Beta

×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Intel's Haswell Moves Voltage Regulator On-Die

Unknown Lamer posted about a year ago | from the march-of-progress dept.

Intel 237

MojoKid writes "For the past decade, AMD and Intel have been racing each other to incorporate more components into the CPU die. Memory controllers, integrated GPUs, northbridges, and southbridges have all moved closer to a single package, known as SoCs (system-on-a-chip). Now, with Haswell, Intel is set to integrate another important piece of circuitry. When it launches next month, Haswell will be the first x86 CPU to include an on-die voltage regulator module, or VRM. Haswell incorporates a refined VRM on-die that allows for multiple voltage rails and controls voltage for the CPU, on-die GPU, system I/O, integrated memory controller, as well as several other functions. Intel refers to this as a FIVR (Fully Integrated Voltage Regulator), and it apparently eliminates voltage ripple and is significantly more efficient than your traditional motherboard VRM. Added bonus? It's 1/50th the size." Update: 05/14 01:22 GMT by U L : Reader AdamHaun comments: "They already have a test chip that they used to power a ~90W Xeon E7330 for four hours while it ran Linpack. ... Voltage ripple is less than 2mV. Peak efficiency per cell looks like ~76% at 8A. They claim hitting 82% would be easy..." and links to a presentation on the integrated VRM (PDF).

cancel ×

237 comments

Sorry! There are no comments related to the filter you selected.

excited (5, Funny)

Anonymous Coward | about a year ago | (#43715815)

come guys, comment, so I know how excited I should be

Re:excited (4, Funny)

Anonymous Coward | about a year ago | (#43715993)

The 1.2V regulator will actually produce 1.199988484939848 volts !

Re:excited (0, Offtopic)

girlintraining (1395911) | about a year ago | (#43716081)

come guys, comment, so I know how excited I should be

Sorry, we were just reading about how the IT crowd will have one final episode, and really rather didn't care about Intel changing something about their product that likely nobody would have ever noticed if some marketdroid hadn't palmed the new Dice overlords a few scheckles to get it on the slashdot front page. So, umm, if you want to know how excited you should be, well... You know how happy a three year old is when they go pee-pee in the grownup toilet all on their own? Imagine you're mom, and the kid is thirty-five, retarded, and you take care of him.

You should be as excited as mom is.

sinking heat? (5, Interesting)

p51d007 (656414) | about a year ago | (#43715829)

with the on die regulator, won't that area of the chip be a tad warmer than the rest of the chip, or will the heat be a moot point?

Re:sinking heat? (0)

Anonymous Coward | about a year ago | (#43715931)

with the on die regulator, won't that area of the chip be a tad warmer than the rest of the chip, or will the heat
be a moot point?

Heat hopefully won't be an issue. Let's hope heat output scales at least somewhat with component size that is 98% smaller.

It'd be nice to see the VRM using the 22nm lithography process but when a processor is quoted as Xnm, this is the smallest feature size on the processor, not all features on the processor - unfortunately.

Re:sinking heat? (2)

dotgain (630123) | about a year ago | (#43716031)

Heat hopefully won't be an issue. Let's hope heat output scales at least somewhat with component size that is 98% smaller.

And why would it do that? A given voltage drop multiplied by the current through it equates to a certain wattage of heat dissipation, regardless of the size of the package.

Re:sinking heat? (2)

adolf (21054) | about a year ago | (#43716099)

Your broad generalization is only if it is a linear regulator. Switch-mode regulators change the game. TFA doesn't seem to indicate which it is.

Re:sinking heat? (5, Informative)

Shavano (2541114) | about a year ago | (#43716167)

It's a switch-mode (Buck) regulator. You can tell from the efficiency curve and the fact that it requires an inductor. It is more efficient than a linear regulator and less efficient than a good external Buck regulator. However, being on-chip it will regulate the voltage better because there won't be significant I*R drop between the regulator output and the load. And as they mention, the cooling fan will be right on top of it, so it is more effectively cooled than an external regulator typically is.

Re:sinking heat? (3, Insightful)

viperidaenz (2515578) | about a year ago | (#43716679)

The voltage regulation issue can easily be solved by having a feedback connection from the die to the external VRM.
There are only two benefits I can see:
1) Higher voltage in to the chip means lower current, which saves power. You I*R formula is slightly wrong, its actually I^2 * R, double the current means 4x the power loss.
2) Lower system cost. the more crap that gets stuffed on the die/in the chip, the less is required on the board. That means fewer components, smaller board area and quicker assembly.
There are of course other benefits that only benefit Intel
a) Fewer external components means they can charge more for their chip without effecting system cost.
b) smaller system = happier customer = will pay more
c) If it does actually result in lower power, then you get more performance or more battery life = customer will pay more

Re:sinking heat? (1)

adolf (21054) | about a year ago | (#43716695)

I figured that it must be that way, but with the power required by a CPU such a regulator must either very noisy and/or require substantial capacitance and/or use ridiculously high frequencies.

So. Filtering? These might be the most expensive mass-produced caps in the world if they're also on-die.

Re:sinking heat? (1)

cats-paw (34890) | about a year ago | (#43716775)

seems like they could have beat the IR drop by simply bringing out sense lines.

it's not obvious too me how this helps intel.

unless of course the motherboard makers aren't doing a very good job with those external regulators because the intel chips now require such high performance regulation.

that's a distinct possibility.

