XMG APEX 15 MAX (E22) [Clevo NH5#VR] {R9-5950X} vs XMG Ultra 17 (M20) [Clevo X170KM-G] {i9-10900K}?

[NBTUser]

Recently, XMG announced the APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} (where # denotes a number indicating non-essential cosmetic differences).
Since this announcement I've been wondering how the XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T) variant} would perform as compared to their XMG Ultra 17 (M20) [Clevo X170KM-G) {Intel i9-10900K (10C/20T) variant}.

While XMG's APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T}'s announcement page lists a benchmark overview comparison**; they do not list the XMG Ultra 17 (M20) [Clevo X170KM-G] {Intel i9-10900K (10C/20T) variant} as one of the laptops compared in their benchmark overview.
I have asked XMG via Reddit to add the XMG Ultra 17 (M20) [Clevo X170KM-G] {Intel i9-10900K (10C/20T)} variant to their comparison benchmark overview, but they have not added this laptop yet.

Absent of this comparison, I have been struggling to figure out how the: AMD Ryzen 9 5950X (16C/32T) [105W TDP] with base clock: 3.4 GHz and max boost clock: up to 4.9 GHz performs in the XMG APEX 15 MAX (E22) [Clevo NH5#VR] as compared to how the: Intel i9-10900K (10C/20T) [125W TDP] with base frequency: 3.7 GHz and boost frequency 5.1 GHz performs in the XMG Ultra 17 (M20) [Clevo X170KM-G].

Specifically, since the XMG APEX 15 MAX (E22) [Clevo NH5#VR] runs the AMD Ryzen 9 5950X (16C/32T) [105W TDP] at: 65W base/88W PPT (Package Power Tracking) whereas the XMG Ultra 17 (M20) [Clevo X170KM-G) runs the Intel i9-10900K (10C/20T) [125W] at: 125W.
As a result, I am trying to figure out how the AMD Ryzen 9 5950X (16C/32T) [65W base/88W PPT] would perform as compared to the Intel i9-10900K (10C/20T) [125W]?

While AMD and Intel list on their respective CPU spec sheets base clocks, base TDPs, and boost clocks; neither lists "boost TDP" in other words how much watts past the base TDP watts you have to hit to get those respective claimed boost clocks. This is further compounded by the fact that both AMD and Intel on their respective spec sheets also do not list "all-core" speeds for their respective CPUs, so I've had to look at tech publication reviews for the AMD Ryzen 9 5950X (16C/32T} which subjectively lists their all-core clock at: 3.85-4.0 GHz (16C/32T) [???W] and Intel i9-10900K (10C/20T)'s Wikipedia page which subjectively lists their all-core clock at: 4.8 GHz (10C/20T) [???W]***.

Given the different all-core speeds, the inherent issue here is this - how do you account for the difference in cores/threads to determine which performs better when they both have different all-core speeds?

Put another way: at what all-core clock speeds (for each) is equal performance reached between the AMD Ryzen 9 5950X (16C/32T} and the Intel i9-10900K (10C/20T)?

For reference, on my Sager NP9672M-G0 [Clevo X170KM-G] {Intel i9-10900K (10C/20T) variant} I can hit a max all-core speed of: 4.7 GHz using around 200W sustainably on the edge of but not hitting thermal throttling; where any higher and thermal throttling kicks in.

The focal point of my question is this: how does a X GHz all-core speed on 16C/32T at 65W/88W PPT (XMG APEX 15 MAX (E22) [Clevo NH5#VR]) compare to a Y GHz all-core speed on 10C/20T at 125W/~200?W boost (XMG Ultra 17 (M20) [Clevo X170KM-G])?

For context, I do data analysis work, so I'm trying to gauge the all-core (aka multicore) performance comparison between these two laptops and their respective desktop CPUs; as I am trying to figure out if there is an increase in all-core performance in purchasing the XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} |65W/88W PPT| over my existing Sager NP9672M-G0 [Clevo X170KM-G] {Intel i9-10900K (10C/20T) variant} |125W/200W boost|.

