Page 2 of 2 FirstFirst 12
Results 17 to 31 of 31

Thread: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

  1. #17
    Member
    Join Date
    Aug 2017
    Posts
    178
    Thanks
    4
    Thanked
    11 times in 9 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    What a lot of people are forgetting is that AMD aren't competing against themselves.
    They are competing against Intel - And on price, performance and power - they are hammering them.
    The only way now, is for Intel to make themselves the cheap option....and let's face it, they may as well, to stop
    the hemorrhaging of market share.

  2. #18
    Senior Member
    Join Date
    Aug 2006
    Posts
    1,851
    Thanks
    12
    Thanked
    91 times in 80 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by 3dcandy View Post
    It's always been the way with Ryzen. Can't see it changing, the cpu cannot boost as high with multiple cores being used especially with a synthetic test like cinebench
    I know there's always an overhead on multithreading (AMD is usually better than intel on this) and in most modern cpu's there's a clock speed drop when using all the cores but my response was more about the 'difference' deksman2 was commenting on.

  3. #19
    Senior Member
    Join Date
    Aug 2003
    Location
    Wonderful Warwick!
    Posts
    3,367
    Thanks
    3
    Thanked
    133 times in 113 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by LSG501 View Post
    I know there's always an overhead on multithreading (AMD is usually better than intel on this) and in most modern cpu's there's a clock speed drop when using all the cores but my response was more about the 'difference' deksman2 was commenting on.
    It's within early samples boosts and yields I guess though. Faster and no higher tdp mainly
    Old puter - still good enuff till I save some pennies!

  4. #20
    Senior Member
    Join Date
    Oct 2006
    Location
    Coventry, United Kingdom, United Kingdom
    Posts
    335
    Thanks
    4
    Thanked
    7 times in 7 posts
    • kellyharding's system
      • Motherboard:
      • MSI B550M Mortar
      • CPU:
      • AMD Ryzen R5 3600
      • Memory:
      • 16Gb Geil
      • Storage:
      • 1TB WD Blue NVMe SSD, 2x 480GB SATA SSD, 4Tb WD BLack HDD, 2x Hiitachi 2Tb HDD
      • Graphics card(s):
      • Gigabyte WindForce GTX 970
      • PSU:
      • BeQuiet! System Power 700W
      • Case:
      • Thermaltake Core V21
      • Operating System:
      • Windows 10 Pro 64Bit
      • Monitor(s):
      • Acer 23", Samsung SM2343BW, NEC 1970nxp
      • Internet:
      • EE 70Mbps

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Ryzen chips were priced cheaper with Ryzen 1 especially as they were an unknown after the Bulldozer derived CPUs.

    By the time Ryzen 3xxx chips were launched that wasn't the case.

    Now with the 5xxx chips, AMD can afford to charge a little more, knowing that they have the best chips in general across all tasks (especially multithreaded).

    The extra price will give them much needed capital for research and develop of zen 4. Zen 4 will of course likely be a new socket with DDR5 and possibly PCI-e 5 support, and maybe more than 2 CCDs on a chip (unless they want to keep desktop topping out at 16c/32t that is).

    Hopefully Big Navi proves a success for them to keep the whole company strong, as we need that competition against Intel and nVidia in the market.
    Desktop MSI B550M Mortar, Ryzen 5 3600, 16GB DDR4 3200, 1Tb NVMe SSD, 2x480GB SATA SSD, 4Tb + 2x2Tb HDD, 3x monitors, Windows 10 Pro 64bit
    Server MSI Z170A Gaming M7, Intel i7-6700K, 16Gb DDR 2400, 120Gb SATA SSD, 1Tb WD Black HDD, 4x 4Tb HDD in RAID5


  5. Received thanks from:

    Jonj1611 (14-10-2020)

  6. #21
    Senior Member
    Join Date
    Jun 2013
    Location
    ATLANTIS
    Posts
    1,004
    Thanks
    1
    Thanked
    25 times in 23 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    AMD better be careful coz intel has billions in cash and fab their own chips meaning they make more cash. If intel would speed up their 10nm and 7nm development by 2022 AMD need to catch 50% market share before then.

  7. Received thanks from:

    CAT-THE-FIFTH (14-10-2020)

  8. #22
    Moosekateer CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    30,041
    Thanks
    3,405
    Thanked
    4,726 times in 3,649 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by philehidiot View Post
    Prices are likely high for the following:
    Except that is what Nvidia started doing,and realised they could charge more and more because they were ahead. Intel too. Apple and Samsung did the same. It also warps the secondhand market ,because it adjusts to new pricing. So you see quiet price creep over generations. Heck,I remember the Athlon 64. Its why Intel kind of needs to be competitive. AMD will price as much as Intel will,if they can get away with it. Look at the GPU market,AMD and Nvidia pricing is relatively close. History will repeat itself.

    Quote Originally Posted by deksman2 View Post
    I wonder why the multi-core scores didn't gain more.

