Look at the single thread scores of those two chips. The pentium is twice as fast as the atom. It is the single thread score that determines 90% of the "feel" of a machine's speed. My G3258 system "feels" faster than any machine I've ever used, mainly because of its 100mS sunspider score from runnign at 4.5 Ghz. But there are of course moments where it lags out / spazzes out, but they are rare and well worth the blazing fast speeds I get 99% of the time.
At the moment, I believe Intel's dual core die is something like 50% iGPU
The whole integrated graphics thing, surprised me how much "enthusiasts" give a crap about em.
...
but they are still going to suck, and suck at a more rapidly increasing rate than the CPU they are attached to it seems like.
...
I came back and everyone is talking about power consumption and integrated graphics, I thought I was on the wrong forums lol...
What's the attraction?
You know a computer is slow when VirtualLarry is complaining.
Well, I think that you're wrong that IGPs are sucking at an increasing rate. I would argue, due to the die size proportions, that they are sucking at a far less rate. You can actually play some (2012-ish, maybe?) games, at low res and low details, on an IGP. Maybe higher than that on an AMD IGP. Well, supposedly.
no way they're going with no iGPU. any thinking to the contrary is a fantasy.
I agree with you. However, the X99 platform should easily satisfy your Alpha requirement.The sucking part I was getting at, that the IGP's performance compared to what is "current" at any given time decreases faster than the cpu it's built in. As in, two years later, your cpu is more relevant/useful but your IGP is waay out of date. Fine for regular folks and some specific uses, but it's puzzling why it's so interesting to enthusiasts. You can buy a generic Rwhatever or some Nvidia that does all that for like forty bucks. Keep that jazz off my CPU imo.
I've noticed that as well, lol.You know a computer is slow when VirtualLarry is complaining.
Something that I think is ironic:
Intel copying ARM and AMD on having large iGPUs.
Reason: The companies that design ARM chips are stuck with really slow cpu designs. There are also usually mobile oriented. So having a large iGPU actually makes some sense to them.
But I wonder if Intel copying that strategy could be their achilles heel. (ie, the large iGPU potentially hurts Intel's ability to compete more than their x86 decoder penalty).
In any event, I would hope to see Intel make a return to GT1 dies in the future. (This, in addition to their GT3 and greater iGPUs, etc).
Larry. I myself own a very nice looking desktop powered by an A4-5000. Its got USB3 at the front, optical drive, PS/2 outputs, VGA and HDMI (nice for legacy devices) and an extra DDR3L slot, I have 16 gigs of ram in this baby. Problem is, it does get rather slow, even for regular browsing (albeit it has 4 cores). So I use it just for storage, movie playback, and downloads. It consumes in the range of 15-20w with a large mechanical 7200RPM drive. Other than super low power use and size, I don't see any other advantages. I see it as more of an "accessory", rather than a proper computer.
It seems that Intel is going this direction as they are slowly making larger, more capable iGPs and it doesn't seem to be hindering their competitive ability in x86 (then again, the competition they're given can be sorted out with one of their lowest SKUs for most uses...)
My 8 years-old E2160 @ 3.0 Ghz is more than twice as fast as that Kabini... Years of progress, lmao :awe: I am still using that oldie purely because I still require some serial/parallel support and that mobo has all of that.I hear ya. My E1-2500 AIO, I just use for listening to internet radio primarily, because the speakers (and Sound Blaster Cinema audio enhancement software) are so good.
If the CPU were faster, it could have been a relatively killer little rig. But, 1.4Ghz Kabini is... slow. So I don't use it as a primary PC, like I do my G3258 and G620.
My G3258 is impressively snappy, even at 3.8 rather than 4.5 like some people.
I have a feeling Apple is one strong reason Intel is making those large iGPUs.
And that is fine, but I also think they should return to a very small GT1 (something like 6 EUs) at some time in the future for the lowend. (Currently all dies are based on GT2 or greater)
P.S. I don't think Intel will have problems with competitors until one of them tries to undercut with a small iGPU. And currently ARM cores are not strong enough to do this. And as far as AMD goes, I actually believe their large iGPUs are one reason they have a hard time competiting with Intel, that and their large cache sizes.
As a primarily mobile user, I'd love a big stocky iGP, but that'd be asking to have two designs; one with GT1 and one with the latest iGP, which is sort of asking quite a bit.
With Ivy Bridge, Intel actually had four different die layouts (dual core with GT1, dual core with GT2, quad core with GT1, quad core with GT2):
http://www.anandtech.com/show/5876/the-rest-of-the-ivy-bridge-die-sizes
With Haswell, as best as I can tell, all the dual core dies became GT2 with Intel disabling 10 EUs for Pentium and Celeron.
Moving forward, it looks like Broadwell will start at dual core GT2 (with Intel disabling 12 EUs for Celeron and Pentium). I would argue that 12 is too much for low end desktop and Intel should begin thinking about a 6 EU dedicated die.
Oh I didn't know that, that's very interesting. Do the quad cores have bigger dies or is it just that the dual cores have disabled cores?
Oh I didn't know that, that's very interesting. Do the quad cores have bigger dies or is it just that the dual cores have disabled cores?
The quad cores have their own dies.
However, it is very interesting that the Ivy bridge quad core with GT1 (132.8 mm2) is not that much larger than the Ivy Bridge dual core with GT2 (118.1 mm2).
With the iGPU even larger now, I wouldn't be surprised if a hypothetical Broadwell quad core with 6 EU GT1 were roughly the same size as a Broadwell dual core with GT2.
wow that a serious piece of crap. does microsoft sell windows with bing? is there any way a home builder load it onto their machine.
Something that I think is ironic:
Intel copying ARM and AMD on having large iGPUs.
Reason: The companies that design ARM chips are stuck with really slow cpu designs. There are also usually mobile oriented. So having a large iGPU actually makes some sense to them.
But I wonder if Intel copying that strategy could be their achilles heel. (ie, the large iGPU potentially hurts Intel's ability to compete more than their x86 decoder penalty).
In any event, I would hope to see Intel make a return to GT1 dies in the future. (This, in addition to their GT3 and greater iGPUs, etc).
iirc, moore's law not only predicts transistor density but also an optimum die size for any process tech. between that and the sheer number of interconnects necessary on a big iron processor like core, the silicon is going to be there. additional processing cores aren't really useful for typical desktop apps (and if you want/need more cores intel likes its market segregation just like airlines), nor are more SoC type features (and chipsets had typically been manufactured on the previous gen equipment anyway to get some additional use out of it), so the graphics is kinda free.