Intel keeps up the unethical SDP scam with “new” 4.5W parts [S|A]

[galego]

From the source:

Intel is back to their normal unethical PR ways with another set of disingenuously fake ‘low power’ SDP numbers. What lowers the bar yet farther is how they did it this time, it shows how little faith the company has in its own BS.

You can tell Intel really doesn’t want anyone asking questions about SDP because the first time they talked about it at CES they wouldn’t even mention that it was the fake SDP numbers rather than the real TDP that they kept mentioning in the keynote. It worked, everyone touted the lower number as a real transistor advance instead of an unethical marketing snow job. This time around they told the press about he new “4.5W” SKUs half an hour before the embargo went up. I am guessing the majority of the press didn’t see the ‘news’ until after the embargo went up, I was sitting in a room with most of Intel’s target audience at the time of the embargo and most didn’t know until I asked them about it. So Intel is trying surprisingly hard to keep people from asking questions about SDP because the company knows if people actually get educated about it the scam is over. So they give you half an hour or less to write-up the story or be left out. If you aren’t there when the embargo goes up, your competitors get all the page views. If you ask questions you are guaranteed to miss the embargo. If you do your job and get educated about the technical details, your story will never get read. It is a purposefully unethical game to keep the press from doing their job of informing our readers.

The saddest part is that if you look around, it worked. Every site has a story about the new 4.5W Haswell CPUs even if they are not new, not 4.5W, or anything other than a scam.

http://semiaccurate.com/2013/07/25/intel-keeps-up-the-unethical-sdp-scam-with-new-4-5w-parts/

The SDP issue at CES is reported here

http://www.theverge.com/2013/1/9/3856050/intel-candid-explains-misleading-7w-ivy-bridge-marketing

Also closely related is the true clock speed of the chips:

Intel "1.5GHz" processors run at 800MHz by default

 

[SiliconWars]

For once I agree completely with Charlie. This SDP stuff is a scam and the chips will never hit their advertised frequencies because they can't go above a set 4.5W TDP. I expect that at such low TDP they will be very mediocre, especially so in graphics. They just weren't designed for that level of power restriction.

The reason why they are doing it I can only guess at, but I assume it's because Bay Trail cannot compete near S800 on performance.

 

[galego]

I have checked what he says and he is rigth: many new sites are reporting the 4.5W in their headlines without adding the needed "SDP" to avoid confusion.

 

[Vesku]

Why else create a new label that differs by only one letter other than to invite misinformation?

I can only imagine what regular information channels are saying when a site like xbitlabs uses this title: "Intel’s New 4.5W Core “Haswell” Chips Blur the Line Between Laptops and Tablets."

 

[kimmel]

I suppose Intel should do what all the Arm guys do and stop giving any number out at all. Maybe then we can stop having these threads about boo hoo Intel gave us a different number than we used to get in desktops. I mean even people here can't even identify the fact that TDP != Power from the wall.

 

[Vesku]

The ARM analogy is off this would be similar to an ARM SoC maker creating their own benchmark suite and then touting how well their SoCs do on that benchmark compared to Intel chips. ARM SoC makers do not need to convince people that their products can handle very small form factors they have to convince people that their products aren't slow. Where as Intel does not need to convince people that their Core architecture isn't slow they have to convince people that they can handle very small form factors.

 

[galego]

The problem is not in the number.

The problem is that SDP is not mentioned in the headlines and many people will take 4.5W to be the TPD.

Another problem is on the own definition of SDP:

I was sitting in a room with most of Intel’s target audience at the time of the embargo and most didn’t know until I asked them about it. So Intel is trying surprisingly hard to keep people from asking questions about SDP because the company knows if people actually get educated about it the scam is over.

 

[frozentundra123456]

So who cares? Again the proof will be in the benchmarks. How long battery life is and what the performance is. I would hazard a guess that 99% of the people who buy a tablet dont even know or care what the TDP of the chip is (or even the model of the chip itself). If the performance and battery life is good, they will like it. If not the market will reject it.