    You'd think that an increase in IPC of 19% would have a more profound effect on multi-core performance.
    I realize the 5xxx series has a 100MhZ lower base clock, but that's a difference of mere 3% (and the loss in performance wouldn't be that much).
    Because IPC for marketing purposes isn't really IPC,more performance per core improvement. So AMD and Intel are including core and uncore improvements into that figure.

    So the effects of having one CCX per CCD,etc are part of those "IPC increases". Also something which needs to be considered,is unlike Zen2,each core now has access to upto 32MB of cache instead of 16MB.

    Clockspeeds also are another consideration as mentioned before. I have heard rumours Zen3 is close to the edge than Zen2 is,so it could be that since they are using the same 7NM process,the CPUs start to throttle under heavy MT load. So single threaded scores are probably a combination of higher realworld clockspeed and doubled L3 cache per core.

    You can kind of see the effect of L3 cache on Zen2. Renoir is noticeably worse than the Zen2 CPUs in certain applications due to its L3 cache reduction.

    I also think gaming performance improvements will be more than application improvements(unless there is more AVX performance). The Zen2 cores is held back by inter-CCX latency it appears. For older engines this will cause a noticeable performance hit which is why Intel Skylake mesh bus CPUs fared much worse than Skylake ring bus CPUs.

    So expect Zen3 improvements to be greatest in games with older engines. I expect newer engines using Vulkan/DX12 might show less of an improvement over Zen2. These will be more CCX aware,especially as the consoles are using CPUs with the older CCX topology.
    Last edited by CAT-THE-FIFTH; 14-10-2020 at 09:26 AM.


    Those despicable Elk,stealing the pond weed!

  9. Received thanks from:

    Sumanji (14-10-2020)

  10. #23
    Registered+
    Join Date
    Dec 2015
    Posts
    28
    Thanks
    0
    Thanked
    4 times in 3 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by [GSV
    Trig;304]Thats not a bad jump for me on my 1600AF, however I'm stuck on a B450M, so unless ASRock actually drop a BIOS for me next year then I'll be sticking with my 1600AF for now and just getting a GPU to replace the RX580-8Gb and see how much I'm actually held back by the CPU..
    Unless you game at high refresh rates, you shouldn't be held back much at all. Fingers crossed for the BIOS updates anyway, though!

  11. #24
    Registered+
    Join Date
    Dec 2015
    Posts
    28
    Thanks
    0
    Thanked
    4 times in 3 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by lumireleon View Post
    People why buy 6 core cpus are extremely on a budget.
    Citation needed on that one. It's only been 3 years since 6-core CPUs were even close to affordable (thanks AMD), and as far as I'm aware people like me who are "extremely on a budget" still buy quad-core.

    Everyone who needs 6-cores on a budget can either buy the 3600 series or wait for a 5600 (non-X) to launch.

  12. Received thanks from:

    Sumanji (14-10-2020)

  13. #25
    Senior Member
    Join Date
    Jul 2003
    Posts
    11,045
    Thanks
    682
    Thanked
    388 times in 271 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    I'm on an old 1920*1080@60Hz monitor and want to update it to 144Hz or above, so I'm thinking GPU first (3070/Navi) then monitor, then CPU at some point next year if needed, not sure if I'm going to stick to 1920*1080 and just increase refresh or increase resolution as well, can't see me being held back by the CPU just yet.

  14. #26
    root Member DanceswithUnix's Avatar
    Join Date
    Jan 2006
    Location
    In the middle of a core dump
    Posts
    11,164
    Thanks
    603
    Thanked
    1,221 times in 1,038 posts
    • DanceswithUnix's system
      • Motherboard:
      • Asus X470-PRO
      • CPU:
      • 3700X
      • Memory:
      • 16GB 3200MHz
      • Storage:
      • 1TB Linux, 1TB Games (Win 10)
      • Graphics card(s):
      • Asus Strix RX Vega 56
      • PSU:
      • 650W Corsair TX
      • Case:
      • Antec 300
      • Operating System:
      • Fedora 30 + Win 10 Pro 64 (yuk)
      • Monitor(s):
      • Benq XL2730Z 1440p + Samsung 2343BW 2048x1152
      • Internet:
      • Zen 80Mb/20Mb VDSL

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by CAT-THE-FIFTH View Post
    . Also something which needs to be considered,is unlike Zen2,each core now has access to upto 32MB of cache instead of 16MB.
    It always did, but required a cross CCX trip to get to the data. Still cheaper than hitting main ram though.

    that could be what is being seen here, scaling is dropping off as core counts increase (no great surprise) so that could just be pressure on the ram interface. We need someone to do a ram scaling test with different ram speeds, ie we just won't know until parts turn up in reviewer's hands.


    Then we get to find out what really matters, how fast it is in PiFast

  15. #27
    Member
    Join Date
    Mar 2018
    Posts
    195
    Thanks
    8
    Thanked
    11 times in 10 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    A 19% increase in IPC doesn't mean a 19% increase in perf/watt. MT workloads end up power limited so the chips can't match the ST gains.

