Thursday, June 13th 2019

AMD's Upcoming $750 Ryzen 9 3950X (16C, 32T) Shown Beating Intel's $2,000 i9-9980XE (18C, 36T)

When we said AMD was readying a presentation on their Ryzen 9 3950X CPUs to awe crowds at E3, we weren't thinking of something of this magnitude. But apparently, it's true: a Geekbench test result has shown AMD's $750, 16 core, 32 thread Ryzen 9 9 3950X beating Intel's 18 core, 36 thread $2,000 i9-9980XE monster. Now, you may be thinking: ok, it beat it because of AMD's announced 4.7 GHz boost, and did so only on single threaded performance, obviously... but you would be wrong.

The Geekbench scores show AMD's Ryzen 9 3950X delivering 5,868 points in single, and 61,072 points in multicore workloads. Intel's i9-9980XE, on the other hand, scores just 5,391 single core, and 46,876 multicore points (on average and at stock clocks of 3,000 MHz base and 3,400 MHz boost). This is an incredible performance difference (particularly in the multicore score), and was apparently done with an engineering sample for AMD's upcoming chip that didn't even run at its announced 4.3 GHz base and 4.7 GHz boost clocks, but at 3.3 GHz and 4.3 GHz respectively. AMD's 105 W TDP, 16 core chip beats Intel's 185 W TDP, 18 core one... Where has the world come? Take the usual dosage of NaCl, and let's keep things in perspective - even if AMD's Ryzen 9 3950X equals, and doesn't beat, Intel's i9-9980XE, it's still a huge win for the red company. Almost as big a win as that huge stone on Lisa's hand.
Sources: Tom's Hardware, Intel i9-9980XE GeekBench score example, GeekBench Ryzen 9 3950X test result
Add your own comment

70 Comments on AMD's Upcoming $750 Ryzen 9 3950X (16C, 32T) Shown Beating Intel's $2,000 i9-9980XE (18C, 36T)

#26
Hossein Almet
If it's true, I may be tempted to upgrade, instead of waiting for Zen 3.
Posted on Reply
#27
TheGuruStud
CrackongThe target isn't "3950x beating 9980xe" cause a reasonable person would know Geekbench wasn't a good benchmark for high core count CPUs at all.
They put this up to bait the reviewers out there to test 3950x against 9980xe right at launch day.
I bet the 9980xe will hold its position but the 3950x will be really close , something like 9960x < 3950x < 9980xe .

Consider their prices, all these benchmark reviewers will put the Intel 9th Gen HEDT lineup into "Overpriced trash" territory.
Hey, if you can afford the CPU and MB, then you can afford the 300 dollar water loop to cool it, too! Intel wins!
Posted on Reply
#28
Crackong
TheGuruStudHey, if you can afford the CPU and MB, then you can afford the 300 dollar water loop to cool it, too! Intel wins!
3950x + GODLIKE = 749 + 799(assumed) + 300 = $1848
9980xe = $1999

With the money of a 9980xe alone, you can buy 3950x+ unnecessary high end MB + water loop and still have money for 32GB of RAM
Posted on Reply
#29
SIGSEGV
I am ready, ready, ready, ready...
I don't think I am gonna upgrade my X470 :lovetpu:
Posted on Reply
#30
TheGuruStud
Crackong3950x + GODLIKE = 749 + 799(assumed) + 300 = $1848
9980xe = $1999

With the money of a 9980xe alone, you can buy 3950x+ unnecessary high end MB + water loop and still have money for 32GB of RAM
Pretty funny that for 1,100 you can have a 3950x/MB/ram, which means an entire high end system with video card is the same price as just a 9980xe. Get bent, Intel.
Posted on Reply
#31
Xzibit
SIGSEGVI am ready, ready, ready, ready...
I don't think I am gonna upgrade my X470 :lovetpu:
More for everyone else...

AMD Robert HallockThe new feature allows you to override the default boost clock by up to +200MHz, in addition to expanded TDC/EDC/PPT limits. The CPU will self-manage from there using its built-in boost and clock management programming.

Ryzen doesn't really have a "single core turbo" clock. Our boost algorithm pursues the highest possible clocks on as many cores as possible until you hit some sort of limit: socket power, core temps, VRM electrical limit, VRM thermal limit, max clockspeed, etc.
Posted on Reply
#32
SIGSEGV
XzibitMore for everyone else...

arrrrghh...
d*mn you. Please stop...
Posted on Reply
#33
bszalman
XzibitMore for everyone else...

How much that's gonna cost?
Posted on Reply
#34
phanbuey
haha wow they're actually ahead of schedule this time.

Im in austin and the local shops here always make fun of what a sh*tshow AMD launches typically are (Ryzen very much included) - it's extra funny considering HQ is down the street. Where the local Fry's had a bunch of chips on hand and no motherboards for 2 weeks after launch.
Posted on Reply
#35
Berfs1
I am amazed at how many ignorant people commented on this post. You wanna know the conditions of the ram? LOOK AT THE SOURCES LMFAO. Wanna know what cooling? Generally speaking it doesn’t matter. For the 9980XE specifically, it runs 3800 MHz at all cores, and idk what the turbo boost is for the 3950X under load. What I will mention is the 3950X score is interesting as it’s memory is clocked at 2063 MHz effective speed, which is way below AMD’s 3200 MHz rating. And it was running X470..
XzibitMore for everyone else...

Weird flex but okay. I say that because you are flexing with shittabyte motherboards.
Posted on Reply
#36
Assimilator
God DAMN I wish people would stop using Shitbench as a performance metric for anything. The only thing that will objectively prove or disprove Ryzen 2's superiority over Intel's HCC CPUs are hard reviews by TPU, AnandTech, etc.
Posted on Reply
#37
Vycyous
Berfs1What I will mention is the 3950X score is interesting as it’s memory is clocked at 2063 MHz effective speed, which is way below AMD’s 3200 MHz rating. And it was running X470..
2063 MHz is 4133 MT/s (DDR4-4133). What is truly interesting is that AMD has shown that performance may actually be reduced by running the memory at anything over about 3600 to 3733 MT/s (1800 to 1866 MHz).
Posted on Reply
#38
ratirt
Maybe Geekbench isn't the perfect benchmark to show the numbers but it still shows something. If the difference is small between the 2 processors I'd agree that it can go either way which one wins in real time and other benchmarks and workloads. The difference is enormous and if the GB doesn't scale well with multicores then it is still advantage for AMD since less cores beat intel's more core product. Of course other reviews would be nice and I'm sure we will get them some time around. We will all be welcome to comment afterwards :)
Posted on Reply
#39
phanbuey
Vycyous2063 MHz is 4133 MT/s (DDR4-4133). What is truly interesting is that AMD has shown that performance may actually be reduced by running the memory at anything over about 3600 to 3733 MT/s (1800 to 1866 MHz).
latency, not performance. You're still going to get the bandwith but increased latency at the high end strap.
Posted on Reply
#40
Vayra86
diatribeI hesitant to believe this just yet. The 18 core Intel should be a least negligible faster than the AMD 16 core counterpart.
Unlikely, the single score of Ryzen is now considerably higher as well. About 5% - per core. So it may be a hair faster on multicore loads, but only if those scale very well.

And let's face it, we've always maintained that those single scores matter a whole lot more than a slightly higher multi,... Intel has nothing on this. Especially if you consider price and the now also relevant damage to HT.
Posted on Reply
#41
Vycyous
phanbueylatency, not performance. You're still going to get the bandwith but increased latency at the high end strap.
According to Ian Cutress at AnandTech, "This ratio should automatically come into play around DDR4-3600 or DDR4-3800, but it does mean that IF2 clock does reduce in half, which has a knock on effect with respect to bandwidth. It should be noted that even if the DRAM frequency is high, having a slower IF frequency will likely limit the raw performance gain from that faster memory. AMD recommends keeping the ratio at a 1:1 around DDR4-3600, and instead optimizing sub-timings at that speed."

If that's truly the case, it's really unfortunate because I was hoping for great memory overclocks and the additional bandwidth and lower latency that usually comes along with that, not one or the other. Speaking of latency, it doesn't appear that it has been improved at all. I run at 3600 MT/s CL14 with my 2700X and latency is generally around 55-60 ns in Aida64, which is about as good as it seems to get with Ryzen. The same 3600 MT/s CL14 (with closely matched sub-timings) on my 7700K has about 33-38 ns of latency, which is a huge difference. I guess we'll find out for sure in a few weeks.
Posted on Reply
#42
xtreemchaos
every dog has its Day, and about time to. looks one hell of a chip i carnt wait for christmas.
Posted on Reply
#43
PanicLake
diatribeI hesitant to believe this just yet. The 18 core Intel should be a least negligible faster than the AMD 16 core counterpart.
The number of core is in no way a measure of performance.
Like those stupid advertisement for smartphone (at least where I live) that say : octa core bla bla bla
Octacore alone doesn't mean anything. Same goes for CPU core count.
Posted on Reply
#44
Konceptz
ArchStupidYou thinking that constitutes any sort of evidence and warrants defense makes you an AMD fanboy and just as bad.
Sorry but no, I just find it hilarious the lengths certain folks go to defend Intel. I buy and build from both camps.
Posted on Reply
#46
Jism
From ZEN1 AMD already had a really good base that only needed improvement, to finally beat the best counterpart of Intel. They succeeded. Case closed.
Posted on Reply
#47
phanbuey
VycyousAccording to Ian Cutress at AnandTech, "This ratio should automatically come into play around DDR4-3600 or DDR4-3800, but it does mean that IF2 clock does reduce in half, which has a knock on effect with respect to bandwidth. It should be noted that even if the DRAM frequency is high, having a slower IF frequency will likely limit the raw performance gain from that faster memory. AMD recommends keeping the ratio at a 1:1 around DDR4-3600, and instead optimizing sub-timings at that speed."

If that's truly the case, it's really unfortunate because I was hoping for great memory overclocks and the additional bandwidth and lower latency that usually comes along with that, not one or the other. Speaking of latency, it doesn't appear that it has been improved at all. I run at 3600 MT/s CL14 with my 2700X and latency is generally around 55-60 ns in Aida64, which is about as good as it seems to get with Ryzen. The same 3600 MT/s CL14 (with closely matched sub-timings) on my 7700K has about 33-38 ns of latency, which is a huge difference. I guess we'll find out for sure in a few weeks.
I hope that's not true... makes sense what he's saying though. I figured the IF bandwidth was high enough that even at 1:2 it wouldn't degrade bandwidth from RAM.
Posted on Reply
#48
windwhirl
I find hilarious that the uploader of these benchmark results is called "blueleader". Oh, the irony...
Posted on Reply
#50
RichF
Unless the latest Intel hardware mitigates against the security flaws and their performance regressions then people should take those into full account for all of these comparisons as well.

For example, having to disable hyperthreading, which several companies have said is necessary for full security, is a big blow to some of Intel's parts. Some people will respond by saying they're not that worried about security. Well, maybe they aren't but some are and should be.
Posted on Reply
Add your own comment
Oct 31st, 2024 19:34 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts