AMD Ryzen 7 5800X3D Retail CPU Gets First Independent Tests

[Airisom]

Well, the ram is running in single rank as well, so if they went up to say 3600 cl14 in dual rank, there should be some more performance gained. Still underwhelming, but apparently he's posting gaming benchmarks tomorrow and comparing against the 12900kf.

 

[trsttte]

Funny that "3200MHz CL14 RAM" is here noted as a downside, as reviewer could have used much quicker RAM.

That's Samsung B die. Was quite expensive, and can be easily overclocked to maximum Ryzen can even handle in 1:1. How many Ryzen users have quicker RAM? And how much does it even matter in benchmarks that do scale - as oposed to Cinebench, also most games outside fringe cases, min frames or ultra high framerate?

And no, "PBO +200MHz overclocking" never worked as AMD explained in that video. A lot of time and resources was spent to even get Ryzen 3000 processors to achieve marketed boost clocks (even for just a fraction of a second), forget achieving anything higher in a meaningful way.

As great as Samsung B die may be (I don't think it matters much anymore, it did in the Ryzen earlier days but currently there are plenty of other chips also doing higher speeds, albeit without so consistent timings) 3200 CL14 is still only 3200 CL14, the eventual OC potential doesn't really matter for the test result.

It's looking like a 5700x, in anything the cache doesn't actively help

A more expensive 5700x at that and with less overclocking options, it's another "brilliant" markting move by AMD where they over hyped a product/technology before having fool proof results to back it up (not saying the X3D is bad, but it doesn't seem to live up to what was promissed and shouldn't have been hyped so much)

 

[Deleted member 185088]

Single core Cinenench R23 (by my opinion most indicative of gaming speeds here) is crap, slower than stock 5800x. Something must be wrong, or this won't beat the 5600x in gaming.

I believe AMD said so, the goal was to improve gaming performance nothing else.

 

[DeathtoGnomes]

I'll reserve my opinion when more thorough reviews are done. These 'its just quick review just to get it out there first' reviews are FUD and should not be taken as serious, especially this one.

 

[TheLostSwede]

I'll reserve my opinion when more thorough reviews are done. These 'its just quick review just to get it out there first' reviews are FUD and should not be taken as serious, especially this one.

It does at least beat just having Geekbench numbers...

 

[5 o'clock Charlie]

The UEFI/AGESA has been available for about a month and new drivers are out, so highly unlikely.
The tests were done with 3200MHz CL14 RAM, so that could account for some of it, but not much.

Thanks again for posting more news on this CPU. I am unsure as to why the RAM would be the deficit in this case scenario. Both gaming and synthetic benchmarks from Gamers Nexus show 3200MHz CL14 RAM is one of the sweet spots for Ryzen, which is what I am currently using.
Hopefully tomorrow's gaming tests will provide more of a sneak preview. As others have said, let's wait and see how this chip performs once official on April 20th from other outlets.
I also wonder if VMs would have any benefit from the extra cache?

 

[Makaveli]

Funny that "3200MHz CL14 RAM" is here noted as a downside, as reviewer could have used much quicker RAM.

That's Samsung B die. Was quite expensive, and can be easily overclocked to maximum Ryzen can even handle in 1:1. How many Ryzen users have quicker RAM? And how much does it even matter in benchmarks that do scale - as oposed to Cinebench, also most games outside fringe cases, min frames or ultra high framerate?

And no, "PBO +200MHz overclocking" never worked as AMD explained in that video. A lot of time and resources was spent to even get Ryzen 3000 processors to achieve marketed boost clocks (even for just a fraction of a second), forget achieving anything higher in a meaningful way.

If PBO+200mhz doesn't work how am I hitting Single core clocks of 5.05Ghz with it on? And I also happen to be using 3200 CL 14 Samsung B die 4x8 dual rank!

 

[ThrashZone]

Hi,
At least run a gaming benchmark

 

[TheoneandonlyMrK]

Well, the ram is running in single rank as well, so if they went up to say 3600 cl14 in dual rank, there should be some more performance gained. Still underwhelming, but apparently he's posting gaming benchmarks tomorrow and comparing against the 12900kf.

If he sticks with one stick of ram they won't be worth much.

 

[TheLostSwede]

Thanks again for posting more news on this CPU. I am unsure as to why the RAM would be the deficit in this case scenario. Both gaming and synthetic benchmarks from Gamers Nexus show 3200MHz CL14 RAM is one of the sweet spots for Ryzen, which is what I am currently using.
Hopefully tomorrow's gaming tests will provide more of a sneak preview. As others have said, let's wait and see how this chip performs once official on April 20th from other outlets.
I also wonder if VMs would have any benefit from the extra cache?

Not the RAM itself as such, but if you use faster RAM, you normally increase the fabric clock, which improves overall performance, if only slightly.
It's also been proven that using Dual Rank memory on modern CPUs increases the performance somewhat.

If he sticks with one stick of ram they won't be worth much.

Single rank ≠ Single Channel.

 

[ThrashZone]

Hi,
There's more to tuning memory than just frequency.

Intel memory performance can alter CB results a lot.

 

[5 o'clock Charlie]

If PBO+200mhz doesn't work how am I hitting Single core clocks of 5.05Ghz with it on? And I also happen to be using 3200 CL 14 Samsung B die 4x8 dual rank!

Which RAM kit are you currently running? Thanks.
Edit: sorry, I missed the system specs on your profile. Is it specifically the NEO or regular RGB version? Not sure if mine 4x8 setup is dual rank. It doesn't specify on gskills site.

 

[Makaveli]

Which RAM kit are you currently running? Thanks.
Edit: sorry, I missed the system specs on your profile. Is it specifically the NEO or regular RGB version? Not sure if mine 4x8 setup is dual rank. It doesn't specify on gskills site.

G.SKILL Trident Z RGB Series DDR4 3200MHz CL14 1.35V Dual Channel Desktop Memory Kit (F4-3200C14D-16GTZR)

I didn't go for the Neo kit as I didn't find the extra cost worth it.

 

[Max(IT)]

Cinebench scores aren't really affected by RAM speed to any great degree.

not completely true. I tested several configurations and RAM speed is not very important but still counts for some difference.
BTW this is not the point here.
More cache won't help Cinebench and such applications at all.
This is a CPU dedicated to gaming...

Uninspiring and late, I would add...

 

[5 o'clock Charlie]

G.SKILL Trident Z RGB Series DDR4 3200MHz CL14 1.35V Dual Channel Desktop Memory Kit (F4-3200C14D-16GTZR)

I didn't go for the Neo kit as I didn't find the extra cost worth it.

Makes sense about the NEO. I assume mine is dual channel and dual rank since mine is basically the same kit without the rgb factor. Thanks.

 

[Makaveli]

Makes sense about the NEO. I assume mine is dual channel and dual rank since mine is basically the same kit without the rgb factor. Thanks.

99% of 8GB's sticks are single rank. So if you are in 4x8 you are in dual rank operation.

 

[ThrashZone]

Hi,
Higher cache frequency would help CB results but not amount.

 

[chrcoluk]

Single core Cinenench R23 (by my opinion most indicative of gaming speeds here) is crap, slower than stock 5800x. Something must be wrong, or this won't beat the 5600x in gaming.

cinebench has never been a gamer workload metric, its commonly used because its simple to run.

My 5600G with its 16meg cache been performant in my non gaming workloads is making more and more sense.

 

[AnotherReader]

cinebench has never been a gamer workload metric, its commonly used because its simple to run.

My 5600G with its 16meg cache been performant in my non gaming workloads is making more and more sense.

It would be worthwhile reading this article from Chips and Cheese. The graph most relevant to this discussion is reproduced below:

Cinebench is the opposite of many games as its L3 hitrate is high. This should be similar to your 5600G as the total L3 that one core can access is the same. The author also notes that

Therefore, AMD’s V-Cache is not a completely superior solution. Rather, it’s an acknowledgement that a one-size-fits-all approach to L3 size no longer cuts it. I suspect AMD will offer both V-Cache and non V-Cache variants of future chips, allowing customers to choose what’s best for their workload.

 

[freeagent]

Good thing I didn’t sell my 5900X to get this like I was going to a couple of months ago. They lost me at reduced clocks and no oc.

 

[QuietBob]

CPU-Z and R23 ST scores are suspiciously low on this sample, only 11% higher than my oc'd 3300X. It looks like the 5800X3D will need extensive tuning to shine - and likely a few driver/AGESA updates.

It would be worthwhile reading this article from Chips and Cheese. The graph most relevant to this discussion is reproduced below:

View attachment 242822

Cinebench is the opposite of many games as its L3 hitrate is high. This should be similar to your 5600G as the total L3 that one core can access is the same. The author also notes that

Thanks so much for this article, a real eye-opener!

 

[Deleted member 24505]

It had better be good for gaming as it's the only other ace in its bag.

 

[Aquinus]

So of course, people test synthetic benches that don't benefit from the cache.

Or they're testing the wrong applications. There are plenty of applications in the server space that really show much much it can help. Check out this and the next 2 pages on Phoronix. I would expect synthetics that use a significant amount of memory to benefit. Granted, EPYC gets a much larger cache, but it goes to show how much it can help.

AMD EPYC 7773X "Milan-X" Benchmarks Show Very Strong HPC Performance Upgrade Review - Phoronix

www.phoronix.com

 

[Konceptz]

Depending on what happens in the next 60-90 days. I might just end up with a 12700k with some 3600mhz DDR4 and call it a day.
The undeniable itch to go back to Intel after years AMD, and only a $40 price difference using 5800x vs 12700k current pricing on microcenter pretty much sums it up for me.

 

[Why_Me]

Depending on what happens in the next 60-90 days. I might just end up with a 12700k with some 3600mhz DDR4 and call it a day.
The undeniable itch to go back to Intel after years AMD, and only a $40 price difference using 5800x vs 12700k current pricing on microcenter pretty much sums it up for me.

AMD Ryzen 7 5800X Vermeer 3.8GHz 8-Core AM4 Boxed Processor - Heatsink Not Included - Micro Center

Get it now! Get the high-speed gaming performance of the world's best desktop processor. Encode faster. Render faster. Iterate faster. Create more, faster with AMD Ryzen processor.

www.microcenter.com

AMD Ryzen 7 5800X $319.99

Intel Core i7-12700 Alder Lake 2.1GHz Twelve-Core LGA 1700 Boxed Processor - Intel Stock Cooler Included - Micro Center

Get it now! 12th Gen Intel Core i7-12700 desktop processor. Featuring Intel Turbo Boost Max Technology 3.0 and PCIe 5.0 & 4.0 support, DDR5 and DDR4 support, 12th Gen Intel Core i7 desktop processors are optimized for gamers and productivity and help deliver high performance.

www.microcenter.com

Intel Core i7-12700 $319.99

Intel Core i7-12700 + Intel B660 Review

Today we have a combo review of all-new Intel hardware, our first look at the affordable Intel B660 platform, the non-K Core i7-12700 and Intel's latest box...

www.techspot.com