 

Links:
1) **XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} Announcement Link: https://www.reddit.com/r/XMG_gg/comments/ujn04y/launch_extended_ryzen_5000_cpu_support_in_xmg/

2) **XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} Benchmark Overview Link: https://docs.google.com/spreadsheets/d/e/2PACX-1vTJdELoyS9gEPIJCb4_y7-d4TlV5V_skKFkzUQFnGH84szj5JdpfA1yxFByUi0cxrJE8igObxrJ6RJ3/pubhtml

3) ***AMD Ryzen 9 5950X All-Core Speed Link: https://www.kitguru.net/components/cpu/luke-hill/amd-ryzen-9-5950x-zen-3-cpu-review/all/1/?PageSpeed=noscript

4) ***Intel i9-10900K All-Core Speed Link: https://en.wikipedia.org/wiki/Comet_Lake_(microprocessor)

[electrosoft]

If the 5950x is working within spec and not subject to thermal throttling it will decimate the 10900k in multi and be darn close if not better in single but it won't because it is physically capped at 88w (yikes).

 

Here is a run on my X170SM-G with a 10900k running at stock tuned no overclock on anything.

 

 

The problem with the AMD laptop from XMG is the hardware is physically capped at 88w which limits the scores. That means the 5950x in it scores roughly the same as a stock 10900k (~16500k). Tuning the curve (which isn't a real OC as it is subject to load variants) nets. They even warn you the 5950x is not recommended for that model and you can see why. It will put an absolute pounding on its VRMs pushing them to max/cap under load everytime. The max  somewhat realistic chip for it is the 5900x which will give you less than 10900k stock performance. The best chip for it is the 5600x IMHO. They won't even offer a 5800X3D because it runs too hot and heavy for that chassis.

 

Again, the problem is the physical hardware limitations of an 88w cap. That just is a deal breaker in every aspect. Even Clevo's new 15.6" Socket 1700 laptop has a cap current cap of 125w and that is BIOS limited. It will probably open up to 150w max.

 

Even XMG recommends only a specific range of CPUs:

 

-----------------------------------------------

Recommendations

Our recommended CPUs for XMG APEX 15 MAX include:

• AMD Ryzen 5 5600X desktop CPU | 6 cores/12 threads | 32 MB cache | max 88 W PPT

• AMD Ryzen 7 5700X desktop CPU | 8 cores/16 threads | 32 MB cache | max 88 W PPT

• AMD Ryzen 9 5900X desktop CPU | 12 cores/24 threads | 64 MB cache | max 88 W PPT (ECO Mode)

The best price/performance ratio is arguably in the AMD Ryzen 7 5700X.

AMD Ryzen 9 5950X is not recommended in XMG APEX 15 MAX unless the user is willing to manually adjust frequency/voltage curve and other manual optimizations in AMD Ryzen Master. Results might be prone to silicon lottery. XMG will not guarantee specific performance results.

System stability is only guaranteed when system runs at factory defaults, including the 4 performance profiles in Control Center: Power Saving, Quiet, Entertainment & Performance."

-------------------------------------------------------

[jaybee83]

i dont get it...the 5950X is SO frugal when it comes to power, can completely stretch its legs at 125W with stock clocks (https://www.techpowerup.com/review/amd-ryzen-9-5950x/19.html).

 

so why cripple it at 88W?!?! they already did that with the previous gen 3950X DTRs from Clevo, and now the same mistake again.... 125W is really not that much, a proper DTR should be able to take care of 150W for the CPU easily, if not more.... (as an example, the 5950X in the techpowerup review oced to 4.5 ghz all core draws 194W in multi threaded)

[electrosoft]

2 hours ago, jaybee83 said:

i dont get it...the 5950X is SO frugal when it comes to power, can completely stretch its legs at 125W with stock clocks (https://www.techpowerup.com/review/amd-ryzen-9-5950x/19.html).

 

so why cripple it at 88W?!?! they already did that with the previous gen 3950X DTRs from Clevo, and now the same mistake again.... 125W is really not that much, a proper DTR should be able to take care of 150W for the CPU easily, if not more.... (as an example, the 5950X in the techpowerup review oced to 4.5 ghz all core draws 194W in multi threaded)

 

Agreed, I am perplexed why they neutered it so badly knowing a similar design PL like Intel would basically afford it all the headroom it needs to run at stock. That laptop with a full 125w PL or even 150w and the 5950x would smoke every laptop out there for multi. It can't touch the NHJ for single and 12th though even if full powered. Nothing can touch 12th gen right now.

