• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Poll: Ryzen 7950X3D, 7900X3D, 7800X3D

Will you be purchasing the 7800X3D on the 6th?


  • Total voters
    191
  • Poll closed .
New world record high average FPS with a 7950X3D and a 4090 In Modern Warfare 2, 423 FPS. :cool:

For reference with the 7950X3D my 7900 XTX scores 430 FPS.
The Warzone maps you did earlier. I make a mistake and did not steer my average/1% low on the plane. I dropped to 260fps then tried rebar off and managed 270fos average. I am not yet finished with Curve Optimiser, and haven’t touched memory as of now. I use your config file but man you are way ahead still. Will the memory make this much difference. Or have you some other secret settings I need to know?

Also how is your clock fixed at 5.2?
 
Last edited:
Can someone help clarify a few things for me about the differences between the 7800X3D & 7950X3D.
One of the upcoming games im interested in, Cities: Skylines 2, has announced it'll "take advantage of all the available processing power of the multicore CPUs." and its got me considering the benefits of the 7950X3D over the 7800X3D.

Am i correct in understanding the construction of the 7800X3D has direct L3 access to all 8 cores, whereas the 7950X3D is split between chiplets, one having 8 cores with L3, the other having 8 cores without (or 40mb type low amounts)?
If the CPU is accessing all 16 cores of the 7950X3D, does it lose the benefit of the 3D-vCache, and actually perform like the non-X3D model only with a lower clock speed - 4.2ghz v 4.5ghz?
Is my understanding close? If CS2 benefits greatly from having more cores, it basically falls into the 'productivity' category which is where the 7950X3D shines. Im thinking it has similar performance for gaming, and significantly better at productivity tasks.

Im not entirely sure the price can be justified for just one game, but if the 7950X3D in 99% of games runs on one CCD with L3 cache and your system processes and whatnot run on the other, i can see benefits in that but maybe its minimal gains?
 
Can someone help clarify a few things for me about the differences between the 7800X3D & 7950X3D.
One of the upcoming games im interested in, Cities: Skylines 2, has announced it'll "take advantage of all the available processing power of the multicore CPUs."

Best thing to do is wait for reviews (and the following performance analysis vids) of the game. A standard 7950x could even be faster here if all the game wants is cores and clockspeed.
 
Last edited:
The Warzone maps you did earlier. I make a mistake and did not steer my average/1% low on the plane. I dropped to 260fps then tried rebar off and managed 270fos average. I am not yet finished with Curve Optimiser, and haven’t touched memory as of now. I use your config file but man you are way ahead still. Will the memory make this much difference. Or have you some other secret settings I need to know?

Also how is your clock fixed at 5.2?
Memory will help for sure.

I've just got some dual rank DIMMs. Trying to figure out how much, if at all, are they faster vs my single rank kit.

Here are my current timings for dual rank, still in the process of lowering voltage but this gets roughly the same latency as my single rank kit at similar timings, albeit with higher SD/DD timings.
PNtRoY1.png

This kit will do 372 RFC though (120ns) whereas my single rank kit will only go 384 RFC (123 ns).
 
Last edited:
A standard 7950x could even be faster here if all the game wants is cores and clockspeed.
Yeah, thats what i was thinking for CS2, however i think for my use case the X3D range is where i'd see more benefit. Without having to care about exactly how it achieved it, i'd seen enough to know the 7800X3D & 7950X3D perform very similar in gaming, but with the CS2 news im trying to understand exactly where their strengths & weaknesses are to be able to evaluate them where benchmarks cant answer it, such as a pre-released game that should utilise & benefit from more cores.

Im hoping to build an AM5/4090 system for Starfield (early Sept release) so waiting for hands-on reviews of CS2 (late Oct) complicates matters. I've been putting it off due to a lack of interest in any games for a good while, thankfully its not like one is an outright bad choice, its just understanding whether the ~50% markup can be justified with my needs.
 
Yeah, thats what i was thinking for CS2, however i think for my use case the X3D range is where i'd see more benefit. Without having to care about exactly how it achieved it, i'd seen enough to know the 7800X3D & 7950X3D perform very similar in gaming, but with the CS2 news im trying to understand exactly where their strengths & weaknesses are to be able to evaluate them where benchmarks cant answer it, such as a pre-released game that should utilise & benefit from more cores.

Im hoping to build an AM5/4090 system for Starfield (early Sept release) so waiting for hands-on reviews of CS2 (late Oct) complicates matters. I've been putting it off due to a lack of interest in any games for a good while, thankfully its not like one is an outright bad choice, its just understanding whether the ~50% markup can be justified with my needs.

With a 4090 purchase, I assume you want to play at 4k? (If you don't don't, throw money into a 4090, it's a complete waste of money at 1440p and under)
At 4k what CPU you have becomes less relevant as you're pretty much always going to be GPU bound.
Intel 13th Gen or AMD 7000 you can't go wrong with either. Just don't buy a 6 core, you want at least an 8 core to keep the 4090 fed.
The X3D cache SKU's are even less relevant at 4k, The caches get the biggest boost at 1080p and to a lesser extent 1440p.
 
