• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The Intel Xeon W-3175X Review: 28 Unlocked Cores, $2999

Also the TDP listed on CPU’s is for stock clocks ignoring any turbo frequencies.

Scanning through this thread I'm amazed that people are still getting TDP wrong after all this time - it is only loosely related to power draw and especially for Intel CPUs:

Thermal Design Power (TDP) represents the average power, in watts, the processor dissipates when operating at Base Frequency with all cores active under an Intel-defined, high-complexity workload. Refer to Datasheet for thermal solution requirements.

Intel's TDP measurement is the amount you need to dissipate for the CPU to work correctly under its normal design parameters - it is not the power draw.

AMD's "TDP" measurement used to more closely reflect their power draw (hence Typical Board Power nomenclature on their GPUs) but I believe with their latest CPUs it is the actual TDP.

And if the thermal and power use of the CPU is really of concern to you then handily Intel makes a datasheet with more information:

hodvsiY.png


If you do the maths on that it puts the stock all core max load at potentially ~380 watt draw.

EDIT: GN's interview does state that:

"At full all-core frequency in that higher power mode, we don’t reach 510W, but we certainly go well beyond 255W, scoring about 380W maximum. If we apply this to the Intel Spec version, and compare to other CPUs, we get the following:"
 
Last edited:
Intel putting 380 Watts on the TDP rating, which is the true TDP rating just doesn't have the same kudos as 255 watts.... ;)
 
Yes that's correct, its basically the total heat generated by the CPU however I need to stress it is for stock operation.
For Intel stock turbo yes its applicable, the turbo frequencies are completely tied to the TDP as to never exceed it.

For all core turbo (which is nothing to do with Intel, it is the motherboard vendor's choice) you will be throttling if your cooling solution isn't up-to the job and the limits aren't removed. The TDP obviously will be way off the charts at this point...


Scanning through this thread I'm amazed that people are still getting TDP wrong after all this time - it is only loosely related to power draw and especially for Intel CPUs:



Intel's TDP measurement is the amount you need to dissipate for the CPU to work correctly under its normal design parameters - it is not the power draw.

AMD's "TDP" measurement used to more closely reflect their power draw (hence Typical Board Power nomenclature on their GPUs) but I believe with their latest CPUs it is the actual TDP.

And if the thermal and power use of the CPU is really of concern to you then handily Intel makes a datasheet with more information:

hodvsiY.png


If you do the maths on that it puts the stock all core max load at potentially ~380 watt draw.

EDIT: GN's interview does state that:

"At full all-core frequency in that higher power mode, we don’t reach 510W, but we certainly go well beyond 255W, scoring about 380W maximum. If we apply this to the Intel Spec version, and compare to other CPUs, we get the following:"
 
Yes that's correct, its basically the total heat generated by the CPU however I need to stress it is for stock operation.
For Intel stock turbo yes its applicable, the turbo frequencies are completely tied to the TDP as to never exceed it.

For all core turbo (which is nothing to do with Intel, it is the motherboard vendor's choice) you will be throttling if your cooling solution isn't up-to the job and the limits aren't removed. The TDP obviously will be way off the charts at this point...

I'm not a fan personally of the way Intel approach this they seem to be deliberately making it hard for people to make an informed comparison.
 
Hmm, silly question time. When/how/why would you ever dissipate less/more power than you draw?

My physics is terrible, btw. I just thought CPUs turned all the mcguffins into heat, as there's no other output?
 
Hmm, silly question time. When/how/why would you ever dissipate less/more power than you draw?

My physics is terrible, btw. I just thought CPUs turned all the mcguffins into heat, as there's no other output?

They are closely related but what Intel is stating is the level of power dissipation required for normal operation i.e. a heatsink with 255 watt capability will allow its turbo system to work within intended limits i.e. time duration before throttling, etc. even if it is pulling 300+ watt at the socket during that time period.

In no way it Intel stating with its TDP measurements - as people are implying - that the CPU will only ever draw upto 255 watt of power as standard - their own datasheet shows the nature of the power draw if that is really important to you (though you do need to research the voltages for your intended application and calculate the wattage from the amperage).

I'm not defending Intel here their approach to it is terrible but people are berating them for something that they've decided Intel are doing when it isn't the case.
 
Hmm, silly question time. When/how/why would you ever dissipate less/more power than you draw?

My physics is terrible, btw. I just thought CPUs turned all the mcguffins into heat, as there's no other output?

Think of the power draw as the top level (this is way over simplified) i.e. what starts at the power headers, you then have loses in the traces, VRM's, pins, LGA package etc... until it reaches the CPU die (every step has resistance), some power is used by the CPU the rest is just heat which should remain within the TDP set by Intel at stock operation give or take.

