to the above:

    What's Going On With The Ryzen 2 Reviews??
    Published on Apr 19, 2018
    What's going on with these Ryzen 2700X and 2600X reviews?? Stay tuned...


    The AMD 2nd Gen Ryzen Deep Dive: The 2700X, 2700, 2600X, and 2600 Tested
    by Ian Cutress on April 19, 2018 9:00 AM EST
    https://www.anandtech.com/show/12625/amd-second-generation-ryzen-7-2700x-2700-ryzen-5-2600x-2600
     
    Last edited: Apr 20, 2018
    TANWare likes this.
  30. jaybee83

    jaybee83 Biotech-Doc

    Reputations:
    4,125
    Messages:
    11,571
    Likes Received:
    9,149
    Trophy Points:
    931
    anandtech out of all places! seriously, their reviews are as detailed and trustworthy as it gets if u ask me. should be interesting to see their regenerated findings :)

    Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk
     
    hmscott likes this.
  31. Talon

    Talon Notebook Virtuoso

    Reputations:
    1,482
    Messages:
    3,519
    Likes Received:
    4,694
    Trophy Points:
    331
    I mean ya it sounds totally legit that the 2700X scores this great compared to the 1800X. Nothing weird with these results at all.

    https://imgur.com/SmJBKkf
     
  32. jaybee83

    jaybee83 Biotech-Doc

    Reputations:
    4,125
    Messages:
    11,571
    Likes Received:
    9,149
    Trophy Points:
    931
    lol, sarcasm alert! :D hey, they ARE looking into it, so lets wait and see what their findings are :)

    Sent from my Xiaomi Mi Max 2 (Oxygen) using Tapatalk
     
    Talon likes this.
  33. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,701
    Messages:
    29,839
    Likes Received:
    59,615
    Trophy Points:
    931
  34. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    I see there are some getting to 4.3 GHz at under 1.4 vcore. I see though the CB R15 scores did not improve over stock that much though.
     
  35. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    I think this solves another mystery. That is why there is no 12nm Vega planned for gamers. At 10% process improvement and even 5% for design the Vega would have finally given the 1080TI a run for its money. AMD would be in the high end graphics card business again for sure. AMD it seems knew well in advance that 12nm was no where near the touted claims.
     
    Papusan likes this.
  36. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    It's interesting that the X model CPU's, like the 2700X reach higher CB scores on stock than OC'd, unless you can get OC above the 4.3ghz range.

    It's thought that the 2700 OC'd will be as good as the 2700x and you can save $30, but IDK, it's going to be close unless that 2700 will do 4.3ghz on OC.

    So far I haven't seen anyone staggering the multipliers, trying for example 4.5ghz on 1 core, leaving the others at stock first off - find the highest OC on 1 core essentially, then stagger the other cores with less multiplier as you go out - 8 cores :)

    So something like:

    45x, 41x, 40x, 40x, 40x, 38x, 38x, 38x, or some thermally / power compatible spread that performs better than all core OC...or maybe the staggering should take into account the CCX cores split so that you balance the thermal / power load evenly across the CCX's?

    When will we see some real OC tuning that fits the curve of cores vs. multiplier for hopefully better performance than stock on the X CPU's?
     
    Robbo99999 likes this.
  37. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    I do not think spread of the OC's will produce that much better results. I do think AMD keeping its word on its promise would but I guess that is not happening, at least not for now. Do not get me wrong, I am not mad over the improvement over stock, I am mad they lied about 12nm LP vs. 14nm LPP plus another 5%. It may just be in the end that 12nm was pushed out too fast and the 10% has not been realized yet but it still makes me mad.

    As I said now that no planned 12nm Vega for gamers makes so much more sense.
     
    Vistar Shook likes this.
  38. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    I think we are talking about different things. I was talking about XFR2 making the X CPU's run so well that it's like it's OC'd out of the box. In order to better XFR2 with OC you need to OC past XFR2.

    AMD didn't make any promises it didn't keep with regards to Ryzen 2.0 performance, I think you are barking up an imaginary tree again. :)

    Tests against 1700 and 2700 at 4.0ghz show 1%-17% improvement between generations, I will look up the review again and post screen shots.

    There are more than 5 screenshot's I'd like to post, so I'll post 5 and include the video again, the graph's are the first part, but the gaming benchmarks that follow are interesting too, if you haven't watched it already:
    1700 vs 2700 4ghz differences 5.jpg
     
    Last edited: Apr 20, 2018
  39. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
  40. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
  41. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    From my understanding 14 LPP was the equal of 16 nm FinFET, but you can point where this is wrong I mean since really there is so little improvement over 14LPP if we continue this track there is little to no improvemt going to 7nm? AMD can not have it both ways here.
     
  42. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    Did you refresh and read the updated post with images and video showing there is significant % improvement between Ryzen 1 and Ryzen 2? That should clear things up for you. There is indeed improvement generation to generation, including better gaming results too, check out the video.
     
  43. James D

    James D Notebook Prophet

    Reputations:
    2,314
    Messages:
    4,901
    Likes Received:
    1,132
    Trophy Points:
    231
    From what I see it's kinda controversial, it's always like this with AMD's new products.
    But if it has 10%+ at the same frequency, would be OK. but anything less feels like not enough.

    And with that power skyrocketing between 4.0 and 4.3 I'd rather stay at 4.0 lowering electricity bill.
     
    hmscott likes this.
  44. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    I do not expect 10% here at the same frequency here, that was never promised. the 10% they were talking about here is from process, meaning increased frequency etc.. The 5% could be IPC from either better design, cache structures and the like.