Last edited:
Yeah, thats what i was thinking for CS2, however i think for my use case the X3D range is where i'd see more benefit. Without having to care about exactly how it achieved it, i'd seen enough to know the 7800X3D & 7950X3D perform very similar in gaming, but with the CS2 news im trying to understand exactly where their strengths & weaknesses are to be able to evaluate them where benchmarks cant answer it, such as a pre-released game that should utilise & benefit from more cores.

Im hoping to build an AM5/4090 system for Starfield (early Sept release) so waiting for hands-on reviews of CS2 (late Oct) complicates matters. I've been putting it off due to a lack of interest in any games for a good while, thankfully its not like one is an outright bad choice, its just understanding whether the ~50% markup can be justified with my needs.
With a 4090, you'll be wanting the 7950X3D for sure. It's just the faster of the two, particularly in a few RT heavy games.

You have a 4090, only right that you should pair it with the fastest gaming CPU tbh.
 
With a 4090 purchase, I assume you want to play at 4k? (If you don't don't, throw money into a 4090, it's a complete waste of money at 1440p and under)
At 4k what CPU you have becomes less relevant as you're pretty much always going to be GPU bound.
Intel 13th Gen or AMD 7000 you can't go wrong with either. Just don't buy a 6 core, you want at least an 8 core to keep the 4090 fed.
The X3D cache SKU's are even less relevant at 4k, The caches get the biggest boost at 1080p and to a lesser extent 1440p.
I hadnt actually considered resolution in any of it, it was more a case of the 4090 being something that'd last me a while and, obscene as its priced, considered the best bang for buck compared to the other 4000 series offerings, and I havent really considered the AMD cards yet.
I have the original Samsung G9, so dual 1440p. I typically run it as 2 1440p screens (2 cables in & out) but there are games i'll want to switch to the full resolution. Pixel count is close to 4K (7.2m vs 8.2m).
 
Memory will help for sure.

I've just got some dual rank DIMMs. Trying to figure out how much, if at all, are they faster vs my single rank kit.

Here are my current timings for dual rank, still in the process of lowering voltage but this gets roughly the same latency as my single rank kit at similar timings, albeit with higher SD/DD timings.
PNtRoY1.png

This kit will do 372 RFC though (120ns) whereas my single rank kit will only go 384 RFC (123 ns).
I was looking at the GSkill 64gb kits and wondering if I should try them also. You have some great latency. These 7950x3d respond very well to memory tuning.

I manage to get 320fps average on the new Warzone map. I’m trying to catch you but I have no chance without proper tuning of memory.
 
I hadnt actually considered resolution in any of it, it was more a case of the 4090 being something that'd last me a while and, obscene as its priced, considered the best bang for buck compared to the other 4000 series offerings, and I havent really considered the AMD cards yet.
I have the original Samsung G9, so dual 1440p. I typically run it as 2 1440p screens (2 cables in & out) but there are games i'll want to switch to the full resolution. Pixel count is close to 4K (7.2m vs 8.2m).
If you play games like warzone for example definitely go 7950x3d with either a 4090 or 7900xtx even if your 1440p each season performance crashes and it's a battle to achieve 200 FPS at some points. I have a gp850 with 180hz at 1440p and have a 7950x3d and a 4090 I haven't optimized my 7950x3d or touched the 4090 only the ram under expo and I get around 200-250 FPS. Maybe I need to learn / figure out how to optimize my 7950x3d but it's very confusing looking in the bios settings haha
 
I was looking at the GSkill 64gb kits and wondering if I should try them also. You have some great latency. These 7950x3d respond very well to memory tuning.

I manage to get 320fps average on the new Warzone map. I’m trying to catch you but I have no chance without proper tuning of memory.
Although not 100% guaranteed, it seems like a lot of the GSkill 6000C30 64GB and the GSkill 6400C32 64GB kits are A die. These two kits specifically:
  • F5-6000J3040G32GX2-TZ5RK
  • F5-6400J3239G32GX2-TZ5RK
You can get them both from a large USA retailer.

In my testing achieving 6200Mhz C28 is easy with 64GB on AM5 with these (A die) kits when using the right timings/voltages.

CL26 and RCD 35 is a bit ballsy and requires higher voltage, but CL28 and RCD 36 and RP 32 should be easy to achieve with these kits.
 
Last edited:
Although not 100% guaranteed, it seems like a lot of the GSkill 6000C30 64GB and the GSkill 6400C32 64GB kits are A die. These two kits specifically:
  • F5-6000J3040G32GX2-TZ5RK
  • F5-6400J3239G32GX2-TZ5RK
You can get them both from a large USA retailer.

In my testing achieving 6200Mhz C28 is easy with 64GB on AM5 with these (A die) kits when using the right timings/voltages.

CL26 and RCD 35 is a bit ballsy and requires higher voltage, but CL28 and RCD 36 and RP 32 should be easy to achieve with these kits.
I can get a deal on the 64gb 6400 kit. Would you say these are the better to go with over the single rank kits?

If it turns out not to be A-die I would be disappointed very much.
 
I can get a deal on the 64gb 6400 kit. Would you say these are the better to go with over the single rank kits?

If it turns out not to be A-die I would be disappointed very much.
It's hard to say, I don't think it make a big difference either way.

All things considered though, as the same or similar timings and same frequency, dual rank should be a little faster overall.

On AMD you'll be able to hit 6200Mhz on both regardless.
 
Back
Top Bottom