 

[galego]

If nobody cares about numbers and if 99% of the people doesn't even know what a TDP is then why is Intel mentioning the "4.5W" in the news?

 

[jpiniero]

If nobody cares about numbers and if 99% of the people doesn't even know what a TDP is then why is Intel mentioning the "4.5W" in the news?

Wall Street

 

[Idontcare]

If nobody cares about numbers and if 99% of the people doesn't even know what a TDP is then why is Intel mentioning the "4.5W" in the news?

Same reason they heavily advertised the benefits of their "3D transistors" to a bunch of people who don't even know what a semiconductor is let alone how the outdated 2D transistors worked.

This has nothing to do with consumers, or even OEMs.

This has everything to do with increasing the perceived resale value of Intel's primary SKU of concern - INTC

And if you happen to be an individual with INTC holdings then you appreciate that Intel's business leaders are doing what they can to increase the resale value of your asset.

 

[frozentundra123456]

Same reason they heavily advertised the benefits of their "3D transistors" to a bunch of people who don't even know what a semiconductor is let alone how the outdated 2D transistors worked.

This has nothing to do with consumers, or even OEMs.

This has everything to do with increasing the perceived resale value of Intel's primary SKU of concern - INTC

And if you happen to be an individual with INTC holdings then you appreciate that Intel's business leaders are doing what they can to increase the resale value of your asset.

Or the same reason AMD advertised how great Bulldozer was for gaming by giving GPU limited benchmarks and how overpriced they claimed intel was by advertising it against the 1000.00 extreme edition cpu which nobody in their right mind would use for gaming if cost was a concern.

 

[Enigmoid]

Anyway I've seen problems in TDP marketing from both sides now. Intel on desktop is remarkably good with its TDP numbers (often prime or IBT will be under TDP). AMD on the desktop tends to sometimes use more power than TDP is rated for.

On mobile both sides are bad. Intel tends to ramp power up to levels greater than the TDP to support boost but with proper cooling is fully able to support Prime + Furmark on CPU cores and IGP (at top boost speeds). AMD tends to keep to strict TDP guidelines (sometimes lower which is nice) but rarely maintains top boost speeds and under gaming load often drops to non boost levels (intel will keep the boost but burn more power than tdp). Under Prime + Furmark trinity has the tendency to really throttle badly. Jaguar doesn't have this problem and richland is better at keeping the clocks up. (Llano almost never boosted at all).

 

[galego]

One thing is the freedom that both Intel and AMD have to rate their processors due to the lack of an unambiguous definition for TDP.

Another thing is marketing numbers without clearly stating that are not TDP numbers. There are huge differences between a marketing number such as 4.5W and a real number such as 11.5W

 

[Idontcare]

One thing is the freedom that both Intel and AMD have to rate their processors due to the lack of an unambiguous definition for TDP.

Another thing is marketing numbers without clearly stating that are not TDP numbers. There are huge differences between a marketing number such as 4.5W and a real number such as 11.5W

What makes 11.5W any more real than 4.5W?

How often do you operate your CPU at its spec'ed TDP?

What number is more relevant to you, the spec'ed TDP that is rarely visited by your processor or the number that is more reflective of the power used by your CPU in the everyday application environment that it actually experiences?

 

[BallaTheFeared]

What makes 11.5W any more real than 4.5W?

How often do you operate your CPU at its spec'ed TDP?

What number is more relevant to you, the spec'ed TDP that is rarely visited by your processor or the number that is more reflective of the power used by your CPU in the everyday application environment that it actually experiences?

According to my cpu it's drawing 86w at 4.8GHz, kill-a-watt says 165w total system, idle says 55w in C7s.

CPU is speced for 84w TDP, does this include the cpu and gpu at full load, under what load, I dare not say fore I do not know!

 

[galego]

What makes 11.5W any more real than 4.5W?

From #7:

The problem is that SDP is not mentioned in the headlines and many people will take 4.5W to be the TPD.

In the context of my posts, "real" did mean "real TDP": 11.5W is the real TPD, 4.5W is not.