    Remember the increased power is under 100% load. Under most usage there is probably little increase in idle power usage. Since my high load is little to none, a 25% increase in power draw for 10% better clocks is of little consequence. YMMV of course.
     
  45. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    I think this is going to change over time, with new BIOS's and expertise gained, he was rushed with this using a BETA BIOS and little help from AMD, as was everyone else.

    Every new process has different characteristics, we don't know yet what the "safe" long term voltages are at 12nm.

    Recommendations to keep voltage low are standard suggestions, but without specific knowledge they are just that, recommendations.

    Again, we need to wait for more information to be released and discovered.

    Besides, this has nothing to do with the 1700 vs 2700X 4.0ghz tests, which was the point - to see there are indeed improvements between 12nm and 14nm, large enough that even you can't complain - they are more than expected, adding in the OC the improvements are even more.

    Ryzen 2.0 is a good improvement over Ryzen 1.0, and much less expense, with nice new higher performance coolers included, with 300mhz-400mhz improved OC (from 3.8ghz-3.9ghz to 4.2ghz-4.3ghz), I don't see what there is to complain about.
     
    Last edited: Apr 20, 2018
  46. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    I again am not complaining over the 4 GHz to 4 GHz, I am complaing about no 10% performance overclock over 14nmLPP for the 12nm LP. Two separate issues here. Where for the 1700x and 1800x under water 4.0 GHz was normally expected and 4.1 on gold and 4.2 on a diamond chip, we should see 4.4 and then 4.5 for gold and 4.6 -4.65 for diamond chips or even better.

    Also 14LPP was known to only be as good as 16nm FinFET, this is why few out there (if anyone) complained about AMD's 14nm, unlike Intel's 14nm, not reaching over 5GHz.
     
    Vistar Shook likes this.
  47. James D

    James D Notebook Prophet

    Reputations:
    2,314
    Messages:
    4,901
    Likes Received:
    1,132
    Trophy Points:
    231
    Ryzen 2.0 looks like more 200MHz improvement in clocks than 400Mhz so lets just agree to give it to AMD and call it 300MHz improvement unless this will be proven wrong by "stable BIOSes or whatever.".
     
    hmscott and ajc9988 like this.
  48. Papusan

    Papusan Jokebook's Sucks! Dont waste your $$$ on Filthy

    Reputations:
    42,701
    Messages:
    29,839
    Likes Received:
    59,615
    Trophy Points:
    931
    AMD has turned around all the stones they had to make the most of the 2nd Gen Ryzen 2700x. This is the reason we now see everywhere the ****y overclockability. Maybe they have hided some of their best bins to after Intel has launched its coming 8 core. But is it enough if this could be true? I don't think so... AMD is Saving the Ryzen 7 2800X for a Rainy Day
     
    Vistar Shook likes this.
  49. TANWare

    TANWare Just This Side of Senile, I think. Super Moderator

    Reputations:
    2,548
    Messages:
    9,585
    Likes Received:
    4,997
    Trophy Points:
    431
    This falls in line to what I had said earlier. As the process matures and they get ready to bin for TR's the better quality CCX's could lead to a 2800x. I am not sure what extra they can get out of them but I am sure this is the plan.

    So I do not think of it so much as saving it for a rainy day. This as even if they get single core to 4.5 GHz it still will not be faster than Intel for gaming. It would make it more competitive to an eight core Intel offering. It also will give better competition to Intel at the TR front as well. It still will not compete against the 7980XE, but it will close the gap somewhat.
     
    jaybee83 and Vistar Shook like this.
  50. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    7,110
    Messages:
    20,384
    Likes Received:
    25,139
    Trophy Points:
    931
    There's no need for a 2800x, the previous range 1700, 1700x, 1800, 1800x were too wide for no difference, there was no special binning for the 1800 / 1800x and getting one didn't guarentee a better OC than getting a 1700 / 1700x.

    It's unlikely there will be faster ThreadRipper 2.o's either, so in the TR 2.0 lineup I wouldn't count on higher clock or OC speeds than what we are seeing with the Ryzen 2.0 results.

    ThreadRipper 1.0 got a bit of a speed bump from process improvements in 14nm from Ryzen 1.0, but unless that happens again in 12nm between Ryzen 2.0 and ThreadRipper 2.0, there won't be any extra speed this time for ThreadRipper.
    Given the average OC from Ryzen 1.0:
    3.8ghz+400mhz = 4.2ghz
    3.9ghz+400mhz = 4.3ghz
    3.8ghz+300mhz = 4.1ghz
    3.9ghz+300mhz = 4.2ghz

    Ryzen 4.0ghz / 4.1ghz were the top OC's in Ryzen 1.0, so they would be equal to 4.3ghz and 4.4ghz in Ryzen 2.0. I don't recall seeing a 4.3ghz Ryzen 1.0, so the single 4.6ghz Ryzen 2.0 CPU would be like 4.2ghz+400mhz from Ryzen 1.0 to Ryzen 2.0.

    300mhz-400mhz is the range, if you only say 300mhz then you can't get some of the higher speeds we are seeing.

    I haven't seen it in print or heard it directly from AMD, but I've heard people quote AMD as saying the expected all core Boost is bumped up by 300mhz from Ryzen 1.0 to Ryzen 2.0, so there is that conservative view from AMD - not OC specifically, just all core Boost - so far we are seeing a bit more. :)

    AMD Ryzen 2700X & 2600X Review | Game Changing?
    AMD 300mhz all core Boost improvement quoted starting 02:45
     
    Last edited: Apr 21, 2018
← Previous pageNext page →