Re:sinking heat? (0)

Anonymous Coward | about a year ago | (#43716129)

certain wattage of heat dissipation

Watts is used to measure power, not heat.

Re:sinking heat? (1)

AvitarX (172628) | about a year ago | (#43716247)

Heat is energy, and it dissipates over time, energy / time = power.

Re:sinking heat? (2)

viperidaenz (2515578) | about a year ago | (#43716719)

Watts is used to measure heat.

http://en.wikipedia.org/wiki/Thermal_resistance [wikipedia.org]

The unit is "degrees per watt". As in "5C/W" = "this heat sink will rise by 5 degrees dissipating 1 watt"

Re: sinking heat? (0)

Anonymous Coward | about a year ago | (#43716155)

that scaling is only true for linear regulators. if they managed to put a switching regulator on the die then the heat dissipation is based on the efficiency and the output power. that efficiency will scale with input voltage but definitely not by as much as a linear regulator. the efficiency is going to scale more with the topology of switching regulator they choose and their ability to limit the resistance of the regulator components and their leakage.

Heat (4, Interesting)

girlintraining (1395911) | about a year ago | (#43715841)

Intel refers to this as a FIVR (Fully Integrated Voltage Regulator), and it apparently eliminates voltage ripple and is significantly more efficient than your traditional motherboard VRM. Added bonus? It's 1/50th the size."

I have yet to come across a voltage regulator that doesn't run hot. Typically, it's one of the hottest components in an electrical circuit. And we're integrated this into a slab of silicon already well-known for getting so hot it can catch fire?

Can someone please tell me why this is a good idea, because all of my experience in electrical engineering says that when things heat up, they become more unstable and prone to failure, and the one thing you do not want going critical is your voltage regulator. If that goes, the whole computer catches fire.

Re:Heat (3, Interesting)

Anonymous Coward | about a year ago | (#43715879)

Most likely, because it's integrated into the CPU itself, the voltage regulator can be made more efficiently and thus save power and heat etc. Discrete parts have their limitations, and doing it on-die might just mitigate that.

Re:Heat (4, Informative)

Anonymous Coward | about a year ago | (#43715881)

You're going to run into that heat anyway, whether it's on the motherboard in general or on the CPU. You can't win. But at least it's better to have heat build-up near a heat-sink, so for high-power conditions it might actually be better to put it on the CPU. I'm also an electrical engineer, but thermals are really a mechanical engineer's realm, so I can't run numbers for you.

Re:Heat (1)

l3iggs (1108141) | about a year ago | (#43716713)

But at least it's better to have heat build-up near a heat-sink

True, that it's a good thing that the CPU heatsink is nearby to cool the regulator, but having it on dye also means that the CPU is nearby which seems to be a bad thing from a system stability standpoint.

Something tells me the folks at Intel have carefully considered all this and that the extra capacity the CPU heatsink now needs to keep the CPU stable is preferable to having an on motherboard Vreg, which is relatively far away from the CPU and that they can't trust to be cooled properly.

Re:Heat (2)

rrhal (88665) | about a year ago | (#43715885)

Being 1/50th the size it will be welcome on mobile devices. Not sure that its a good thing for your gaming desktop.

Re:Heat (4, Informative)

Shavano (2541114) | about a year ago | (#43716183)

Being 1/50th the size it will be welcome on mobile devices. Not sure that its a good thing for your gaming desktop.

That 84 watts is going to rip through your mobile device's battery pretty damn fast.

Compared to a traditional regulator? (1)

perpenso (1613749) | about a year ago | (#43716549)

Being 1/50th the size it will be welcome on mobile devices. Not sure that its a good thing for your gaming desktop.

That 84 watts is going to rip through your mobile device's battery pretty damn fast.

Don't we need to compare it to a traditional regulator implementation before we come to that conclusion? Assuming pretty damn fast means faster than current Atom based devices.

Re:Heat (2)

ebno-10db (1459097) | about a year ago | (#43716267)

Being 1/50th the size it will be welcome on mobile devices.

It's not clear how they measure "1/50th the size". I could be wrong but it sounds like marketing hype. With a switching regulator the inductors and capacitors generally take up much more real estate than the chip. If they have some magic way to reduce the inductor and capacitor sizes it isn't mentioned in the article (and that would be a much bigger deal than just putting the regulator on the die).

Re:Heat (4, Interesting)

petermgreen (876956) | about a year ago | (#43716331)

You can reduce the inductor and capacitor sizes a lot by increasing the switching frequency. Of course doing so will likely increase your switching losses but it may still be worth it if it lets you put the regulator closer to the load. Especially given the ever lower voltages that modern chips are running at.

Re:Heat (4, Interesting)

ebno-10db (1459097) | about a year ago | (#43716373)

You can reduce the inductor and capacitor sizes a lot by increasing the switching frequency.

But you can do that w/ an external regulator too. Apparently the secret is on-chip inductors. Now that's impressive. I'm surprised that some of the "analog" companies making switchers didn't come up with that first. I know Intel has good fab tech, but this seems more like the sort of funky thing analog guys would come up with first.

http://www.psma.com/sites/default/files/uploads/tech-forums-nanotechnology/resources/400a-fully-integrated-silicon-voltage-regulator.pdf

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716415)

The size of external components on any switcher is made by the frequency of the switcher! Multi-megahertz switchers have tiny discreets.
If I were to guess, I would say that they are incorporating a very high frequency switching core that would require small external discreets. Really, there is no other way that I know of to do this. IOO, there will be a switching controller and a big ole FET or two. The discreets (cap and inductor) will be external.