  16. #28
    root Member DanceswithUnix's Avatar
    Join Date
    Jan 2006
    Location
    In the middle of a core dump
    Posts
    11,164
    Thanks
    603
    Thanked
    1,221 times in 1,038 posts
    • DanceswithUnix's system
      • Motherboard:
      • Asus X470-PRO
      • CPU:
      • 3700X
      • Memory:
      • 16GB 3200MHz
      • Storage:
      • 1TB Linux, 1TB Games (Win 10)
      • Graphics card(s):
      • Asus Strix RX Vega 56
      • PSU:
      • 650W Corsair TX
      • Case:
      • Antec 300
      • Operating System:
      • Fedora 30 + Win 10 Pro 64 (yuk)
      • Monitor(s):
      • Benq XL2730Z 1440p + Samsung 2343BW 2048x1152
      • Internet:
      • Zen 80Mb/20Mb VDSL

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by edmundhonda View Post
    A 19% increase in IPC doesn't mean a 19% increase in perf/watt. MT workloads end up power limited so the chips can't match the ST gains.
    AMD claim a 20% improvement in power efficiency, better than the IPC increase.

  17. #29
    Senior Member
    Join Date
    Aug 2003
    Location
    Wonderful Warwick!
    Posts
    3,367
    Thanks
    3
    Thanked
    133 times in 113 posts

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Anyone looking at this thread might want to take a look at the latest Intel review up on ere...

    Intel have basically canned a processor because they are losing to AMD so much... and yet the opening part of the review on Hexus states that even so the chip might be at 125tdp but often pulls 175w under load. And the review agrees that it's not a great buy, is old tech but still doesn't say it's not worth upgrading. Bizarre really but I guess we haven't seen the 5000 series reviews yet
    Old puter - still good enuff till I save some pennies!

  18. #30
    Banhammer in peace PeterB kalniel's Avatar
    Join Date
    Aug 2005
    Posts
    29,876
    Thanks
    1,624
    Thanked
    3,053 times in 2,476 posts
    • kalniel's system
      • Motherboard:
      • Gigabyte Z390 Aorus Ultra
      • CPU:
      • Intel i9 9900k
      • Memory:
      • 32GB DDR4 3200 CL16
      • Storage:
      • 1TB Samsung 970Evo+ NVMe
      • Graphics card(s):
      • nVidia GTX 1060 6GB
      • PSU:
      • Seasonic 600W
      • Case:
      • Cooler Master HAF 912
      • Operating System:
      • Win 10 Pro x64
      • Monitor(s):
      • Dell U2311H
      • Internet:
      • rubbish

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by 3dcandy View Post
    Anyone looking at this thread might want to take a look at the latest Intel review up on ere...

    Intel have basically canned a processor because they are losing to AMD so much... and yet the opening part of the review on Hexus states that even so the chip might be at 125tdp but often pulls 175w under load. And the review agrees that it's not a great buy, is old tech but still doesn't say it's not worth upgrading. Bizarre really but I guess we haven't seen the 5000 series reviews yet
    175W under load is below spec I bet - the 10900K is rated to 250W under load I would expect the 10850 to be at least 230W

  19. #31
    Moosekateer CAT-THE-FIFTH's Avatar
    Join Date
    Aug 2006
    Location
    Not here
    Posts
    30,041
    Thanks
    3,405
    Thanked
    4,726 times in 3,649 posts
    • CAT-THE-FIFTH's system
      • Motherboard:
      • Less E-PEEN
      • CPU:
      • Massive E-PEEN
      • Memory:
      • RGB E-PEEN
      • Storage:
      • Not in any order
      • Graphics card(s):
      • EVEN BIGGER E-PEEN
      • PSU:
      • OVERSIZED
      • Case:
      • UNDERSIZED
      • Operating System:
      • DOS 6.22
      • Monitor(s):
      • NOT USUALLY ON....WHEN I POST
      • Internet:
      • FUNCTIONAL

    Re: Cinebench R20 scores for all Ryzen 5000 CPUs appear online

    Quote Originally Posted by DanceswithUnix View Post
    It always did, but required a cross CCX trip to get to the data. Still cheaper than hitting main ram though.

    that could be what is being seen here, scaling is dropping off as core counts increase (no great surprise) so that could just be pressure on the ram interface. We need someone to do a ram scaling test with different ram speeds, ie we just won't know until parts turn up in reviewer's hands.


    Then we get to find out what really matters, how fast it is in PiFast
    Renoir does seem to be able to run IF at a higher speed IIRC. Would be interesting to see how much Zen3 is limited by IF frequency.

    I wonder if the power improvements are done to more efficient IF. IIRC,Zen1 IF consumed a lot of the power budget as core counts rose.


    Those despicable Elk,stealing the pond weed!

Page 2 of 2 FirstFirst 12

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •