My desktop PC doesn't run on battery! Also...I think that MS needs to make some changes to their O.S. in order to understand and use efficiently this mix of cores and also some firmware/bios changes - so it is not only up to them... And to conclude: https://i.imgur.com/wVwibyo.png

That's ok for smartphones where you need to save battery, but i don't see the point in desktops, maybe in laptops.....

Next feature would return of big turbo button for turning on big cores..

ruthan:

Next feature would return of big turbo button for turning on big cores..

Intel TURBO Edition (with a massive big red button on front of the case). I'm loving it already. https://i.ytimg.com/vi/cDphUib5iG4/hqdefault.jpg

ruthan:

Next feature would return of big turbo button for turning on big cores..

You hit the nail here... Also this could happen if there is a bug and the "big cores" refuse to wake up with cpu intensive software: https://i.imgur.com/yJOmtNh.jpg

I think that big button would be blue and they license Big Blue name for it:)

ruthan:

Next feature would return of big turbo button for turning on big cores..

:) 386SX33MHz? I had one of these, 10MHz normal, 33MHz Turbo - with big button 🙂

butjer1010:

:) 386SX33MHz? I had one of these, 10MHz normal, 33MHz Turbo - with big button 🙂

Some 386SX is watching you. DONT ... PUSH ... MY ... BUTTON! 😀:D:D EDIT:

Hilbert Hagedoorn:

[...] Big.little architecture. [....]

Anybody else read "Little Big Adventure" and was immediately beamed back into the 90's? 😀

ruthan:

Next feature would return of big turbo button for turning on big cores..

Turning them OFF, turbo button slowed the CPU

Art Attack at Intel.

FINALLY I've been asking for this feature in x86 for years! I don't know if it's going to work the way I expect it to, but I'm glad this is the first step forward toward CPUs with different types of cores. Not all tasks are equal, so to create a one-size-fits-all CPU isn't very productive. I just hope the K/X models will allow you to independently overclock the small cores. That way, you get blazing fast single-threaded performance, and then use the big cores for highly complex tasks.

This is one of those rumors that either is completely false or has got it so wrong on what the CPU's purpose. Either way this big little concept only is needed in severely thermal constrained scenarios like in a phone or tablet so there is no chance Intel is doing something like that in a desktop CPU.

schmidtbag:

FINALLY I've been asking for this feature in x86 for years! I don't know if it's going to work the way I expect it to, but I'm glad this is the first step forward toward CPUs with different types of cores. Not all tasks are equal, so to create a one-size-fits-all CPU isn't very productive. I just hope the K/X models will allow you to independently overclock the small cores. That way, you get blazing fast single-threaded performance, and then use the big cores for highly complex tasks.

They already solved this with boost clocks. We can have 64 cores from AMD in a HEDT desktop CPU so it's not like we cant have enough cores in the thermal envelope. Now mind you in a phone, tablet maybe even a very small laptop I could see a case for the extra power savings but a desktop there is no need. Which is where this rumor is wonky because this is specifically a desktop architecture.

JamesSneed:

They already solved this with boost clocks. We can have 64 cores from AMD in a HEDT desktop CPU so it's not like we cant have enough cores in the thermal envelope. Now mind you in a phone, tablet maybe even a very small laptop I could see a case for the extra power savings but a desktop there is no need. Which is where this rumor is wonky because this is specifically a desktop architecture.

Boost clocks are not a way of solving this. I'm talking about cores that are physically different. Generally speaking, the more complex your architecture is, the lower it's maximum potential clock. If Intel's "little" cores lack a lot of advanced instructions (such as SSE4 or AVX), I'm sure you could squeeze at least another 1GHz out of them, while also drawing less power per clock.

Removing SSE4 would be stupid. however AVX1/2 are not needed at all to achieve a good performance per watt in multimedia decoding.

schmidtbag:

Boost clocks are not a way of solving this. I'm talking about cores that are physically different. Generally speaking, the more complex your architecture is, the lower it's maximum potential clock. If Intel's "little" cores lack a lot of advanced instructions (such as SSE4 or AVX), I'm sure you could squeeze at least another 1GHz out of them, while also drawing less power per clock.

Surely that could just complicate things more than needed? Having CPU's have steps has solved a lot of these issues, allowing them to clock down if only needing to open application or having them run full for a game for example. Even then stepping up and down depending whats needed. Though I don't think taking out an advanced instruction would make that much difference, I think its an issue of thermals with hitting those high clocks rather than clock instructions which aren't always used. Also removing them would also be removing their power making increasing the GHZ pointless surely?

Ricepudding:

Surely that could just complicate things more than needed? Having CPU's have steps has solved a lot of these issues, allowing them to clock down if only needing to open application or having them run full for a game for example. Even then stepping up and down depending whats needed.