The moment you override anything (clock, voltage, turbo settings, load line etc...), the official TDP is meaningless.
 
Think of the power draw as the top level (this is way over simplified) i.e. what starts at the power headers, you then have loses in the traces, VRM's, pins, LGA package etc... until it reaches the CPU die (every step has resistance), some power is used by the CPU the rest is just heat which should remain within the TDP set by Intel at stock operation give or take.

The moment you override anything (clock, voltage, turbo settings, load line etc...), the official TDP is meaningless.

That is true but the problem is people are berating Intel for claiming the CPU only uses upto 255 watt which isn't a claim Intel is making at all - the TDP is the level of dissipation required for correct stock operation and the length of time stock turbo functionality will run fully spun up, etc. the datasheet above shows that the CPU is quite capable of going way beyond 255 watt power draw with all core operation potentially hitting ~380 watt at the socket and turbo capable of around 500 watt.

On the other hand it is not a great approach for getting a feel for power draw for comparison purposes even with Intel's own CPUs let alone other brands where AMD's approach gives a better ball park for actual power use under normal circumstances.
 
That is true but the problem is people are berating Intel for claiming the CPU only uses upto 255 watt which isn't a claim Intel is making at all - the TDP is the level of dissipation required for correct stock operation and the length of time stock turbo functionality will run fully spun up, etc. the datasheet above shows that the CPU is quite capable of going way beyond 255 watt power draw with all core operation potentially hitting ~380 watt at the socket and turbo capable of around 500 watt.

On the other hand it is not a great approach for getting a feel for power draw for comparison purposes even with Intel's own CPUs let alone other brands where AMD's approach gives a better ball park for actual power use under normal circumstances.

Its impossible to make power comparisons between two completely different architectures, way to many variables involved.
Given the target audience of this product, power draw shouldn't even be a consideration... this is only the beginning it will be commen place in the next 12-24 months for CPU's to exceed 500-600W+ of power, core counts are going to go up... power per core is staying constant give or take(multi-die designs improve it slightly)... if you want cores at high clocks you have to pay for them.
 
TDP is directly proportional to Power. Always. >90% of Power drawn by a processor is turned into heat by the processor, where the losses in the VRM and copper are pretty minimal.

TDP is defined by the manufacturer, and is some fraction of the functional power draw of a processor under a "nominal" load.
It would make sense if the TDP was the target power draw used by the processor's adaptive clocking as a power target, but I've never seen evidence that this is the case.

I prefer the common GPU review metrics of performance per watt, average power and peak power. All you need is these over a spread of workloads and you can actually do a comparison.
 
TDP is directly proportional to Power. Always. >90% of Power drawn by a processor is turned into heat by the processor, where the losses in the VRM and copper are pretty minimal.

TDP is defined by the manufacturer, and is some fraction of the functional power draw of a processor under a "nominal" load.
It would make sense if the TDP was the target power draw used by the processor's adaptive clocking as a power target, but I've never seen evidence that this is the case.

I prefer the common GPU review metrics of performance per watt, average power and peak power. All you need is these over a spread of workloads and you can actually do a comparison.

People generally talk about TDP as if it was the maximum (stock) power draw from the PSU of the CPU but Intel stopped using it like that a long time ago for instance when a cooler is used that matches or exceeds the rated TDP it will allow short time turbo bursting as per the described Intel behaviour during which the CPU will draw way more power from the socket than the rated TDP.

Not saying I agree with the way Intel is dealing with the information but people are berating Intel for their perception of what Intel is claiming versus what Intel is actually claiming:

intel said:
"The processor TDP is the maximum sustained power that should be used for design of the processor thermal solution. TDP is a power dissipation and component temperature operating condition limit, specified in this document, that is validated during manufacturing for the base configuration when executing a near worst case commercially available workload without AVX as specified by Intel for the SKU segment. TDP may be exceeded for short periods of time or if running a very high power workload."

Further confused by the fact that if you want a sustained all core boost you will go way above the rated TDP in both thermal solution and correlation with power draw which is all in the datasheet.

Personally don't really care about power draw unless it is in the context of a portable device but it is another measure of how lacking a product might be when the overall performance for that product isn't good.
 
https://wccftech.com/intel-cascade-...il-cascade-lake-x-to-follow-at-computex-2019/

if i remember correctly, Cascade Lake-X is 22 cores. What the Aorus X299 Master and MSI Meg were created for

https://www.msi.com/Motherboard/MEG-X299-CREATION

My basket at Overclockers UK:
Total: £400.49 (includes shipping: £10.50)​

makes sense to launch after the W-3175X .... could put a lot of people off buying it for cheaper price
 
Back
Top Bottom