Re:Heat (5, Interesting)

viperidaenz (2515578) | about a year ago | (#43716795)

If you core requires 1V and 90 watts you need to transfer 90A through your PCB traces, up in to the chip, across the bond wires (if there are any) and on to the die.
If your die has a regulator on board and accepts 12V instead, and is 80% efficient you only need to transfer 9.4A. You've just lowered your resistive losses by about 100x. If the connection between the external VRM and die is 0.001ohms, at 90A you waste 8.1W. at 9.4A you waste 0.088W.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716439)

I'm fairly certain that they were refering to the space occupied by the regulator... stripping all of the casings, leads, and the length of interconnects would shrink the circuit considerably.

Re:Heat (5, Informative)

fuzzyfuzzyfungus (1223518) | about a year ago | (#43715891)

My suspicion(if only for die-space reasons, it isn't purely cosmetic that contemporary VRMs occupy a substantial amount of board space) is that is this a 'marketecture' summary of Intel moving some additional voltage adjustment and power gating functions on die, to support dynamic adjustment of power to the greater number of components(multiple CPU cores, possibly independently clocked, GPU, RAM controller, PCIe root, etc.); but we'll still see a bunch of chunky power silicon under serious heatsinks clustered around the CPU socket.

Given that much of the contemporary power savings are achieved by superior idling, rather than absolute gains in maximum power draw, Intel is either going to have to keep moving power regulation on die, or start dedicating even more pins to tiny voltages at nontrivial currents, with the associated resistive losses; but that won't necessarily change the fact that the circuitry that brings the 12v rail down to what the CPU wants is a pretty big chunk of board.

Re:Heat (2)

girlintraining (1395911) | about a year ago | (#43715963)

My suspicion(if only for die-space reasons, it isn't purely cosmetic that contemporary VRMs occupy a substantial amount of board space) is that is this a 'marketecture' summary of Intel moving some additional voltage adjustment and power gating functions on die, to support dynamic adjustment of power to the greater number of components(multiple CPU cores, possibly independently clocked, GPU, RAM controller, PCIe root, etc.); but we'll still see a bunch of chunky power silicon under serious heatsinks clustered around the CPU socket.

That's the only plausible thing I could come up with as well. The control logic could go into the CPU, but I don't see how pulling 12V down to fractions of a volt is going to happen on the die itself without it burning a hole through the board; heatsink or not, you can't escape Ohm's Law.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716057)

you can't escape Ohm's Law.

Actually you can. It's called a switching power supply.

Re:Heat (5, Funny)

girlintraining (1395911) | about a year ago | (#43716127)

you can't escape Ohm's Law.

Actually you can. It's called a switching power supply.

In other news, a Nobel Prize in Physics was awarded to Anonymous Coward of Slashdot today, after discovering that the laws of physics do not apply to switching power supplies... His next research proposal is on solving the energy crisis by designing keyboards to detect when someone is angry and then increasing the key resistance by piezoelectric effect to generate energy. While it would generate only marginal amounts of power when used by 99.975% of the population, it was recently discovered that the remainder are actually Linux and Apple fanboys who, if fed a regular diet of dismissives via their computer screen, will so furiously hit the keyboards that power for entire cities is easily achievable.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716255)

Actually, not all devices obey Ohm's Law, so I wouldn't be surprised if switching power supplies don't. Inductors and capacitors, for example, are not ohmic.

Re:Heat (2, Informative)

Anonymous Coward | about a year ago | (#43716451)

Actually, no real device is ohmic at all. Even a resistor will heat up with increasing current causing an increased resistance that is non-linear. For us EEs, ohmic devices are our massless pullies and frictionless inclines.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716539)

Well yeah, but most EE's don't deal with circuits at that level unless they're doing special stuff. For most non-sensitive circuits you don't have to worry about the frequency response of a resistor.

Re:Heat (4, Insightful)

girlintraining (1395911) | about a year ago | (#43716595)

Actually, no real device is ohmic at all. Even a resistor will heat up with increasing current causing an increased resistance that is non-linear. For us EEs, ohmic devices are our massless pullies and frictionless inclines.

I'm not a certified EE, but I have built electronic circuits. I know there's a lot of ways to 'cheat' on paper; switching power supplies don't get rid of ohm's law though, they're simply more efficient. Ohm's law is about the relationship between resistance, voltage, and current. Those relationships are derived from the physics about electron exchange between different materials. Now yes, capacitors and inductors both run 90 degrees out of phase between voltage and current so it can appear to be violating ohm's law, but if you apply a correction factor you'll see it's pretty close to parity. When you get down to really small discrete components, like a transitor for example, measurement inaccuracy and time domains will really start to screw with you, but ohm's law still holds even down to that scale.

Ohm's law is the reason for these changes Intel is making: An attempt to remove parasitics from the circuits, which all boil down to resistance; Whether it's phase-shifted forward because of capacitors, or backwards because of inductors, or because of components that create those effects, doesn't really matter.

Now you're right, a purely ohmic device doesn't exist. Even resistors can generate small amounts of phase shift. But that doesn't make them "massless pullies" or "frictionless inclines". Ohm's law is still useful for the same reason the OSI 7 layer model is still useful, despite no network yet having been designed that perfectly adheres to it...

