News AMD's Ryzen 9 5950X Shatters PassMark Records

[Endymio]

Does no one proofread their work these days? So many grammatical errors.
Quality of the site keeps diminishing.

Hey, what's a few possessive "it's" errors between friends?

 

[EridanusSV]

Till actual benchmarks show up, these passmark results with 1 CPU of unknown settings should be ignored. Everything else you'll compare it to would have thousands of results averaged

Till actual race shows up, these SSC Tuatara results with 1 run on the unknown desert of Nevada should be ignored. Everything else you'll compare it to would have thousands of results averaged. My Prius could probably defeat it.

 

[Endymio]

Till actual race shows up, these SSC Tuatara results with 1 run on the unknown desert of Nevada should be ignored.

Nice try, but the Tuatara run was observed by an independent organization, under publicly-published and tightly-controlled conditions. I think saying we should ignore entirely this benchmark is a bit strong, but the results certainly should be taken with a boulder of salt.

 

[EridanusSV]

Does it even matter? We've been here before with the 3900X and 3950X. We're going to know the outcome on this you intel simps

 

[fakerboy80]

There's a Ryzen 5 5950x???

Oh boy, i'm willing to upgrade to 5000 series, Getting a new motherboard, Ram. And also a new PC case.

Since i have an AMD Phenom II X4 970 BE with an RX 5700 XT.

 

[Arbie]

Forget about g-unit111. Take my money!

 

[Flayed]

There's a Ryzen 5 5950x???

Oh boy, i'm willing to upgrade to 5000 series, Getting a new motherboard, Ram. And also a new PC case.

Since i have an AMD Phenom II X4 970 BE with an RX 5700 XT.

That will be a huge performance increase

 

[spongiemaster]

Don't forget that Zen 3's is 19% higher than Zen 2. Zen 2's IPC was already 7-10% higher than Sky Lake derivative CPUs.

IPC is very application specific. 10% is the upper bound for 3000 series vs Comet Lake, not the norm. 3000 IPC is consistently worse than Comet Lake in games.

 

[Grindayzier]

This should be death blow to intel.
Waiting for the same to Nvidia.

I'm Intel/Nvidia fan by the way

 

[helper800]

IPC is very application specific. 10% is the upper bound for 3000 series vs Comet Lake, not the norm. 3000 IPC is consistently worse than Comet Lake in games.

You can have higher IPC but lower performance in single-threaded applications like games because the lower IPC processor (intel) is running at a clock speed that more than makes up for its lower IPC. You seem to be confusing single-threaded performance with IPC. Ryzen 3000 is about 5-10% higher IPC than Intel's comet lake, however, comet lake runs about 15% higher clock speeds on average in comparison to more than make up the difference.

 

[Juan_Bijero]

Until I see independent reviews of the new AMD cpus and video cards, I'm not willing to argue about what is true or what isn't true. If AMD knocks it out of the park, then more power to them! It feels like we finally have some real competition. That said, I have already made my bed, so I will sleep in it with no regrets.

 

[Conahl]

IPC is very application specific. 10% is the upper bound for 3000 series vs Comet Lake, not the norm. 3000 IPC is consistently worse than Comet Lake in games.

and they get that performance cause of clock speeds and the use of more power.

You can have higher IPC but lower performance in single-threaded applications like games because the lower IPC processor (intel) is running at a clock speed that more than makes up for its lower IPC. You seem to be confusing single-threaded performance with IPC. Ryzen 3000 is about 5-10% higher IPC than Intel's comet lake, however, comet lake runs about 15% higher clock speeds on average in comparison to more than make up the difference.

exactly. a lot of people dont seem to realize this. clock comet lake and ryzen 3000 at the same clock speeds, and see who is faster.

 

[spongiemaster]

You can have higher IPC but lower performance in single-threaded applications like games because the lower IPC processor (intel) is running at a clock speed that more than makes up for its lower IPC. You seem to be confusing single-threaded performance with IPC. Ryzen 3000 is about 5-10% higher IPC than Intel's comet lake, however, comet lake runs about 15% higher clock speeds on average in comparison to more than make up the difference.

No, I'm not confusing anything. Techspot did a comparison of multiple AMD Ryzen generations and I think it was a 9900k. All CPU's were set to 4GHz to test IPC, and the Intel chip came in first in every single game tested. No clock advantage for Intel, still won all of them. The 3000 series won most of the non-gaming tests as it does on average have higher IPC than any Skylake variant, but the overall victory was low to mid single digits. It definitely wasn't 10% on average.

 

[mdd1963]

The supposed 43% improvement (from 3rd gen to 4rth) in CS:GO frame rates was/is impressive!

