Intel Xeon W-3175X 28-core Processor Now Available at $2,999

[Deleted member 158293]

Asus Dominus REX , 2000$.
And that's the only one available.

Yeeeeah, that actually makes more sense... that is one complex motherboard...

Especially given the fact per the Anandtech article I just finished reading Asus only seemed to have gotten an order for 500 of these.

You know... thinking about it, it still seems oddly cheap, hard to recoup any costs spread over just 500 units!?

 

[NicklasAPJ]

Yeah no. I would take a 2990WX over this any time of the day. Cheaper and more powerful than this and runs great on most Linux distros.

Is cheaper, but not more powerfull.

But agree, the price is too much, even for me.

I dont wanna paid 1000 Dollers for a mortherbord + that cpu.

Im wating for next gen Intel cpus instead, or if AMD show something that can smash 7980 XE to my use.

 

[efikkan]

Yea there is optimized for Intel or to be specific "crippled for AMD".

That's not an intelligent answer based in facts, but FUD.

There is so many cases where AMD and Intel perform about same but in some other cases AMD is suddenly horrible and Intel isnt. And its not core count or core speed either. Not talking about AMD losing some 20%, where its expected, since one AMD core at 4GHz isnt equiv of one Intel core at 4GHz, but cases where AMD suddenly has 50% less performance for no reason.

"for no reason"? It's called a bottleneck.
You have certain characteristics tied to the architecture, and then there is certain characteristics tied to the model, such as Threadripper 2970WX and 2990WX, where the core configuration can be causing high latency which affects some workloads more than others.

 

[Deleted member 158293]

Can't really compare the 2990WX to anything in Windows until the Microsoft fixes the scheduler.

Until then it's mostly a Linux CPU.

 

[efikkan]

Can't really compare the 2990WX to anything in Windows until the Microsoft fixes the scheduler.

And who's fault is that?
Should Microsoft now really add bloat to their kernel to workaround CPU makers' "design flaws"?

 

[xkm1948]

Can't really compare the 2990WX to anything in Windows until the Microsoft fixes the scheduler.

Until then it's mostly a Linux CPU.

It is 2990WX, W means workstation. It is MEANT for Linux and workstation only. Any researcher with a tiny bit of self-respect would not use a consumer version of Windows 10 for workstation level work. Just saying.

 

[M2B]

Programmers/developers can design an application which favours a specific architecture or even a specific CPU more, but they simply don't care about these things.
As efikkan said, it's just a myth.
X86 cores are simply designed to do anything you throw at them.
It's AMD's fault as if an application doesn't favour their architecture.

 

[Sasqui]

And it runs on a socket 775 platform no doubt.

 

[Smartcom5]

…
The "optimized for Intel" claim is a myth.

Yup, it's a myth which totally wasn't settled by real-world hard cash in magnitudes of billions. … oh wait!
Will you do your homework? Please!


Smartcom

 

[efikkan]

Yup, it's a myth which totally wasn't settled by real-world hard cash in magnitudes of billions. … oh wait!

Will you do your homework? Please!

This old BS again? You go do your homework. Nothing really uses Intel's own compiler.
A lot of software uses Microsoft's compiler, but more and more uses the much more advanced gcc and llvm compilers.

 

[Patriot]

And who's fault is that?
Should Microsoft now really add bloat to their kernel to workaround CPU makers' "design flaws"?

It has been shown earlier schedulers worked better than the current microsoft one, when there is an update, and a subsequent degradation of performance... YES it is the OS manufacturers fault.
Especially when other OS's just work from the beginning and much better.

As for Intel compiler issues... Intel has it's compiler everywhere and had to add a disclaimer to certain benchmarks due to the explicit, suck on AMD commands.
https://www.ftc.gov/sites/default/files/documents/cases/091216intelcmpt.pdf
That cost them a Billion or so.

It's time to ditch fanboyism and look at the facts.
Don't whore yourself out to intel or amd, they don't care about you, just want to make money.
Intel just has a habit of cheating when it can't win.



None of this matters, this chip isn't being sold, maybe maybe 1000 of them and not to us...
Took them over a year to bin enough 8180s to not be "caught" in a fake product.
As a historical overclocker... I would still love to play with one with that 1700w chiller, I just wouldn't want to pay the power bill...
I mean... harbringer did pull 1450w from the wall running f@h @ 3.3ghz on 48c. I don't know what it pulled @ 3.8ghz running cine 9.5.
At ~3.2ghz this passes up what the equivalent intel machine could do performance wise, 4p Sandybridge... using 750w.