 

As it stands now, if you're in the market for the best multi performance in a true desktop CPU the NHJ is top dog for a 15.6" form factor and the X170SM/KM for 17.3".

 

 

[Taraquin]

Using curve optimizer you can get quite good scores with 5950X even at 88W PPT. I agree that they should have made higher cap possible, but you can get quite good scores at 88W and neg CO 🙂 If it's possible to lower soc, vddg, vddp etc you can free up aditional performance for cores aswell. 

[1610ftw]

On 5/20/2022 at 4:50 PM, NBTUser said:

Recently, XMG announced the APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} (where # denotes a number indicating non-essential cosmetic differences).
Since this announcement I've been wondering how the XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T) variant} would perform as compared to their XMG Ultra 17 (M20) [Clevo X170KM-G) {Intel i9-10900K (10C/20T) variant}.

While XMG's APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T}'s announcement page lists a benchmark overview comparison**; they do not list the XMG Ultra 17 (M20) [Clevo X170KM-G] {Intel i9-10900K (10C/20T) variant} as one of the laptops compared in their benchmark overview.
I have asked XMG via Reddit to add the XMG Ultra 17 (M20) [Clevo X170KM-G] {Intel i9-10900K (10C/20T)} variant to their comparison benchmark overview, but they have not added this laptop yet.

Absent of this comparison, I have been struggling to figure out how the: AMD Ryzen 9 5950X (16C/32T) [105W TDP] with base clock: 3.4 GHz and max boost clock: up to 4.9 GHz performs in the XMG APEX 15 MAX (E22) [Clevo NH5#VR] as compared to how the: Intel i9-10900K (10C/20T) [125W TDP] with base frequency: 3.7 GHz and boost frequency 5.1 GHz performs in the XMG Ultra 17 (M20) [Clevo X170KM-G].

Specifically, since the XMG APEX 15 MAX (E22) [Clevo NH5#VR] runs the AMD Ryzen 9 5950X (16C/32T) [105W TDP] at: 65W base/88W PPT (Package Power Tracking) whereas the XMG Ultra 17 (M20) [Clevo X170KM-G) runs the Intel i9-10900K (10C/20T) [125W] at: 125W.
As a result, I am trying to figure out how the AMD Ryzen 9 5950X (16C/32T) [65W base/88W PPT] would perform as compared to the Intel i9-10900K (10C/20T) [125W]?

While AMD and Intel list on their respective CPU spec sheets base clocks, base TDPs, and boost clocks; neither lists "boost TDP" in other words how much watts past the base TDP watts you have to hit to get those respective claimed boost clocks. This is further compounded by the fact that both AMD and Intel on their respective spec sheets also do not list "all-core" speeds for their respective CPUs, so I've had to look at tech publication reviews for the AMD Ryzen 9 5950X (16C/32T} which subjectively lists their all-core clock at: 3.85-4.0 GHz (16C/32T) [???W] and Intel i9-10900K (10C/20T)'s Wikipedia page which subjectively lists their all-core clock at: 4.8 GHz (10C/20T) [???W]***.

Given the different all-core speeds, the inherent issue here is this - how do you account for the difference in cores/threads to determine which performs better when they both have different all-core speeds?

Put another way: at what all-core clock speeds (for each) is equal performance reached between the AMD Ryzen 9 5950X (16C/32T} and the Intel i9-10900K (10C/20T)?

For reference, on my Sager NP9672M-G0 [Clevo X170KM-G] {Intel i9-10900K (10C/20T) variant} I can hit a max all-core speed of: 4.7 GHz using around 200W sustainably on the edge of but not hitting thermal throttling; where any higher and thermal throttling kicks in.

The focal point of my question is this: how does a X GHz all-core speed on 16C/32T at 65W/88W PPT (XMG APEX 15 MAX (E22) [Clevo NH5#VR]) compare to a Y GHz all-core speed on 10C/20T at 125W/~200?W boost (XMG Ultra 17 (M20) [Clevo X170KM-G])?

For context, I do data analysis work, so I'm trying to gauge the all-core (aka multicore) performance comparison between these two laptops and their respective desktop CPUs; as I am trying to figure out if there is an increase in all-core performance in purchasing the XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} |65W/88W PPT| over my existing Sager NP9672M-G0 [Clevo X170KM-G] {Intel i9-10900K (10C/20T) variant} |125W/200W boost|.