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716747)

Ohm's law is the reason for these changes Intel is making:

Ohm's law as applied to the connection between the regulator and the load, not because of how Ohm's law applies (or doesn't apply...) to the regulator.

I know there's a lot of ways to 'cheat' on paper; switching power supplies don't get rid of ohm's law though, they're simply more efficient.Ohm's law is about the relationship between resistance, voltage, and current.

And there are vast categories of devices that are not ohmic, not even in the more generalized AC circuit sense using complex impedance. Basic semiconductor devices can be very non-linear relation between current and voltage, or not even have a pure functional relationship between the two due to hysteresis or other dependences. You have the first sentence backwards, as the on paper cheat is the idea that everything is broken up into simple, linear ohmic (complex or not) components, while the real world has a lot of disregard for that.

Re:Heat (1)

Charliemopps (1157495) | about a year ago | (#43716285)

Who said they're going to keep it at 12 volts? A VRM can also be the transformer, but it doesn't have to be. They could ramp down the voltage on the board just like they always did and just have a VRM on the chip that is maintaining a steady voltage. If you read the article they are barging about how little fluctuation they are getting. So it seems like what they are doing here is adding basically an extra regulator on chip so they can have extremely stable voltage. I'm guessing as small as things are getting now, just the trip across the motherboard can have noticeable fluctuations on supply voltages due to EM interference and temp fluctuations, so having a regulator on the chip lets them get more precise. Maybe that precision will let them do some magic on the chip to increase performance or something?

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716363)

The control logic could go into the CPU, but I don't see how pulling 12V down to fractions of a volt is going to happen on the die itself without it burning a hole through the board; heatsink or not, you can't escape Ohm's Law.

You should look up how switching power supplies work, which gets around this problem quite well. Hobbyists may still use their 7805 and 7812 because of how simple they are, but even for years now a diy electronics person can make an equivalent switching circuit for fractions of a dollar. Efficiencies are typically 70+%, and can go upwards of 95% And like many semiconductor devices, it is nonlinear and doesn't follow Ohm's law (although you can sometimes still define a local impedance for the voltage to current ratio...). The efficiency in many designs gets high with larger voltage drop. Where a linear 12 V to 1.2 V regulator would only be 10% efficient and burning up say 108 W if delivering 10 A on the output, a well built switching supply delivering the same current and same voltage drop can dissipate only 1 W or less in the regulator itself.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716425)

The CPU is already dissipating tens of watts. A switching mode power supply can easily achieve their claimed ~80% efficiency. This means overall there is only a roughly 20% increase in total heat dissipation and most of this is likely offset by improvements that can be made to the chip because of the integrated Vreg.

Re:Heat (5, Informative)

Omega Hacker (6676) | about a year ago | (#43716509)

Ohm's law is completely irrelevant to this situation *in the form you describe*. "Burning a hole through the board" would be possible and a simple function of Ohm's law only if they were using a linear regulator to generate the Vcore. But VRM's have been switching DC/DC converters since the 486 days. They achieve a voltage conversion by switching the incoming voltage on and off *very fast*, which results in an output voltage that's a function of the input voltage and the duty cycle of the on/off switching. An inductor (current-smoothing) and capacitor (voltage smoothing) give a nice clean DC voltage.

The differences between on-motherboard VRMs and this new in-package (it's technically a separate chip...) are significant. First off, physically moving it closer means that you're not sending 100+ Amps of current over the 3-4 centimeters of generally very thin copper traces on the PCB, they're sent millimeters through die-bond wires, or even through a solid substrate (no idea what Intel does at that level). There's your Ohm's law coming into play at that level, but the power losses there are relatively minimal since you're talking maybe a few tenths of an ohm. Die-bond wires are going to drop that to 10's of milli-ohms probably, so nothing major but still a positive effect.

The main reason this will generate a lot less heat is because of the *frequency* of the switching. Because this on-board VRM is so much smaller, it can switch the input faster (shorter wires, less parasitic capacitance, less ringing, etc.). This in turn means smaller value components required, e.g. the switch from the monster inductors seen on the motherboard (at maybe 1-2MHz switching) in the slide to the tiny chip-scale inductors on the FIVR (at 10's or 100's of MHz). The end result of all of this is that switching losses get significantly smaller. It's those losses that create heat local to the regulator. If they can for example go from an 80% efficient VRM to an 90% efficient FIVR for a 100W CPU load, they reduce the switching losses from 25W to 11.1W.

Re:Heat (1)

swalve (1980968) | about a year ago | (#43716731)

Sure you can. Voltage regulators aren't just resistors any more. You divide the 12v @ 1 amp into 1v @ 1amp at 12 different spots. More or less. Think of it like TDMA, if that helps. Switching voltage regulators are super efficient. And even if there isn't an efficiency gain in the VRM, they will likely be one since the processor will be operating at tighter voltage tolerances. The VRM will be closer to the load and be able to react to load shifts quicker, meaning the processor spends less time slightly over-volted. (If the processor needs 1.2 volts or it crashes, an external regulator might have to have a setpoint of 1.3 volts so it never goes under. But that means it wastes a lot of wattage on heat. If it can instead have a 1.225 setpoint, you are saving power that would normally be shed as heat.

Re:Heat (4, Interesting)

Virtucon (127420) | about a year ago | (#43715905)

Well even at 10W I'm wondering how they'll address the heat.
With the density of circuits in the adjacent silicon I would wonder how they're providing enough isolation to prevent it from becoming a very small brick.

