Jump to content

Ryzen Preorder


90 replies to this topic

#81 gaIaxor

    Member

  • PipPip
  • 24 posts

Posted 08 March 2017 - 04:04 PM

There is a bug in the W10 scheduler. So SMT off, W7 or pegging MWO threads only to real cores (on same CCX) may increase performance on Ryzen.
Can someone check that? (Temporary pegging MWO threads with the task manager while on Testing Grounds won't change any settings.)

Quote

Windows 10′ scheduler correctly identifies Intel’s hyper-threads as lesser performing than principal core threads and schedules tasks in a way that’s takes advantage of the additional throughput without negatively impacting performance. Unfortunately the scheduler currently is not able to differentiate principal core threads from virtual SMT threads with Ryzen and in fact sees 16 thread Ryzen 7 processors as processors with 16 physical cores with equal resources per thread.
...
It also incorrectly identifies the amount of cache available per thread. Adding up the amount of L2 and L3 cache Windows 10’s scheduler “thinks” is there totals to an insane 136MB of cache, when Ryzen 7 in fact only has 20MB of L2+L3 cache combined.

http://wccftech.com/...-scheduler-bug/

#82 Dr Cara Carcass

    Member

  • PipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 643 posts

Posted 09 March 2017 - 04:28 AM

View PostSmokeyjedi, on 08 March 2017 - 01:33 PM, said:

SO cara cara or whatever is literally the only person in the whole world who could post @ 4.2ghz on AIR/ with no added voltages.......

Pretty interesting indeed.


so smoke yeti is the blind guy.......

#83 Smokeyjedi

    Member

  • PipPipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 1,040 posts
  • LocationCanada

Posted 09 March 2017 - 01:34 PM

View PostCara Carcass, on 09 March 2017 - 04:28 AM, said:


so smoke yeti is the blind guy.......

this make literally no sense, Care to try again?

Im calling your oc, out its a lie, you are not the only person on earth to post 4.2ghz with no voltage, ROFL. care to share what YOUR smoking? must be better than mine.

#84 Bill Lumbar

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 2,073 posts

Posted 09 March 2017 - 03:50 PM

View PostxWiredx, on 07 March 2017 - 08:09 AM, said:

Yeah, so... the original Some Performance Tests thread is here: https://mwomercs.com...formance-tests/

It would be nice if we could get a new post going in similar fashion with Ryzen-based systems.

Again, the testing methodology for me was pretty straight-forward: I piloted a heavy/assault and ran the FRAPS bench for 180 seconds. This was decent for me since about 1 minute was spent getting into position for the fight, about 1 minute was spent trading and repositioning or lightly fighting, and then one minute was spent in heavier action.

I documented which maps I was on, too, since there are some somewhat wide swings in performance depending on the map.

I did not test FP nearly as rigorously. Or whatever it is called now. I haven't played in about a month, taking a break to enjoy some other games (like Breath of the Wild).

ALSO, for those who were questioning it... a guy with a 5GhZ FX 8350 ran these numbers using my same methodology back then because he too claimed his FX chip (which even AMD admits is an utter failure) did magical things:

Please tell me you are not talking about me Wired.... lol

#85 ninjitsu

    Member

  • PipPipPipPipPipPip
  • FP Veteran - Beta 2
  • FP Veteran - Beta 2
  • 402 posts

Posted 09 March 2017 - 06:19 PM

I have to boost to 1.35V to get a stable 4.0GHZ.

#86 Dr Cara Carcass

    Member

  • PipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 643 posts

Posted 10 March 2017 - 05:11 AM

View PostSmokeyjedi, on 09 March 2017 - 01:34 PM, said:

this make literally no sense, Care to try again?

Im calling your oc, out its a lie, you are not the only person on earth to post 4.2ghz with no voltage, ROFL. care to share what YOUR smoking? must be better than mine.


No yeti, i dont care to try again, not after how you approached me. If you werent such an ******* i would have posted a screenshot and explained what happened - but not this way.

View Postninjitsu, on 09 March 2017 - 06:19 PM, said:

I have to boost to 1.35V to get a stable 4.0GHZ.


Isnt 1.35V the standard setting?

#87 ninjitsu

    Member

  • PipPipPipPipPipPip
  • FP Veteran - Beta 2
  • FP Veteran - Beta 2
  • 402 posts

Posted 10 March 2017 - 05:28 AM

View PostCara Carcass, on 10 March 2017 - 05:11 AM, said:


Isnt 1.35V the standard setting?


For my chip, the stock voltage is 1.18750. I think I see what's happening with your OC, Cara. It seems like the Asus utility it raising your voltage for you. 1.35V is definitely higher than stock voltage. I think 4.2 is possible on air. I top out at 47C with 1.35V using a Corsair H110i and fans at 50%.

Edited by ninjitsu, 10 March 2017 - 05:50 AM.


#88 Peter2k

    Member

  • PipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 2,032 posts
  • LocationGermany

Posted 10 March 2017 - 05:46 AM

View PostCara Carcass, on 10 March 2017 - 05:11 AM, said:


Isnt 1.35V the standard setting?


voltage is a bit more complicated on Ryzen

the chip controls it automatically at stock clocks

when you overclock it sets itself higher automatically

Quote

Information from AMD Regarding Overclocking. As a general guideline: a CPU voltage of up to 1.35 V is acceptable for driving everyday overclocks


Quote

AMD Ryzen processors do not use pre-programmed VID tables.

- Therefore, there is no fixed Vcore when the CPU runs in its out-of-box condition.

- Default Vcore will vary depending on workload and will range from 1.2-1.3625V.

- Overclocking an AMD Ryzen processor will snap the voltage to 1.3625V, but this value can be changed.


#89 Dr Cara Carcass

    Member

  • PipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 643 posts

Posted 10 March 2017 - 03:12 PM

View Postninjitsu, on 10 March 2017 - 05:28 AM, said:


For my chip, the stock voltage is 1.18750. I think I see what's happening with your OC, Cara. It seems like the Asus utility it raising your voltage for you. 1.35V is definitely higher than stock voltage. I think 4.2 is possible on air. I top out at 47C with 1.35V using a Corsair H110i and fans at 50%.


Yeah i relized that yesterday - probelt says 1.488V software says 1.35V

#90 Smokeyjedi

    Member

  • PipPipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 1,040 posts
  • LocationCanada

Posted 10 March 2017 - 06:07 PM

View PostCara Carcass, on 10 March 2017 - 03:12 PM, said:


Yeah i relized that yesterday - probelt says 1.488V software says 1.35V

Ya I was ribbin ya, But The ******* factor was misread I was referring to your comment just above, I couldnt make sense of what you were saying........

And Sorry to be so hard *** about the facts but If youve been here long enough (5-6 years) youd be used to my direct throat punch methodology....sorry

#91 Dr Cara Carcass

    Member

  • PipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 643 posts

Posted 11 March 2017 - 07:10 AM

View PostSmokeyjedi, on 10 March 2017 - 06:07 PM, said:

Ya I was ribbin ya, But The ******* factor was misread I was referring to your comment just above, I couldnt make sense of what you were saying........

And Sorry to be so hard *** about the facts but If youve been here long enough (5-6 years) youd be used to my direct throat punch methodology....sorry


How does the time you spend here rectify your way of treating people? I have been aropund since closed beta - does that make it any better when i am an ******* to others?

But well i appreciate that u were nice enough to give a word of appology.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users