 

Links:
1) **XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} Announcement Link: https://www.reddit.com/r/XMG_gg/comments/ujn04y/launch_extended_ryzen_5000_cpu_support_in_xmg/

2) **XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} Benchmark Overview Link: https://docs.google.com/spreadsheets/d/e/2PACX-1vTJdELoyS9gEPIJCb4_y7-d4TlV5V_skKFkzUQFnGH84szj5JdpfA1yxFByUi0cxrJE8igObxrJ6RJ3/pubhtml

3) ***AMD Ryzen 9 5950X All-Core Speed Link: https://www.kitguru.net/components/cpu/luke-hill/amd-ryzen-9-5950x-zen-3-cpu-review/all/1/?PageSpeed=noscript

4) ***Intel i9-10900K All-Core Speed Link: https://en.wikipedia.org/wiki/Comet_Lake_(microprocessor)

 

You would probably be better off with the new 15" model that takes Intel chips if you make the change and if your programs can make use of the efficiency cores. Both should give you an at least 10-15% higher multi core over the X170 not because it is a superior design but because in their infinite wisdom Clevo decided to not give their best chassis (X170) a proper upgrade to socket 1700 where it could then sport 25k+ cinebench R23 scores with the current generation and more later. Some users already got the Clevo with the Intel socket and and @win32asmguy is planning to put a 12900k in there, might want to check out that thread:

 

Personally unless you want to play with new stuff I do not think that this is a good time to go for a replacement as the 15" units sucks compared to the X170 except for portability and even there Clevo went out of their way to make things worse than they had to be. Making two power bricks mandatory is bullshit if you mostly use the CPU and I hope they will change that assinine design in the future.

 

So if you have to then better look at the scores of both the new Intel and AMD 15" Clevo unit and compare them but I would wait until the end of the year for a true X170 successor that would do a lot better due to superior cooling and the possibility to use up to 32 core intel processors on it:

https://videocardz.com/newz/intel-showcases-13th-gen-core-raptor-lake-cpu-with-24-cores-and-32-threads

And maybe there will even be for the first time a big DTR with Thunderbolt and the ability to hold even higher core AMD chips of the next generation.

 

 

 

 

[1610ftw]

On 5/20/2022 at 9:28 PM, jaybee83 said:

i dont get it...the 5950X is SO frugal when it comes to power, can completely stretch its legs at 125W with stock clocks (https://www.techpowerup.com/review/amd-ryzen-9-5950x/19.html).

 

so why cripple it at 88W?!?! they already did that with the previous gen 3950X DTRs from Clevo, and now the same mistake again.... 125W is really not that much, a proper DTR should be able to take care of 150W for the CPU easily, if not more.... (as an example, the 5950X in the techpowerup review oced to 4.5 ghz all core draws 194W in multi threaded)

Looks to me like Clevo is still heavily biased towards Intel. Look at their new top of the line BGA-book that is Intel only - they will probably sell about 10 of those but who cares if Intel pays for it. Hopefully all of that will pay off with a proper 1700 socket 17"+ unit in the future. 

[electrosoft]

1 hour ago, 1610ftw said:

Looks to me like Clevo is still heavily biased towards Intel. Look at their new top of the line BGA-book that is Intel only - they will probably sell about 10 of those but who cares if Intel pays for it. Hopefully all of that will pay off with a proper 1700 socket 17"+ unit in the future. 

 

This. Clevo had a golden opportunity to really apply all those great power limits and cooling to a much less power hungry and better CPU (before 12th) with the 5000 series and did nothing with it. If they took the hardware of the X170SM/KM and made and had made an AMD variant? I would have been all over it. Hopefully they do something with AM5 better than trying to shoehorn a 5950x in what is available now and fight an uphill battle and still never achieve proper 5950x performance even with all the tweaks.

 

 

[1610ftw]

3 hours ago, electrosoft said:

 

This. Clevo had a golden opportunity to really apply all those great power limits and cooling to a much less power hungry and better CPU (before 12th) with the 5000 series and did nothing with it. If they took the hardware of the X170SM/KM and made and had made an AMD variant? I would have been all over it. Hopefully they do something with AM5 better than trying to shoehorn a 5950x in what is available now and fight an uphill battle and still never achieve proper 5950x performance even with all the tweaks.