Re:Heat (3, Insightful)

Anonymous Coward | about a year ago | (#43715961)

Considering that they've already started shipping an actual product, perhaps you should switch modes--from skeptic to sleuth. Start from the proposition that (a) it's possible or (b) they're leaving something out of the marketing jargon. There are a million ways they could do it wrong, and likely only a few ways to do it right. If you start from the proposition that Intel is shipping a working product, then it should be much easier to figure out.

Re:Heat (1)

swalve (1980968) | about a year ago | (#43716761)

Some of the pentium 4 chips dissipated over 100 watts. I think they know how to move heat off of silicon.

Re:Heat (1)

RightwingNutjob (1302813) | about a year ago | (#43715907)

I can see the logic behind shortening the length of the wire carrying 'clean' power and getting it away from all the other components (read: noise sources) on the motherboard. It also takes the thinking burden away from the chip integrator and motherboard designer (which is a non-negligible bonus for both marketing and engineering).

Re:Heat (1)

Barlo_Mung_42 (411228) | about a year ago | (#43715919)

I'm not an EE so I won't pretend to fully understand this particular case but I like it when tech companies reach a bit and try something hard. This may or may not be a good idea but I'm still excited about it.

Re:Heat (2)

citizenr (871508) | about a year ago | (#43715943)

I'm not an EE so I won't pretend to fully understand this particular case but I like it when tech companies reach a bit and try something hard. This may or may not be a good idea but I'm still excited about it.

Thats what she said.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43715921)

TFA says it has better efficiency, but it's not clear how much better.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43715945)

Perhaps it's vastly more efficient than a traditional voltage regulator? It is, you know, silicon fabbed by Intel. The undisputed leader in chip fabrication technology. (Seriously. They're conservatively at least 2 generations ahead of everyone else)

Really, this is a pretty good thing. One of the weak points of many motherboards is the shit voltage regulators they come with. (And the biggest real selling point of premium motherboards) Pulling this in to the package will eliminate one more unknown and should allow Intel to better optimize their chips, not having to deal with whatever crappy noisy power being sent by the lowest-bid mobo the PC maker slapped in in to the case.

Expect more things like this in the future.

And less things like it in the further future. Systems integrate and de-intgergrate on a cycle as new technologies become prominent. We're currently in an integration cycle.

Re:Heat (1)

ebno-10db (1459097) | about a year ago | (#43716295)

Perhaps it's vastly more efficient than a traditional voltage regulator? It is, you know, silicon fabbed by Intel. The undisputed leader in chip fabrication technology. (Seriously. They're conservatively at least 2 generations ahead of everyone else)

Maybe one generation ahead, but that's in digital chips. While things like switching regulators can be built on digital processes (it's been done before) it's generally not the optimal process. Maybe they've come up with some clever ways to build better switchers on a digital process than previously, but it's hard to believe it's better than a process designed for stuff like this.

Re:Heat (1)

Anonymous Coward | about a year ago | (#43715947)

Say you have the choice between using a different SoC (maybe an ARM based system) that needs only one power rail, and an Intel SoC that did not include these regulators. Which will cost less to engineer, and be a smaller, simpler solution?

Remember the one of the many forms of Halswell is a low power solution, with TDP of 10 watts...

Re:Heat (1)

sshir (623215) | about a year ago | (#43716025)

May be that they still use discrete FETs, it's just control circuitry is on die now. (I'm speculating)

Re:Heat (1)

Shavano (2541114) | about a year ago | (#43716233)

May be that they still use discrete FETs, it's just control circuitry is on die now. (I'm speculating)

That's an interesting question. They can make a FET with as big a voltage and current rating as they want by just making it many times the size of a logic FET. But they also need thicker gate oxide to prevent Vdg breakdown at multiples of the normal logic operating voltage. Not sure how they would do that. It could add a process step, which would make it expensive.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716075)

Because CPU nowadays runs much cooler then they use to. The regulator is more efficient (less heat) in the cpu and any extra heat should be easily managed. Most voltage regulators run hot because they often rely on limited cooling unlike CPU which contains quite a bit of cooling (thermal pad/paste, sizable heatsink, constantly running fan).

There also this thing call temperature sensors which CPU already have which can basically dethrottled the cpu (less heat) or shut down in case of emergencies. Heat has always been an issue with CPU, one well researched and dealt extremely well with nowadays.

Basically you have many pros for very little cons

Pro:
-More efficient (less heat and electricity)
-Save space (no dedicated chip which takes up more room)
-Save costs (no dedicated chip which would cost more)

Con:
-Less silicon space for CPU which can be used for like cache
-More heat? (debatable since you could easily use that space for something else that generates as much heat)

Re:Heat (5, Informative)

Kjella (173770) | about a year ago | (#43716091)

Can someone please tell me why this is a good idea

The long story is here (PDF) [psma.com] . Motherboard will still do the heavy lifting from 12V to 2.4V, but the integrated VRM will distribute it. Advantage is extremely clean, fine-grained, low-latency and flexible power supply to deliver exactly as much power to where it's needed and probably - this is just speculation on my part - allowing the CPU to work on a wider range of voltages since there's less noise and ripple so you don't need the same tolerance limits. It sounds perfect for smart phones, tablets and laptops that are primarily battery-limited, nice to have for average machines but potentially an issue for overclockers. All you need is cooling though, it shouldn't limit overclocking if you can keep the temp down.

Re:Heat (1)

Shavano (2541114) | about a year ago | (#43716175)

Intel refers to this as a FIVR (Fully Integrated Voltage Regulator), and it apparently eliminates voltage ripple and is significantly more efficient than your traditional motherboard VRM. Added bonus? It's 1/50th the size."

I have yet to come across a voltage regulator that doesn't run hot. Typically, it's one of the hottest components in an electrical circuit. And we're integrated this into a slab of silicon already well-known for getting so hot it can catch fire?

Can someone please tell me why this is a good idea, because all of my experience in electrical engineering says that when things heat up, they become more unstable and prone to failure, and the one thing you do not want going critical is your voltage regulator. If that goes, the whole computer catches fire.

It's cooled by your CPU fan.

Re:Heat (1)

marcosdumay (620877) | about a year ago | (#43716477)

Thanks, that gets the overall picture.

So, the idea is that they'll get some very nice inductors on die, capable of replacing some much more expensive external ones. Also, they can distribute the load to a lot of paralel circuits, creating the right tension for each part of the chip, and reducing the loss of each circuit.

But really, at 90W, a embebing a 76% efficient (not really an exceptional result) conversor means that you'll need to dissipabe other 28W at peak power. Well, I can't say if this is worth it. Certainly, Intel engineers can say, but won't, and the marketeers will lie anyway.

Re:Heat (0)

Anonymous Coward | about a year ago | (#43716483)

Can someone please tell me why this is a good idea

1. Precise voltage control as implemented by CPU manufacturer.
2. Less net system heat.
5. Less total power consumption.
3. Fewer discrete board components.
4. Smaller board size.
6. Lower total cost.

Not necessarily in that order. Don't be thick headed. FPUs were once discrete components. Are you still butt-hurt about FPUs being integrated?

Never bet against integration because eventually you're wrong. Integration is why computers succeed.

But did they fix the TIM? (0)

Anonymous Coward | about a year ago | (#43715863)

I don't want to delid my Haswell CPU.

On-Die die risk? (0)

Anonymous Coward | about a year ago | (#43715871)

As long as incorporating logical features to a chip is nice, is it prudent to incorporate such risky function to it? Wouldn't it become more vulnerable to highly out of control voltage variations? Will it become more easy to burn?

Re:On-Die die risk? (1)

eclectro (227083) | about a year ago | (#43716905)

I suggest the tag for this story be 'whatcouldpossiblygowrong"

Interesting (0)

Anonymous Coward | about a year ago | (#43715915)

Interesting possibilities but as TFA says, it could cause a lot of pain for overclockers.

Full presentation (5, Informative)

AdamHaun (43173) | about a year ago | (#43716005)

You can find the full slide set in PDF format here [psma.com] .

If I read this right, it really is a fully on-chip switching regulator, inductors and all. They already have a test chip that they used to power a ~90W Xeon E7330 for four hours while it ran Linpack. (Or a virus -- it says Linpack in the summary page.) Voltage ripple is less than 2mV. Peak efficiency per cell looks like ~76% at 8A. They claim hitting 82% would be easy, and there are "additional advancements that cannot be reported at this time" planned for the future.

The slides have bunch of other technical details about testability features, too, which is always neat to see.

Re:Full presentation (2)

girlintraining (1395911) | about a year ago | (#43716143)

hey already have a test chip that they used to power a ~90W Xeon E7330 for four hours while it ran Linpack. (Or a virus -- it says Linpack in the summary page.)

Respectable viruses take issue with your comment that Linpack is anything like them. Viruses do useful work.

Re:Full presentation (1)

ebno-10db (1459097) | about a year ago | (#43716341)

inductors and all

Now that's impressive, and I suspect the real secret to this. Not to say the semi design is trivial, but without the on-chip inductors you wouldn't have much. They weren't clear about it, but perhaps it means getting rid of the big filter caps, and relying on smaller caps for each regulator. It also explains the fast response time with a bunch of smaller regulators.

Re:Full presentation (0)

Anonymous Coward | about a year ago | (#43716721)

From a switchmode supply point of view, they are shitty. 75% at 10A Yes, they can do a much bigger version, but it would eat up lots of die real estate.
So from a cost point of view, Intel is better off using their silicon area for logic and hang off POL around the board.

What is interesting is that, they can make them on die, low ripple etc and it uses a lot of digital technology.

See here for _MUCH_ better parts from their analog competitors http://www.linear.com/product/LTM4620
LTM4620 - Dual 13A or Single 26A DC/DC Module Regulator with integrated magnetics
At 10A output, you get close to 90% efficiency and they take 5V or 12V input directly. The two outputs can be wired in parallel to get the 26A output.

P.S. I use a lot of switch mode power supplies in my designs in module and discrete forms..

THE GNAA (-1)

Anonymous Coward | about a year ago | (#43716023)

shall return...

Re:THE GNAA (-1)

Anonymous Coward | about a year ago | (#43716289)

One can only hope.

Trollaxor would be even better, but he _still_ hasn't released the philes on the real reason why Taco left Slashdot.

Not a bad idea (1)

EmagGeek (574360) | about a year ago | (#43716033)

An on-die power controller still needs to have external capacitors, especially at the power levels we're talking about.