 

[frozentundra123456]

From #7:



In the context of my posts, "real" did mean "real TDP": 11.5W is the real TPD, 4.5W is not.

Yes, and they didnt say 4.5 watts was the TDP, they said it was the SDP.

 

[SPBHM]

One thing is the freedom that both Intel and AMD have to rate their processors due to the lack of an unambiguous definition for TDP.

Another thing is marketing numbers without clearly stating that are not TDP numbers. There are huge differences between a marketing number such as 4.5W and a real number such as 11.5W

I'm not seeing that, anandtech.com

"Haswell Goes Fanless: 4.5W SDP Parts in Limited Volumes Later this Year"

he even goes to explain a little of what SDP means, the first thing you see is the slide provided by Intel

IF TDP and SDP were the same Intel would have no need to mention both I guess... I would prefer Intel to keep using their "TDP" numbers and that's all, but it's not Intel using "SDP" numbers as TDP.

 

[Silverforce11]

What makes 11.5W any more real than 4.5W?

How often do you operate your CPU at its spec'ed TDP?

What number is more relevant to you, the spec'ed TDP that is rarely visited by your processor or the number that is more reflective of the power used by your CPU in the everyday application environment that it actually experiences?

Its a pretty big deal when you rate max TDP versus "average" usage scenario. Because the form factor they are dealing with, thermals could kill these devices outright if its not designed to cope with "max".

More exploding smartphones/tablets is the last thing we want, when its in our pants or on our laps.

 

[monstercameron]

but how dynamic are these scenarios for intel to be providing a single number and not a range? and what is the scenario specified in the design?

 

[galego]

Some headlines omitting any mention to "SDP":


Intel teases 4.5W Haswell CPU for fanless tablets

Intel announce 4.5W Haswell chip for ultra-thin fanless devices

Intel's 4.5 Core Y Haswells Make Way for Fanless Laptops, Chip ...

Intel To Make 4.5W Chips To Enable More Fanless And Tablet ...

Intel To Launch 4.5W Haswell Processors Later This Year

Intel confirms 4.5 watt Haswell processor aimed towards fanless ...

Intel teases new Haswell chip to allow fanless tablets

Intel teases Haswell chips in fanless tablets by the end of the year

Chip Haswell 4.5W không cần quạt tản nhiệt

Intel 預覽適用於無風扇平板的4.5W Y-series Core 處理器

Intel: CPU Haswell da 4.5 Watt per tablet e PC fanless

 

[BallaTheFeared]

Its a pretty big deal when you rate max TDP versus "average" usage scenario. Because the form factor they are dealing with, thermals could kill these devices outright if its not designed to cope with "max".

More exploding smartphones/tablets is the last thing we want, when its in our pants or on our laps.

This would only make sense if OEMs weren't aware of what the chips drew and what they needed to cool them.

You'd then have to assume if that wasn't true than OEMs don't test their products but rather simply package them and send out "Sue ME" fliers with every purchase.

 

[SPBHM]

Some headlines omitting any mention to "SDP":

how much of that is Intel's fault?
we clearly have sites reporting it correctly, we clearly have Intel material showing TDP and SDP as different things...

was Intel trying to create this confusion?

sites like Anandtech clearly got it right and are not hiding any relevant info.

 

[Arkaign]

http://i44.tinypic.com/20tko03.jpg

As patently obvious, the TDP-style desktop numbers are sort of useless for application in tablets/phones. Hence why ARM doesn't really disclose that type of information. You're not going to load a tablet with something like Prime or IBT, it's just not that kind of device (perhaps in a few years that will change).

[redacted]

SDP clearly doesn't mean TDP. If anyone thinks that, they need to eat a shotgun ASAP. SDP is meaningful for the kind of devices that are functionally almost incapable of fully loading due to inherent factors of their usage/OS/software as an overall design.

[redacted]

1) We're not 4Chan
2) Thread crapping will not be tolerated. You've been here long enough to know that, Arkaign.
-ViRGE