overclocked 1950x Leak "Otherworldly Performance"

T

tangoseal

[H]F Junkie
Joined
Dec 18, 2010
Messages
9,743
Another WCCFTECH article.... if you read it (with a grain of salt of course) they are simply reporting what someone else is reporting.

In two days time we will see, but pending this is accurate then the silicon is matured a little more and apparently Threadripper Silicon can hit 4.0-4.1ghz at a stunningly low voltage of 1.25v (4.0g) - 1.4v (4.1g)

A quote from the article.... "Not even the world record Cinebench R15 run of the 7900X at 5.96GHz with liquid nitrogen cooling is enough to surpass Threadripper’s score."

Granted we are only seeing comparison between a 10 core Intel and a 16 core AMD. Of course there are 6 more cores to pound out more score. But considering they cost exactly the same the AMD is obliterating the Intel part in both performance per dollar and "holy shit factor".

Screenshot from Article ...
Geekbench score is obnoxiously high!
AMD-Ryzen-Threadripper-1950X-overclocked-4.1GHz.jpg


Cinebench posting a 3337 score (remember the threadripper on LN2 @ 5ghz hit 4000+)
AMD-Ryzen-Threadripper-1950X-overclocked-4.0GHz.jpg


Another quote from the Wccftech article:

"The Ryzen Threadripper 1950X is by far the most powerful desktop processor we have seen tested to date and once overclocked its performance jumps from the realm of very impressive to, there’s no better word for it, otherworldly. The embargo on Threadripper reviews goes up this upcoming Thursday and we can’t wait to see what else this chip can do"'

Link to original article:
http://wccftech.com/amd-threadripper-1950x-oced-tested-otherworldly-performance/
 
tangoseal said:
Care to clarify?

Are you saying it bulldozes the competiton, or are you comparing it to the failure of Bulldozer? A little ambiguity in your statement mate.
Click to expand...

I don't think Bulldozer was a failure. Certainly not for people who use their computers with threaded workloads.

The term "ThreadDozer" is not only humorously ironic but in a sense is very appropriate.
 
It's so good to see AMD back in the seat. I'm enjoying watching the flames roar under Intels ass (their fanboys as well).
 
tangoseal said:
Care to clarify?

Are you saying it bulldozes the competiton, or are you comparing it to the failure of Bulldozer? A little ambiguity in your statement mate.
Click to expand...

i wouldn't exactly call bulldozer a failure, did it not match the competition? sure. but the architecture design worked the way it was suppose to and it was an interesting outside the box attempt at multi-threaded computing. so i still tip my hat to them for trying something different. i mean if you really thing about it ryzen's still a cousin architecture to bulldozer with it's 2x4core CCX's vs bulldozers 4x2 core modules.
 
I think at this point no one takes stuff from WCCFTECH seriously. I would believe it if it said that Threadripper did the dishes and mows the lawn and folded laundry as well ;) .

Just a disclaimer , not saying you can't reach these scores but just saying that it is prolly not valid for a 24/7 setup ....
 
Pieter3dnow said:
I think at this point no one takes stuff from WCCFTECH seriously. I would believe it if it said that Threadripper did the dishes and mows the lawn and folded laundry as well ;) .

Just a disclaimer , not saying you can't reach these scores but just saying that it is prolly not valid for a 24/7 setup ....
Click to expand...

Believe? What you are saying is akin to .... um.... A video camera captures a murder on tape but no one in the jury believes the camera. The camera didn't make up the story. It only reported what it saw. In the case of this particular set of benchmarks and claims, ftech didn't claim anything. They are just reporting what they came across.

I don't get what everyone has against wccftech, I mean some of it I get, but really on the level you are claiming, sigh!
 
It's ok, this is posted on tpu as well and over there there were 5 pages of intel boys having a huge fit. Unpossible!!!?
 
Are there any situations where a high clocked 10 core cpu would outperform a lower clocked 16 core? I mean where the 16 generally outperforms the 10, would there be exceptions?
 
When the first Ryzen Blender benchmarks came out, my two12c/24t workstation were beating the Ryzen chip's score by about 5 and 6 seconds, respectively. The Ryzen was clocked at 3.4 GHz and my two workstations are clocked at 2.93 and 2.7, respectively. Now that Ryzen processors can be overclocked to about 4 GHz and improvements have been made to the memory sub-system, I'd guess that it would do the renders faster by a second or two (using the same binary from December of last year).
 
But the real question is, can it run Crysis??

Can't wait until tomorrow's embargo lift! Might be my next pc =)
 
ir0nw0lf said:
Someone needed to hand out pacifiers in that thread over there... :D
Click to expand...

Good god .... you hit the nail on the head. I've never seen more piss poor crybaby sucklings ever like over at tpu. We have our fair share but hot damn they're like all little infants over there.
 
Asia might come out first as it will be Aug. 10th there in 6 hours. Just have an NDA on the NDA lift.
 
Well I hope Amazon ships tomorrow morning... I paid a whopping 6 dollars for next day air lol...Amazon prime love it.

Anyways I hope I get my Threadripper this weekend. Im tired of looking at my ghetto ass Caselabs tray as a test bench ITX temp setup. I stripped my cases down getting ready for the monster build. This Asus 1080ti Poseidon is literally screaming to be under water.

Btw I heard that the Hyper 212 evo pictured below will cool a Threadripper? I dont know about that shit now lol

20170808_000431.jpg
 
I felt ashamed bolting my 212 onto Ryzen. It'll do for now but methinks threadripper would boil that thing
 
ecmaster76 said:
I felt ashamed bolting my 212 onto Ryzen. It'll do for now but methinks threadripper would boil that thing
Click to expand...

they should make an adapter: TR4 to 2x AM4 bracket. So you can use 2 am4 coolers to keep the threadripper temperate :)
 
tangoseal said:
Believe? What you are saying is akin to .... um.... A video camera captures a murder on tape but no one in the jury believes the camera. The camera didn't make up the story. It only reported what it saw. In the case of this particular set of benchmarks and claims, ftech didn't claim anything. They are just reporting what they came across.

I don't get what everyone has against wccftech, I mean some of it I get, but really on the level you are claiming, sigh!
Click to expand...

It is not my fault they post a lot and a lot of it tends to not get close to what you end up with. People tend to see headlines and not look at it from a perspective that is within reason , for overclocking results it is worse because you can do a benchmark run on voltage you will never run 24/7 and on cooling solutions where fan speed is screaming your ears off, not to say those are not valid I'm saying that the expectations derived from this content is what WCCFTECH tends to do to much.
 
mnewxcv said:
Are there any situations where a high clocked 10 core cpu would outperform a lower clocked 16 core? I mean where the 16 generally outperforms the 10, would there be exceptions?
Click to expand...

Think any Ryzen CPU released so far versus a 7700k for gaming.

Anything that's capable of using the extra cores will run away, anything that's thread-limited will run faster on a CPU with fewer but faster cores.
 
IdiotInCharge said:
Think any Ryzen CPU released so far versus a 7700k for gaming.

Anything that's capable of using the extra cores will run away, anything that's thread-limited will run faster on a CPU with fewer but faster cores.
Click to expand...

Am I right to say most thread limited applications are limited to 1, 2, or 4 cores? Applications that if a 10 core were faster than the 16, then a quad core would likely be faster than the 10 in. I'm hoping with the recent and upcoming emergence of high core count desktop cpus that programmers of games and other software will start taking advantage of them. No sense in having a 32 thread cpu being your gaming performance bottleneck.
 
This is nit cool. So I did this preorder thing. Amazon was like yeah... released on the 10th... nope my orders page just sitting on well email you a shipping date once we know. So much for bs advertising Amazon. Now well see if they ship by next Friday. And to think I was dissing new egg and some newegg shoppers are getting shipping notifications
 
tangoseal said:
This is nit cool. So I did this preorder thing. Amazon was like yeah... released on the 10th... nope my orders page just sitting on well email you a shipping date once we know. So much for bs advertising Amazon. Now well see if they ship by next Friday. And to think I was dissing new egg and some newegg shoppers are getting shipping notifications
Click to expand...
You didn't learn from Ryzen apparently lol
 
mnewxcv said:
Am I right to say most thread limited applications are limited to 1, 2, or 4 cores? Applications that if a 10 core were faster than the 16, then a quad core would likely be faster than the 10 in.
Click to expand...

Essentially, but it's not a static, broad-brush concept: it's going to differ not just from application to application, but even workload to workload. This is less true for games than it is for content creation, but it's still prevalent in say the difference in CPU usage between a game's single-player component and its multiplayer component with VOIP and streaming added on top.

The single-player component might be fastest with a 'faster' (IPC x clockspeed) CPU with relatively fewer cores, while the multiplayer scenario might be fastest with a 'slower' CPU with relatively more cores, even though the game itself is running a little bit slower. Essentially, if you have as many cores as your task can saturate, you need a 'faster' CPU to realize any gains, and if you have not saturated your CPUs cores, then adding cores can help.

[QUOTE="mnewxcv, post: 1043155129, member: 156147"I'm hoping with the recent and upcoming emergence of high core count desktop cpus that programmers of games and other software will start taking advantage of them. No sense in having a 32 thread cpu being your gaming performance bottleneck.[/QUOTE]

We're all thinking/hoping the same thing, but the challenge is that progress on this area, necessary as it is, is slow going because it just isn't easy. That's why Intel's higher performance consumer CPUs are still seen as effective for high-end gaming, as one can't (yet) have both.
 
You must log in or register to reply here.
Back
Top