But, the problem areas in a switching supply are EMI and stray inductances on the board slowing down the turnon of the mosfet. Because the turnon is slowed down, the mosfet spends time in the ohmic region, which creates excess heat.

An on-die gate driver routed directly to the gate with no trace length has no stray inductance, and an on-die gate probably also has less capacitance than a discrete component. So, switching times are much much faster, and there is far less loss per cycle in the ohmic region. That's a Good Thing(TM).

EMI will also be reduced just because there aren't a bunch of high speed traces running around, and the thing can run in the multi-MHz range, so less energy is switched per cycle, and maybe even the switch inductor can even be small enough to simply be drawn in silicon.

I would not be surprised to see efficiency in the 95+% range even coming from 12V down to 0.9V or whatever voltage this thing runs at, so you'll be throwing an extra 6W or so into a 120W package. Not bad.

Re:Not a bad idea (1)

ebno-10db (1459097) | about a year ago | (#43716309)

even coming from 12V

Way too high of an voltage for these sorts of semi processes. I think it starts at 2.4V.

Re:Not a bad idea (1)

serbanp (139486) | about a year ago | (#43716475)

I would not be surprised to see efficiency in the 95+% range even coming from 12V down to 0.9V or whatever voltage this thing runs at, so you'll be throwing an extra 6W or so into a 120W package. Not bad.

Bollocks! Since the internal VR uses the same process as the CPU itself, it can't sustain high input voltages, therefore a one-stage 12V to 0.9V conversion is just a pipe dream.

The longer pdf presentation actually shows the motherboard-level 12V to 2.2V VR, which would be still rated for the full power (85W plus margin). OTOH, it's quite impressive that the 22nm process has support for 2.2V CMOS.

As others mentioned already, Intel is just trying to solve the power distribution issue, not eliminating the main down-conversion stage, which will *always* be external.

As a CPU VR designer myself, I'm very interested in seeing how this concept will play out. Beside the issues of more peak power dissipation on the CPU die and increased EMI, there is a long-term reliability issue involved, especially on the power stage; switching inductive loads creates ringing, which will degrade in time (through HCI) the switching transistors.

Re:Not a bad idea (1)

ebno-10db (1459097) | about a year ago | (#43716597)

switching inductive loads creates ringing, which will degrade in time (through HCI) the switching transistors

What's "HCI"?

Re: Not a bad idea (0)

Anonymous Coward | about a year ago | (#43716947)

Hot carrier injection

Re:Not a bad idea (1)

ebno-10db (1459097) | about a year ago | (#43716629)

a one-stage 12V to 0.9V conversion is just a pipe dream. The longer pdf presentation actually shows the motherboard-level 12V to 2.2V VR

Do current CPU VR's do a one-stage jump from 12V to 0.9V? If so it seems they'd have an efficiency advantage by avoiding a double conversion. 12V to 0.9V seems like a big jump for a buck converter, but perhaps there's another way.

Yawn (0)

Dachannien (617929) | about a year ago | (#43716073)

Wake me up when they move the keyboard on-die.

Re:Yawn (2, Funny)

wbr1 (2538558) | about a year ago | (#43716105)

Apparently your e-peen is already small enough to fit on die.

Re:Yawn (2)

Dachannien (617929) | about a year ago | (#43716697)

Seriously? And here I thought I was being clever. I bow to the master.

cheap oakleys (-1, Offtopic)

janet668 (2922459) | about a year ago | (#43716111)

The birth of the first fake ray bans [cheaprboutlet.com] aviator sunglasses in 1937, glasses style, as always, retains its classic teardrop-shaped appearance, design technique is relatively moderate, but the innovative style more lasting appeal. 75 years, Ray ban outlet [cheaprboutlet.com] aviator generation after generation eager to re-interpretation of the dream of the self-style choice, they continue to draw inspiration from walking in the forefront of fashion tide people. Cheap ray bans [cheaprboutlet.com] ambermatic limited edition lenses like sunglasses engine. Recalling the history of the development of the Ray-Ban lens and its quality has always been the key to success. In 1978, with the birth of the Dreaming lens, lens products range greater extension.

From a former power supply designer - Neat! (5, Interesting)

jimmyswimmy (749153) | about a year ago | (#43716161)

That's some amazing work. The current state of the art in CPU power supply designs hasn't changed in 15 years. 12V in, low voltage out, and the output voltage has been moving lower and lower for years, with designs below 1 V. If you figure you had a few percent of tolerance in the early years when everything ran off 2.5V and that few percent remains constant, then at 1 V you have almost no room for slop. So there are a lot of output capacitors there, both those electrolytics (you always hear people complaining about them but they're CHEAP) and ceramics. The ceramics cost a fortune and you need a lot of them to get your tolerance down - the first half microsecond of a load step is entirely the ceramic capacitor's response, not the controller or anything else. Moving part of the VR onboard allows them to reduce the parasitics significantly and they can probably tolerate a little higher tolerance as a result, but moreover they can get rid of some of those ceramics in the whole system - ultimately many of those on the motherboard.

So this is taking a lot of cake out of company mouths. Analog, Intersil, IRF, ON, who else - manufacturers of controllers, MOSFETs. Inductors, ceramic and 'lytic vendors are all going to lose out a bit here. Potentially Intel can reduce the platform cost vs. AMD as well, which is interesting. There is still an onboard VR but it will be 12 - 2.4 V, wherever they think the sweet spot is for efficiency and size. And the first real change in this industry for a long time. Cool work.