It works well enough for ARM that companies like Qualcomm and Samsung use the same concept on their flagship CPUs. In the case of ARM, a well-programmed scheduler will use the little cores for background tasks, prioritize the big cores for foreground tasks, and use all cores whenever something is extra demanding. We don't know exactly how Intel is adjusting the small cores. It's possible they're just making their own iteration of what ARM is doing, where there's only going to be low power and high power cores. If the small cores are literally smaller, that means their architecture is changed in some way. That could mean fewer instructions, it could also mean smaller caches (which would also be beneficial to performance). Either way, fewer transistors means less heat and fewer errors, which therefore allows for higher clock speeds and better low-clock efficiency. Depending on the task, this can yield better overall performance. To start with, I wouldn't be surprised if Intel's small cores will be based on 10nm, due to the limited yield that produces. But if Intel gets a decent scheduler in place, they could really surge ahead in performance-per-watt. x86 cores are becoming too big, trying to compensate for such a wide array of workloads without being especially good at any of them.

Though I don't think taking out an advanced instruction would make that much difference, I think its an issue of thermals with hitting those high clocks rather than clock instructions which aren't always used. Also removing them would also be removing their power making increasing the GHZ pointless surely?

Have you seen the wattage difference when something like AVX is used? It's startling how much a CPU heats up because of such instructions. If a core is built to not handle such workloads, there's less of a concern over what it's potential peak thermal output will be, which will allow it to overclock higher. Most advanced instructions aren't used in everyday software. SSE3 is pretty much as high as most programs go, and SSE3 is also about as advanced as non-x86 CPUs go. So, most programs built to be multi-platform don't support such instructions. However, those instructions are expensive to add to a CPU. They're complex to engineer and they take up a lot of transistors. Remove them from the cores for applications that don't need them (like most) and you get a much simpler, cheaper, and lower-power core. But then retain them for the applications that do need them.

schmidtbag:

That could mean fewer instructions, it could also mean smaller caches (which would also be beneficial to performance). Either way, fewer transistors means less heat and fewer errors, which therefore allows for higher clock speeds and better low-clock efficiency. Depending on the task, this can yield better overall performance. To start with, I wouldn't be surprised if Intel's small cores will be based on 10nm, due to the limited yield that produces. But if Intel gets a decent scheduler in place, they could really surge ahead in performance-per-watt. x86 cores are becoming too big, trying to compensate for such a wide array of workloads without being especially good at any of them.

I can see where you are coming from but least on the desktop market i think power is less of an issue, low-clock efficiency last i checked was very good, and intel do Y models which are made for just that low-clock power. to me it's an odd market, now if these were coming out on laptops I could totally see the market, i just fail to see the appeal on the desktop front. I know the mobile market do it, and again same with Laptops i can totally see the appeal, having power if needed. But Desktops i feel dont have this power issue and they have low power models if needed.

schmidtbag:

Have you seen the wattage difference when something like AVX is used? It's startling how much a CPU heats up because of such instructions. If a core is built to not handle such workloads, there's less of a concern over what it's potential peak thermal output will be, which will allow it to overclock higher. Most advanced instructions aren't used in everyday software. SSE3 is pretty much as high as most programs go, and SSE3 is also about as advanced as non-x86 CPUs go. So, most programs built to be multi-platform don't support such instructions. However, those instructions are expensive to add to a CPU. They're complex to engineer and they take up a lot of transistors. Remove them from the cores for applications that don't need them (like most) and you get a much simpler, cheaper, and lower-power core. But then retain them for the applications that do need them.

I get your point but the heat comes from doing AVX workloads, and they normally can make a CPU downclock due to the heat. But removing it beyond maybe making the CPU less complex to make, i still think other instructions run into just as many heat issues. Same way running games on SSEX instructions can cause high temps even with no overclock. I don't see them removing instructions and it being a massive jump, unless they most anything high, in which case what would be the point in the CPU? You'll run it at what 6ghz but be able to run nothing on it, where is the point in that. Now i can see them being cheaper and less power hungry, but unless all you want to do is run chrome which seems to work on SSE1/2 far as i could tell from a little google. I fail to see the joy in it. or more to the point what would be the need for the high GHZ. Most likely there is a reason they have not gone this route, and we can see that mobile processors running windows do have issues, good example of this is the Surface X which has to run older versions of adobe to work.

You has not to be Nostradamus to see lows compatibility issues and incompatibilities with older SW, as when HT came or SMP with Ryzen, unless they will make it fully abstract on CPU level.. But i already see some threads named like.. pleas help my games is running on weak core.. how to switch it.. return of CPU affinity p*orn.

The Windows scheduler is garbage. Big/Little works on Android because they have their choice of many competent Linux CPU schedulers to pick and optimize from. Windows can barely tell the difference between a SMT/HT thread and a free core...and even when it can, many programs are hard-coded to run on a specific core (and Windows is not smart enough to force these to re-assign). Which is part of why Ryzen has performance issues on Windows with certain programs. Windows also likes to re-assign threads to different cores repeatedly.. Because this makes sense on monolithic Intel to try to spread the load out between all the cores, but when some cores are superior to others, it causes other issues