This chip, at stock non-turbo clocks uses more than a 64 core rome chip will use... How the tables have turned.
Performance at high power sucking clocks should be interesting for workstation workloads, many render engines can't use more than 64 threads.
AKA some workloads will run better on this than dual 8180s or server chips... even if it uses more power than dual chips.
Curious on actual benchmarks and not speculation.

 

[efikkan]

It has been shown earlier schedulers worked better than the current microsoft one, when there is an update, and a subsequent degradation of performance... YES it is the OS manufacturers fault.

No, it's AMD's fault that they launched a product where some cores have substantially higher latency than others.

This Threadripper generation will soon be replaced anyway. It's unreasonable for Microsoft to add bloat to their codebase for two CPU SKUs.

As for Intel compiler issues... Intel has it's compiler everywhere and had to add a disclaimer to certain benchmarks due to the explicit, suck on AMD commands.

No. The software we see in benchmarks are not compiled with Intel's compiler. It has really not been popular since the 90s, nearly no one in the industry uses it.

 

[cdawall]

I've read a review on this new CPU from Anandtech and it was a fairly interesting read.. AMD wins in some areas, Intel in others. One thing I read was the price $3000... Then I heard the CPU requirements and power usage for stock, then overclocked and whilst I care little about power usage, stock is rated at 255w, Asus then did a turbo version which was 510w.. Now at stock speeds it never reached that high but overclocked it did require a little help with the water to keep it cool.. And also the max temp of the CPU was reported to be 120c (auto shut down)

Take a look here....

Some quotes from Anandtech page I found interesting -

At 4.3 GHz, we were hitting over 500W peak load (confirmed by wall meter), which is the limit of the cooling setup provided. Compared to the 4.0 GHz result, we calculated that the CPU actually used 17% more power overall to get a 7% increase in performance.

Here the power is overall a bit lower, but we can see that the score isn’t rising much at 4.0 GHz, again due to our CPU temperature sensor showing 110ºC very easily. In this instance, the power consumption between 3.9 GHz and 3.6 GHz increased by 14%, while the score rose 10%.

Intel sent an EKWB Phoenix cooler which is rated for much higher power consumption, but arrived too late for our testing. We’re planning on doing an overclocking review, so this should help. But what our results show is that when Intel showed that 5.0 GHz demonstration using a water chiller they really did need it. Users might look into investing in one themselves if they want this chip.

And then the topping on the cake...

But What About That 5.0 GHz? How Much Power?
We took some of our benchmark values for power and frequency, extrapolated them with a power curve, and we estimate that at 5.0 GHz, this chip is likely to be drawing in excess of 900W, perhaps as high as 1200W. Yes, Intel really did need that 1700W water chiller.

But in all seriousness, if you have a 28 core CPU, don't you expect to have a high power draw?? Maybe it's just me but common sense to me is - more cores, more speed, more heat....

As Anandtech points out, it's more of a case of a non held back performance CPU... Question is, who would buy one?

At 4.2ghz my 2990wx pulls over 850w at the wall. That's through a seasonic prime platinum. Subtract 150w for the rest of the system as a good guess for load wattage. That was sustained. It peaked higher.

 

[xkm1948]

At 4.2ghz my 2990wx pulls over 850w at the wall. That's through a seasonic prime platinum. Subtract 150w for the rest of the system as a good guess for load wattage. That was sustained. It peaked higher.

How's temp for your XSPC block on the 2990WX

 

[Darmok N Jalad]

Threadripper 2 based on Zen2 and AMD’s chiplet design will likely fix the flaws with Threadripper 1.5’s 4 CCX design. Once that memory passthrough issue is solved for half the cores, the scaling issues should go away. Or at least that’s the theory, anyway. It would certainly seem worthwhile to wait if you can, as the TDP of this Intel chip is insane.

 

[hat]

I heard these two guys were interested in buying:

 

[Deleted member 158293]

It is 2990WX, W means workstation. It is MEANT for Linux and workstation only. Any researcher with a tiny bit of self-respect would not use a consumer version of Windows 10 for workstation level work. Just saying.

No arguement from me there... gaming is the only use I have left for running a windows session.

 

[lexluthermiester]

This is interesting;

Steve's going to do a live stream as well in a little while.

 

[R0H1T]

Yeeeeah, that actually makes more sense... that is one complex motherboard...

Especially given the fact per the Anandtech article I just finished reading Asus only seemed to have gotten an order for 500 of these.

You know... thinking about it, it still seems oddly cheap, hard to recoup any costs spread over just 500 units!?

I wonder if they ought to be gold plated, considering the costs

 

[Deleted member 158293]

Looks like De8bauer pulled over 1225 watts over just one 24 pin connector + auxiliary?!

I wonder if they ought to be gold plated, considering the costs

Well... there's an odd little small display for a kinda useless animation, so why not have some real 24k gold bling on there? Or you could just straight up buy about 1.5 oz of gold for the same price.

 

[Argyr]

The fact you are identifying 'AMD and Intel 'folks' is already so wrong on so many levels to begin with.

How about letting that go, and taking a fresh approach to this, considering we are all consumers first and foremost. Has it occurred to you that people switch 'sides' all the time, too? Look around a bit and you will see. What's more interesting here is the reasoning behind those switches. As is the reasoning behind favoring one brand or the other. Who knows, maybe at some point we'll learn something.

I really value your usual unbiased pacifist wisdom.
I've been a PC enthusiast for 24 years. I had the chance to taste all the different flavors and you have to make up your mind after a while. I made up my mind: AMD will be nr.2 underdog forever to me. They had the chance to impress me but failed to do so. I had long and prosperous decades with my usual setup.

To widen the perspective a bit: We're living in a consumerist age and there's just too many manufacturers to pick of. You have to make up your mind, mostly based on experience: I had a whirlpool fridge and dishwasher die on me. Never going to buy Whirlpool again

 

[medi01]

• Intel Turbo Boost Technology 2.0 that delivers frequencies up to 4.3 GHz.

Isn't this that uber chip that was demoed at 5Ghz in summer, cough?

Ayup, 2990WX at Newegg right now for $1,729 vs this.

Yep, and that not even including hidden costs such as 1.5k for Intel mainboard.

ntel defense force? Strange, it's quite the opposite.

Sure thing, bro.
Nobody. Not like there are people defending outrageous nvidia pricing or not like any Anadntech article on AMD gets flooded with team green (very apparently paid) shills spitting all sorts of FUD. One of them, under nick chizow, was particularly "impressive".

We're living in a consumerist age and there's just too many manufacturers to pick of.

Yeah. Intel, AMD, something that is still producing that Citrix (or were they called Shmitrics? I don't remebmer) CPUs.
Plenty of choices.
And on GPU side of things: greedy green, red underdog and blue "we'll be there in a couple of years". So hard to choose

 

[Darmok N Jalad]

Isn't this that uber chip that was demoed at 5Ghz in summer, cough?

Yep, and that not even including hidden costs such as 1.5k for Intel mainboard.


Sure thing, bro.
Nobody. Not like there are people defending outrageous nvidia pricing or not like any Anadntech article on AMD gets flooded with team green (very apparently paid) shills spitting all sorts of FUD. One of them, under nick chizow, was particularly "impressive".


Yeah. Intel, AMD, something that is still producing that Citrix (or were they called Shmitrics? I don't remebmer) CPUs.
Plenty of choices.
And on GPU side of things: greedy green, red underdog and blue "we'll be there in a couple of years". So hard to choose

The 28C 5.0ghz demo was pure madness. Intel used a 1HP, 1700W chiller on the CPU, and the VRM array had a massive heatsink with 4 high-speed fans on it. An interesting proof of concept, I suppose, but process tech will need to come a long way for something like that to be economical.

https://www.anandtech.com/show/12907/we-got-a-sneak-peak-on-intels-28core-all-you-need-to-know

 

[Smartcom5]

…
No. The software we see in benchmarks are not compiled with Intel's compiler. It has really not been popular since the 90s, nearly no one in the industry uses it.

Then, oh so wise guy, please explain me the following: Why on earth even today software is shipped being compiled with such compilers while featuring given flags? … while runnign significant faster even on today's AMD-hardware, if such flags are removed?
You got any viable explanation for this? … I mean, besides another personal vendetta fueled by a fanboy's blind hatred?

I'm goddamn serious here, explain it please! It still affecting even today's software running on actual AMD hardware.


Smartcom

 

[bajs11]

It is rated for a 500 W TDP, thanks to the use of a triple 120 m radiator

hmm is that a 120 meter radiator?