Re:From a former power supply designer - Neat! (1)

gr8_phk (621180) | about a year ago | (#43716209)

Where did you see the input voltage for this thing being 2.4V? I was going to ask what it is.

Re:From a former power supply designer - Neat! (1)

petermgreen (876956) | about a year ago | (#43716435)

Afaict it's not explicitly stated but it's implied by a caption on page 24 of the slides pdf that someone linked above.

Re:From a former power supply designer - Neat! (0)

Anonymous Coward | about a year ago | (#43716417)

The current state of the art in CPU power supply designs hasn't changed in 15 years

This is only true if you ignore the entire ARM ecosystem.

This switcher responds in nanosecond to load step (0)

Anonymous Coward | about a year ago | (#43716733)

With the 100 MHz cycle and sixteen phases, the output capacitors have to hold up the output for less than a nanosecond, as there
is a new pulse every 10 nS / 16 = 0.625 nS.

Very impressive indeed.

Many thanks for the link to the PDF of the presentation.

This technology could have application in AM broadcast band radio transmitters, and in audio power amplifiers.

Peter Traneus Anderson

yay house fires (0)

Anonymous Coward | about a year ago | (#43716879)

nothing says progress like a good ol house fire

Peanuts! (-1)

Anonymous Coward | about a year ago | (#43716215)

Try this sometime when you are bored!

            1) Take one pound of raw peanuts (not roasted!)

            2) Shell them, saving the skins and discarding the shells.

            3) Eat the nuts.

            4) Grind up the skins and roll them into a cigarette, and smoke!

            You'll have fun, believe me!

Re:Peanuts! (-1)

Anonymous Coward | about a year ago | (#43716275)

Try this sometime when you are bored!

Jack Off.

I am totally impressed (1)

overshoot (39700) | about a year ago | (#43716301)

... that they can incorporate the inductor and capacitors for a 90 W switchmode regulator onto silicon. This is a breakthrough in physics, not just in semiconductor processing.

Re:I am totally impressed (0)

Anonymous Coward | about a year ago | (#43716431)

This! I'm excited too. Great for the industry.

Re:I am totally impressed (1)

ebno-10db (1459097) | about a year ago | (#43716605)

This is a breakthrough in physics, not just in semiconductor processing.

What sort of a breakthrough in physics? Have they found a way around Maxwell's equations or something?

Re:I am totally impressed (0)

Anonymous Coward | about a year ago | (#43716701)

Realize that the necessary inductor for this kind of supply will be much smaller than you would need from normal switched mode power supply (smps) chips. The reason for this is that a fairly normal/average smps will run at a frequency between.. 100kHz on the low end to may a couple of MHz on the high end. This VR runs anywhere from 30MHz to 140MHz (programmable range). As frequency increases the needed inductance drops. So you can see, even compared to the high end chips at a slow speed the needed inductance will be 15 times smaller. Comparing high end and high speed it is 70-100 times less inductance. Higher switching frequencies do come with the drawback of harder to control EMI and more significantly lower efficiency.

Also, you need less capacitance since the regulator will be so close to the transistors on the chip due to having less parasitic inductance due to the closeness.

I'm not surprised they went this way. They are called /integrated/ circuits after all.

Re:I am totally impressed (2)

jimmyswimmy (749153) | about a year ago | (#43716765)

Not such a big breakthrough as you'd think. As you increase the switching frequency you can decrease the value of inductor and capacitors required. Last CPU supply I built - 10 years ago! - used 100 nH inductors at 300 kHz per phase. I skimmed the PSMA article but there was mention of MHz operating speeds, not at all unheard of these days, so the components ought to be much smaller. A 10 nH inductor and some hundreds of pF of capacitance seems very feasible without stretching the bounds of silicon technology at all.

Time to call my broker, (1)

Fengpost (907072) | about a year ago | (#43716357)

and dump my Maxim stock!

Re:Time to call my broker, (0)

Anonymous Coward | about a year ago | (#43716545)

I dump a stock of manjuice while reading Maxim!

Re:Time to call my broker, (0)

Anonymous Coward | about a year ago | (#43716607)

Good god why? Have you never heard of Internet Porn?

2010. Not news, just unattributed slides. (0)

Anonymous Coward | about a year ago | (#43716501)

So we have some slides for a February 2010 presentation, about a widely known technology thats going to be included in some upcoming parts.

Now here is something more news like: Today http://hothardware.com/ posted an article calming to be news, but was actually just a portion of an 2010 Intel slide deck with the date and copyright removed. Tech fans seem excited by their content, but are ignoring the illegal complete lack of attribution regarding the stolen slides, as well as the fact that there is no actual news there.

Rotten idea for performance (0)

ChrisMaple (607946) | about a year ago | (#43716583)

CPU chips are performance-limited by heat. Adding the regulator on-chip, dumping heat into the chip without adding processing capability, is a net loss.By making the chip bigger, it decreases yield and makes it more expensive to produce. Lose-lose.
For moderate-performance applications where CPU yield is already high, the cost reduction achieved by simplifying the motherboard might make this a winner.

Re:Rotten idea for performance (2, Funny)

Anonymous Coward | about a year ago | (#43716693)

You should contact Intel - I bet they didn't even consider this.

So much for overvolting.... (0)

Anonymous Coward | about a year ago | (#43716621)

'Nuff said.

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?
or Connect with...

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>