TR Forums

I just ran Cinebench R20 on my Ryzen 2700X at stock speeds.

MP: 3913 cb
SP: 425 cb

Seems legit to me. I'm happy with my Ryzen. Plus, it supports ECC.

DancinJack wrote:

chuckula wrote:

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

Was that running Windows on the Mac?

No. I'd be VERY surprised if anyone got a 2400G running on macOS.

Uh... I meant was his MacBook running Windows.

I kind of assumed the 2400G was running Windows.

#MinisterOfGerbilAffairsAchievementUnlocked

mikewinddale wrote:

I just ran Cinebench R20 on my Ryzen 2700X at stock speeds.

MP: 3913 cb
SP: 425 cb

Seems legit to me. I'm happy with my Ryzen. Plus, it supports ECC.

What are the clock speeds with stock, in your case? (like what is the average all core speed)

Stock these days are kind of annoying.

thecoldanddarkone wrote:

What are the clock speeds with stock, in your case? (like what is the average all core speed) Stock these days are kind of annoying.

4,349.0 MHz single core (turbo), and 3,949.1 MHz all-core.

Maximum CPU temperature (while running Cinebench R20) is 68.5C (Tdie, without the +10C offset), with a be quiet! Dark Rock Pro 4.

chuckula wrote:

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

Was that running Windows on the Mac?

No, I just have a 256 GB SSD and Logic Pro X, so I don't really have the room. I'm aware that across OSes it's possible that performance is different.

DancinJack wrote:

chuckula wrote:

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

Was that running Windows on the Mac?

No. I'd be VERY surprised if anyone got a 2400G running on macOS.

Kentucky basketball gettin the better of DancinJack tonight. He did say "on the Mac", meaning on the MBP, or so I presume.

derFunkenstein wrote:

:lol: Kentucky basketball gettin the better of DancinJack tonight. He did say "on the Mac", meaning on the MBP, or so I presume.

Kansas, but yeah.

Dang it I shoulda known that. I saw KU and assumed Kentucky despite your flagrant Kansas fandom.

derFunkenstein wrote:

Dang it I shoulda known that. I saw KU and assumed Kentucky despite your flagrant Kansas fandom.

Or the fact that my location says Kansas.

And Kentucky is "UK."

All you can say about Cinebench is...wait, is there something to say about Cinebench?

Glorious wrote:

oh thank god I needed some High-Test Chuckula to get me through the rest of this day.

OP delivers!

Sad!

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

MBP's probably experiencing thermal throttling.

(And off-topic, but GAWD I hate the new MBP keyboards...)

mikewinddale wrote:

thecoldanddarkone wrote:

What are the clock speeds with stock, in your case? (like what is the average all core speed) Stock these days are kind of annoying.

4,349.0 MHz single core (turbo), and 3,949.1 MHz all-core.

Maximum CPU temperature (while running Cinebench R20) is 68.5C (Tdie, without the +10C offset), with a be quiet! Dark Rock Pro 4.

Thank you for satisfying my curiosity!

just brew it! wrote:

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

MBP's probably experiencing thermal throttling.

(And off-topic, but GAWD I hate the new MBP keyboards...)

I don't know about throttling, exactly, but it's not running full speed. I've installed Intel's Power Gadget and it shows the CPU and with all 8 threads busy it hovers around 3.3 GHz and a CPU temp hovering around 83-85 Celsius. It's got a single-core max turbo of 3.8 GHz. Meanwhile the Ryzen is running at 3.8 with all 8 threads pumping. And that's with the CPU performing graphics duties.

Guess I'll have to temporarily clear out some space and install Windows on the Mac to really see what it's capable of.

I, for one, welcome our new Cinebench overlords.

Just tried Cinebench R20 on my new system.

i9-9900K Stock at 3.6 GHz

CPU - 4715 CB
CPU (single core) - 479 CB
MP Ratio - 9.84x

Khali wrote:

Just tried Cinebench R20 on my new system.

i9-9900K Stock at 3.6 GHz

CPU - 4715 CB
CPU (single core) - 479 CB
MP Ratio - 9.84x

What is the cpu clocked at while it's running this. 4715, is probably a decent all core speed.

Thank you

my puny 6700K needs to get upgraded.

c'mon sonny cove

9900k @ Stock
4699/515

Running 4.7GHz all-core for the first half of the benchmark, but toward the middle heat starts to saturate my 240mm rad and some cores are at 4.6GHz, toward the end a few cores are seeing 4.5GHz. The single core benchmark is pretty much pegged at 5GHz.

Rather interestingly the score actually goes down when I overclock to 5GHz because the higher applied voltage causes the chip to generate more heat and it starts throttling sooner. I ended up with a multi-core score of 4666. Clocks were on average 100MHz less than when the chip was set to stock, dipping as low as 4.3GHz toward the end of the benchmark. The single core score was essentially unchanged. This was with ambient temps of 25-26°C.

Windows Store only? Ugh. WHY.

synthtel2 wrote:

just brew it! wrote:

synthtel2 wrote:

(I usually run SMT-less now because apparently Windows' scheduler is hot garbage,

It's not just the fault of the scheduler; SMT is really hard to schedule for optimally. Blaming the scheduler when SMT performs poorly on certain workloads is, in a sense. blaming the messenger.

It's nothing inherent to SMT, it's the thing where Windows wants to make my threads share cores even when there are plenty of actually free cores to go around. Doing better would literally be as simple as not assigning work to any even-numbered logical cores unless all odd-numbered logical cores are already in use. I can also blame Windows because I dual-boot and Windows very clearly has a problem with this that Linux doesn't.

MacOS does a pretty decent job scheduling every other core (primary thread, if you will), as well. Spikes are almost always visible on every other thread.

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

Is your Ryzen a desktop? I can imagine that there could potentially be thermal throttling involved in a slim notebook design when running all-cores all-out sustained for multiple minutes.

2400G is a desktop processor. I've already covered all the rest of it on page 2.

Waco wrote:

Windows Store only? Ugh. WHY.

1-click install, 1-click uninstall, and centralized updates are pretty compelling reasons.

anotherengineer wrote:

HA

Called it.

Read the title. Guessed the author, guessed right.

No point reading this thread.

Thanks for your input.

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

Why are you comparing a 2018 65W TDP desktop CPU to a 2017 45W TDP laptop CPU?

Usacomp2k3 wrote:

Waco wrote:

Windows Store only? Ugh. WHY.

1-click install, 1-click uninstall, and centralized updates are pretty compelling reasons.

That's not much different than properly designed applications on the older API.

Anyone with a multi-socket system or a higher core count Threadripper care to run this benchmark? Would love to see the scaling across multiple dies/sockets/NUMA domains.

End User wrote:

derFunkenstein wrote:

Amusingly (based on the thread title), or perhaps not, my Ryzen 5 2400G system is faster than the entry level 2017 MacBook Pro 15" with the Core i7-7700HQ. The AMD system scored 1740 and the MacBook rang in at 1685. I expected it to be kind of close but I figured the Mac would win. ¯\_(ツ)_/¯

Why are you comparing a 2018 65W TDP desktop CPU to a 2017 45W TDP laptop CPU?

Because it's what I have. Why not?

Core i7-8086K @ 5 GHz all cores / DDR4-3933 (but I have a bunch of stuff running):
499 CB ST | 3851 CB MT
Ryzen 7 1700 @ 3.9 GHz all-core / DDR4-3200:
381 CB ST | 3756 CB MT
Core i7-4790K @ 4.4 GHz all-core / DDR3-2133:
386 CB ST | 1965 CB MT

Cinebench is trash. ('ω')

techguy wrote:

Anyone with a multi-socket system or a higher core count Threadripper care to run this benchmark? Would love to see the scaling across multiple dies/sockets/NUMA domains.

The program has a few examples from such systems.

2x Xeon Platinum 8168 (48C/96T): 16536 CB MT
4x Xeon Processor E7-4890v2 (60C/120T): 12986 CB MT
1x Ryzen Threadripper 1950X (16C/32T): 6670 CB MT

Hey, this'll be fun to run. I can only wonder what the performance delta will be like between my 7940x and the FX 6300 I'm putting together as a Frankenputer/project PC this week... I'd include my Ryzen 1700 too, but it's not running Windows. Nevertheless, I'll update with my scores when I can.