 

 

 

 

Both Intel and AMD are coming out with extrmely powerful stuff so hopefully Clevo or others will be able to do something with that.

It is about time that somebody - anybody - is pulling out some of the stops that have hampered DTR development over the last years. They used to be a lot more powerful and now with more power hungry hardware we are supposed to get by with a smaller power envelope? This can only lead to frustration and ultimately the complete demise of the true DTR.

[electrosoft]

On 6/16/2022 at 2:35 AM, Taraquin said:

Using curve optimizer you can get quite good scores with 5950X even at 88W PPT. I agree that they should have made higher cap possible, but you can get quite good scores at 88W and neg CO 🙂 If it's possible to lower soc, vddg, vddp etc you can free up aditional performance for cores aswell. 

 

Unfortunately that will still only get you so far limited at 88w trying to shoehorn a 5950x and leave you at the mercy of varying boosts depending on load (which can be good or bad). On my 5800x desktop after optimizing CO and PBO after the Ageisa update that neutered it I ended up going classic static OC for consistency and stability. Still, it is better than nothing if you're trying to extract optimal performance of the 5950x in a severely underpowered chassis.

 

X170'esque chassis would handle a 5950x with ease. I wish Clevo would offer a 17.3" DTR but as @1610ftw said hopefully Clevo will offer something compelling with AMDs new lineup along with 13th gen.

 

 

[Taraquin]

4 hours ago, electrosoft said:

 

Unfortunately that will still only get you so far limited at 88w trying to shoehorn a 5950x and leave you at the mercy of varying boosts depending on load (which can be good or bad). On my 5800x desktop after optimizing CO and PBO after the Ageisa update that neutered it I ended up going classic static OC for consistency and stability. Still, it is better than nothing if you're trying to extract optimal performance of the 5950x in a severely underpowered chassis.

 

X170'esque chassis would handle a 5950x with ease. I wish Clevo would offer a 17.3" DTR but as @1610ftw said hopefully Clevo will offer something compelling with AMDs new lineup along with 13th gen.

 

 

On my 5600X allcore speed went from 4.35 to 4.6 using -30 neg CO. Comparable allcore OC is 4.6GHx which consumes around 76W, but I miss out on the 4.85 SC boost so in general I would say CO is superior to allcore OC, especially if you game.

[electrosoft]

2 hours ago, Taraquin said:

On my 5600X allcore speed went from 4.35 to 4.6 using -30 neg CO. Comparable allcore OC is 4.6GHx which consumes around 76W, but I miss out on the 4.85 SC boost so in general I would say CO is superior to allcore OC, especially if you game.

5600x is actually one of the recommended CPUs for the 15" XMG model and most likely the most optimal one to use.

 

Depending on the game you use it will vary how hard it hits the CPU and I don't like fluctuations when gaming as loads can vary that is why for my 5800x I just went static all core @ 4.7 as I found it bouncing all around with dips below 4.7 at times. A game like BFV (or similar) that can hit the CPU decently hard and on multiple cores would most likely run better fixed and OCd but a game that hits it lighter and SC would not. Basically, It is a case by case and I don't like that. I like to know where my CPU will be and where it will stay.

 

I wanted to like CO but that -30 CO would crash and burn under real loads unless you have a golden of golden chips. Once I actually started to dial in my CO profile and run real loads or even games like WoW and observed the CPU in action I would see too many fluctuations between ~4.9 and 4.5 especially in raids and PvP. I mean in the overall scheme of things it isn't a deal breaker or maker either way but I prefer consistency and stability and just scaling cores under CB23 would BSOD my 5800x so I had to start lowering my CO and even with a golden chip for true stability I found CO was sitting in the -5 to -15 on some nowhere near -30 which really was for very light loads. Even through it all I liked my Tomahawk X570 + 5800x + B-Die combo but the USB ports and the issues (which are still a problem for many users to this day) was a deal breaker for me as I work extensively with external media and to have them just time out up to the day I sold it and moved on was unacceptable.

 

AMD is promising decent OC and boosts on their next gen chips so we will see what they bring to the table. I'm hoping its something juicy because the baby cores approach from Intel stacking on more e-cores is not too exciting.