 

[helper800]

No, I'm not confusing anything. Techspot did a comparison of multiple AMD Ryzen generations and I think it was a 9900k. All CPU's were set to 4GHz to test IPC, and the Intel chip came in first in every single game tested. No clock advantage for Intel, still won all of them. The 3000 series won most of the non-gaming tests as it does on average have higher IPC than any Skylake variant, but the overall victory was low to mid single digits. It definitely wasn't 10% on average.

Games are not the best metric for trying to figure out IPC improvements for a few reasons. Even that article you reference has this to say on that:
"You can’t really use games for measuring IPC gains but they’re still interesting to look at for clock-for-clock comparisons. One obvious issue with games is that they’re not always CPU bound and even when they are, the degree to which they are CPU bound can vary."
Every other 100% CPU bound benchmark test put the AMD processors clock-for-clock ahead of the Intel CPU's by about the aforementioned amount.

 

[Jim90]

No, I'm not confusing anything. Techspot did a comparison of multiple AMD Ryzen generations and I think it was a 9900k. All CPU's were set to 4GHz to test IPC, and the Intel chip came in first in every single game tested. No clock advantage for Intel, still won all of them. The 3000 series won most of the non-gaming tests as it does on average have higher IPC than any Skylake variant, but the overall victory was low to mid single digits. It definitely wasn't 10% on average.

These persistent variable level attacks on AMD are not going unnoticed. I'm guessing you'll be looking to change paymasters soon.

 

[Darkbreeze]

This should be death blow to intel.

That's funny as can be. Even if AMD started selling CPUs that had forty or fifty percent performance increases over the latest Intel CPUs, that would still not be true. They are simply too diverse.

 

[Groveling_Wyrm]

This should be death blow to intel.
Waiting for the same to Nvidia.

I'm Intel/Nvidia fan by the way

If it does become a death blow to Intel, it would mean that Intel is taking this laying down, and not investing/researching in new tech, which is unrealistic. Eventually Intel will come out with something. Will it be competitive? I don't know. It better be if Intel wants to keep their market share. They also need to work on their other lines. AMD is definitely pushing hard to get more market share.

For AMD to increase their production to gain more market share, they would have to talk to TSMC and get them to increase their production. TSMC is already maxed out on orders for the next 2-3 years from multiple sources, on all their die sizes, with all their money being re-invested in newer tech, and new fabs/retooling fabs to keep up. TSMC isn't expecting to increase production significantly in the 5 year plan they have. AMD can't go anywhere else to make their processors, either. With this problem, AMD will need extra cash, and the only place they can get that cash is from charging higher prices.

The last time AMD had a significant lead, Intel released the Core 2 processors and then the I series, and they wiped the floor with AMD, leading to a 10+ year time span where they didn't have any competition from AMD, until the Ryzen. Who knows what Intel will pull out of their hat in the next 5 years.

I will give you one thing though. This is putting increasing pressure on Intel to put up or shut up. I would expect that they intended the 5950x to compete with what they thought Intel would put out, but since Intel is having so many problems they are falling face first into this.

No question now, though. With the 5950x, AMD is the king of desktop processors. We just don't know for how long.

 

[geogan]

Well I just hope there is actual availability of this new 5950X and it's not like the NVidia 3080 shambles....

Hopefully the fact that you already have to have Ryzen 470/570X type motherboard, RAM etc setup already, and its a costly change/upgrade to set up in addition, will limit the competition for sales this time, not like GPUs where anyone with any sort of hardware setup can still use them.

 

[jeremyj_83]

No, I'm not confusing anything. Techspot did a comparison of multiple AMD Ryzen generations and I think it was a 9900k. All CPU's were set to 4GHz to test IPC, and the Intel chip came in first in every single game tested. No clock advantage for Intel, still won all of them. The 3000 series won most of the non-gaming tests as it does on average have higher IPC than any Skylake variant, but the overall victory was low to mid single digits. It definitely wasn't 10% on average.

Another issue with using games is that developers have put MASSIVE amounts of optimizations for the Core architecture in them. While running at the same clock speed works to a certain extent, depending on the CPU you will have some other differences. For one that article has both CPUs running 3200MHz RAM. While 3200MHz is the official spec for Ryzen, that is a memory over clock for Intel which official spec is only 2933MHz. In this case both CPUs are base clocked at 3.6GHz so the CPU over clock is the same, however, had they used the 3900X instead the Intel would have had a higher percentage over clock on frequency. Ideally what you want to do is let your CPUs run at their normal frequencies & official RAM spec and then normalize the numbers per GHz. When Ryzen 3000 was released, Anandtech did a study of the IPC using SPEC. While I don't like that the 9900K got the memory OC to 3200MHz, they did let the CPUs run their normal frequencies and then normalized the results per GHz. https://www.anandtech.com/show/14605/the-and-ryzen-3700x-3900x-review-raising-the-bar/6

 

[JBB_123]

A little history lesson. Back in June 2019, it was reported that the Ryzen 5 3600 posted a score of 2979 in single threaded performance. Currently, this same Ryzen 5 3600 cpu has a Passmark single threaded performance score of 2584.