 

 

[Ishayin]

On 5/20/2022 at 3:50 PM, NBTUser said:

Links:
1) **XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} Announcement Link: https://www.reddit.com/r/XMG_gg/comments/ujn04y/launch_extended_ryzen_5000_cpu_support_in_xmg/

2) **XMG APEX 15 MAX (E22) [Clevo NH5#VR] {AMD Ryzen 9 5950X (16C/32T} variant} Benchmark Overview Link: https://docs.google.com/spreadsheets/d/e/2PACX-1vTJdELoyS9gEPIJCb4_y7-d4TlV5V_skKFkzUQFnGH84szj5JdpfA1yxFByUi0cxrJE8igObxrJ6RJ3/pubhtml

3) ***AMD Ryzen 9 5950X All-Core Speed Link: https://www.kitguru.net/components/cpu/luke-hill/amd-ryzen-9-5950x-zen-3-cpu-review/all/1/?PageSpeed=noscript

4) ***Intel i9-10900K All-Core Speed Link: https://en.wikipedia.org/wiki/Comet_Lake_(microprocessor)

 

Thanks for the reference links! Following that rabbit hole led me to enjoying watching 3DAndStuff's latest laptop modding adventure which is both rather interesting and informative on topics like VRMs, nickel plated heat sinks, and stabilised liquid metal TIMs:

https://www.youtube.com/watch?v=Q50QU8a776E

 

[jaybee83]

here is to hoping that there is indeed a worthy X170 DTR successor from Clevo aside from the ghastly X270 recently published 😅 but hey, maybe thats just the all Intel version, lets hold out for an all AMD or AMD+Nvidia socketable version? 😛 

[Taraquin]

On 6/18/2022 at 1:27 AM, electrosoft said:

5600x is actually one of the recommended CPUs for the 15" XMG model and most likely the most optimal one to use.

 

Depending on the game you use it will vary how hard it hits the CPU and I don't like fluctuations when gaming as loads can vary that is why for my 5800x I just went static all core @ 4.7 as I found it bouncing all around with dips below 4.7 at times. A game like BFV (or similar) that can hit the CPU decently hard and on multiple cores would most likely run better fixed and OCd but a game that hits it lighter and SC would not. Basically, It is a case by case and I don't like that. I like to know where my CPU will be and where it will stay.

 

I wanted to like CO but that -30 CO would crash and burn under real loads unless you have a golden of golden chips. Once I actually started to dial in my CO profile and run real loads or even games like WoW and observed the CPU in action I would see too many fluctuations between ~4.9 and 4.5 especially in raids and PvP. I mean in the overall scheme of things it isn't a deal breaker or maker either way but I prefer consistency and stability and just scaling cores under CB23 would BSOD my 5800x so I had to start lowering my CO and even with a golden chip for true stability I found CO was sitting in the -5 to -15 on some nowhere near -30 which really was for very light loads. Even through it all I liked my Tomahawk X570 + 5800x + B-Die combo but the USB ports and the issues (which are still a problem for many users to this day) was a deal breaker for me as I work extensively with external media and to have them just time out up to the day I sold it and moved on was unacceptable.

 

AMD is promising decent OC and boosts on their next gen chips so we will see what they bring to the table. I'm hoping its something juicy because the baby cores approach from Intel stacking on more e-cores is not too exciting.

 

 

I haven`t noticed fluctuations in performance when gaming with pbo+co. Lows are the same as static, but I get slightly better overall performance due to higher single core. Could be a few games where it matters, but haven`t noticed it in any of the games I play (SOTTR, TW3, CP2077). Within the same power limit my CO+PBO is superior to allcore OC on my setup, but I`m quite lucky and can run -28x1, -29x1 and -30x4 on my 5600X. both 4.6 allcore and my current pbo (it runs at 4.6 avg) uses around 75-76W, but I miss out on 250MHz singel core using allcore oc which lowers avg fps in some games like SOTTR.

[electrosoft]

8 hours ago, Taraquin said:

I haven`t noticed fluctuations in performance when gaming with pbo+co. Lows are the same as static, but I get slightly better overall performance due to higher single core. Could be a few games where it matters, but haven`t noticed it in any of the games I play (SOTTR, TW3, CP2077). Within the same power limit my CO+PBO is superior to allcore OC on my setup, but I`m quite lucky and can run -28x1, -29x1 and -30x4 on my 5600X. both 4.6 allcore and my current pbo (it runs at 4.6 avg) uses around 75-76W, but I miss out on 250MHz singel core using allcore oc which lowers avg fps in some games like SOTTR.