The 3900X had a score above 3000 for a time as well, but its score is now at 2730.

AMD Ryzen 5 3600 Benchmark Leaked, Dominates Intel's i9 9900K in Single-threaded Performance

We're almost exactly a week away from the official launch of AMD's highly anticipated Ryzen 3000 series processors on July 7, and yet the leaks just keep pouring in from all over the techsphere. And this time around we've got another absolute whopper for ya. So let's dig in! AMD Ryzen 5 3600...

wccftech.com

Hoping that the scores being reported stay robust, but I would think that these drop a little as there is more sampling. Hopefully not by 300 or 400 points.

 

[spongiemaster]

Games are not the best metric for trying to figure out IPC improvements for a few reasons. Even that article you reference has this to say on that:
"You can’t really use games for measuring IPC gains but they’re still interesting to look at for clock-for-clock comparisons. One obvious issue with games is that they’re not always CPU bound and even when they are, the degree to which they are CPU bound can vary."
Every other 100% CPU bound benchmark test put the AMD processors clock-for-clock ahead of the Intel CPU's by about the aforementioned amount.

Me: 30% better IPC than Intel seems unlikely
You: No, that's where they should based on these unverified/unsourced numbers I'm using
Me: No, those numbers don't look accurate, I know it is slower in games for example
You: You don't know what you're talking about, you're wrong
Me: No, here are actual test results
You: Those numbers don't count because I don't want them to, no counter source provided

The obfuscation and denial of real world results just doesn't stop. The world is flat everyone. If all you got is personal attacks against me, then you got nothing. Barring something completely unexpected, the 5000 series is going to sweep pretty much, if not, all IPC tests here. Those results will count to me, just like the current ones do, and I bet they'll suddenly count for you too. We'll see how many of you put asterisks next to AMD's winning results.

 

[TerryLaze]

You can have higher IPC but lower performance in single-threaded applications like games because the lower IPC processor (intel) is running at a clock speed that more than makes up for its lower IPC. You seem to be confusing single-threaded performance with IPC. Ryzen 3000 is about 5-10% higher IPC than Intel's comet lake, however, comet lake runs about 15% higher clock speeds on average in comparison to more than make up the difference.

Games are not the best metric for trying to figure out IPC improvements for a few reasons. Even that article you reference has this to say on that:
"You can’t really use games for measuring IPC gains but they’re still interesting to look at for clock-for-clock comparisons. One obvious issue with games is that they’re not always CPU bound and even when they are, the degree to which they are CPU bound can vary."
Every other 100% CPU bound benchmark test put the AMD processors clock-for-clock ahead of the Intel CPU's by about the aforementioned amount.

Both of these post are excellent arguments on why this passmark result is irrelevant to like 99% of people.

TLDR:
We have no idea of the all core clocks that zen 3 will get
(if it's high it will also burn a lot more power if it is energy efficient it will be pretty low all core clocks)
For most people on the desktop platform games are the heaviest workload they will ever run.

Still no iGPU on most, if any, models which a lot of people think of as useless but it helps OEMs and it provides 4k playback, streaming and several other conveniences.

As already said by others TSMC is pretty solidly booked out so the amount of ZEN 3 that AMD can produce is going to be very limited (compared to the volume of intel at least)

AMD is going to do very well and is going to increase their profitability but there is no need to make it out as if intel is suddenly going to do badly.

We have three continuous years of both companies growing side by side.

 

[spongiemaster]

A little history lesson. Back in June 2019, it was reported that the Ryzen 5 3600 posted a score of 2979 in single threaded performance. Currently, this same Ryzen 5 3600 cpu has a Passmark single threaded performance score of 2584.

The 3900X had a score above 3000 for a time as well, but its score is now at 2730.

AMD Ryzen 5 3600 Benchmark Leaked, Dominates Intel's i9 9900K in Single-threaded Performance

We're almost exactly a week away from the official launch of AMD's highly anticipated Ryzen 3000 series processors on July 7, and yet the leaks just keep pouring in from all over the techsphere. And this time around we've got another absolute whopper for ya. So let's dig in! AMD Ryzen 5 3600...

wccftech.com

Hoping that the scores being reported stay robust, but I would think that these drop a little as there is more sampling. Hopefully not by 300 or 400 points.

What are you doing? You're not allowed to say anything that even resembles negative about AMD here unless you want to be labeled a paid Intel shill.

 

[jeremyj_83]

Still no iGPU on most, if any, models which a lot of people think of as useless but it helps OEMs and it provides 4k playback, streaming and several other conveniences.

The OEMs have access to the 4000G series which are 4c/8t, 6c/12t, and 8c/16t CPUs. All of those are available in 35W & 65W TDP variants. I think that most OEMs are happy with those choices. I would love to have those CPUs become available for retail purchase.