 

Did you ever test your chip to see how it rates? (Plat, Gold, Silver). I know also when tuning my brothers B550M + 5600x I went with CO vs static because his workloads are very light and at most he plays fortnite with a 1650 so for him he will most likely experience the nice sc boosts but he wanted a smaller form factor so I built him an mATX Gigabyte system and tuning CO coupled with the 240mm I put in there and the system usually stays near dead silent for everything he does and the GPU is loudest component. He's had it for over a year now and loves it but he said it is running a smidge louder and laggy so I'll give it a yearly tune up. But when I had my 5800x system running WoW shadowlands (my main game) I would hit raids and PvP and see that ~4.9 boost drop to 4.7 or lower and at that point there was no advantage especially after I went back in with CB23 to root out varying core instabilities under load and found quite a few core loads (let's say 3 cores) would crash under load but 1 or 8 wouldn't. It would depend on the load and how many cores were being pushed. If you're ok with a dirty OC and instabilities that are there but don't affect your workloads that works too.

 

It definitely does depend on the tasks at hand. You can run CB23, start with a single threaded run and slowly start adding in threads per 10 min runs and you will find instabilities quickly that will force you to back off of aggressive CO offsets quickly and/or your clocks will tank.

 

In the end, whatever works for you I just prefer consistent clocks and stability in all scenarios but whatever gets you those fps! 🙂

 

 

 

 

[Taraquin]

On 6/20/2022 at 6:45 PM, electrosoft said:

 

Did you ever test your chip to see how it rates? (Plat, Gold, Silver). I know also when tuning my brothers B550M + 5600x I went with CO vs static because his workloads are very light and at most he plays fortnite with a 1650 so for him he will most likely experience the nice sc boosts but he wanted a smaller form factor so I built him an mATX Gigabyte system and tuning CO coupled with the 240mm I put in there and the system usually stays near dead silent for everything he does and the GPU is loudest component. He's had it for over a year now and loves it but he said it is running a smidge louder and laggy so I'll give it a yearly tune up. But when I had my 5800x system running WoW shadowlands (my main game) I would hit raids and PvP and see that ~4.9 boost drop to 4.7 or lower and at that point there was no advantage especially after I went back in with CB23 to root out varying core instabilities under load and found quite a few core loads (let's say 3 cores) would crash under load but 1 or 8 wouldn't. It would depend on the load and how many cores were being pushed. If you're ok with a dirty OC and instabilities that are there but don't affect your workloads that works too.

 

It definitely does depend on the tasks at hand. You can run CB23, start with a single threaded run and slowly start adding in threads per 10 min runs and you will find instabilities quickly that will force you to back off of aggressive CO offsets quickly and/or your clocks will tank.

 

In the end, whatever works for you I just prefer consistent clocks and stability in all scenarios but whatever gets you those fps! 🙂

 

 

 

 

I think it was platinum in hydra (3 cores could run 5.1GHz+ on 1.35v, 1 5GHz), 2 was in the 4.9 territory. Age 4 and Cyberpunk is the most CPU demanding I play. I have tuned ram (4000cl16 1t B-die) so it will have more CPU headroom than a 5600X running 3600 xmp. 

 

Ryzen 5k is very temp sensitive, as long as temps are below 70C it will run allcore about 150MHz faster vs 80C. Static OC tends to give more stable temps, while CO+PBO give more fluctuations, if a certain temp threshold is reached using PBO+CO allcore will drop and I can see that this may be beneficial in certain games that hit allcore heavy. Early patches of CP2077 used some avx2, that was a true CPU killer. 

[amirbahalegharn]

if only clevo make a good laptop capable of 125W cpu for AM5 platform,just knowing that AMD will support that socket for 4-5 Years means with bios updates(official,unofficial) ,you can really get a perfect laptop in terms of upgrading path in Processor thus making it much easier and distinguishable to masses... ryzen 7100 or 7600x will be so different in performance compared to ryzen 10100 or 10700x.... the only challenger & competitor for this X86 laptop in 2025 -at least in terms of processing performance- , are those which will be equipped with ARM processors rivals of apple silicon.