First Gracemont laptop available for sale.
Anybody got disposable $500 to buy and test this laptop?
same tile with ptl?所以 WCL 规格有点泄露了
2+4LPE
4MB 内存侧缓存
整体
更好的 iGPU 与 N100
终于对 N100 进行了不错的升级
128-bit memory isn't strictly necessarily at this level. The N100 was quite a bit faster than the predecessor despite 64-bit memory because the CPU was that much superior, and the anemic GPU wasn't that memory BW bound. Not to mention newer DDR standards become faster too.So WCL specs kind of leaked
2+4LPE
4MB Memory Side Cache
Monolithic
Better iGPU vs N100
Finally a decent upgrade over N100
64 bit
If it makes feel any better, think of it as 2x32 bit The side cache should also help a bit, assuming it has comparable bandwidth to previous implementation. On LNL it wasn't much faster than system RAM, but in this case it might end up so.64 bit
Yeah you are right it's a value tier product and the media engine has been bumped to be same as that is in Panther Lake and the iGPU part is going from Xe to Xe3 there are other improvements so this is definitely a good value part.If it makes feel any better, think of it as 2x32 bit The side cache should also help a bit, assuming it has comparable bandwidth to previous implementation. On LNL it wasn't much faster than system RAM, but in this case it might end up so.
The specs for WCL are very decent for a small & efficient chip that can power all kinds of small devices.
On Intel platform for client yes DMR may have SMT though.Question... SMT is already dead no?
DMR have SMT.On Intel platform for client yes DMR may have SMT though.
Intel Darkmont, which is an E-core architecture used in Intel's recent CPUs, particularly in Xeon processors like Clearwater Forest, does not feature Hyper-Threading. Intel has made a strategic shift away from Hyper-Threading in some of its newer architectures, including Arrow Lake and Lunar Lake. This change is primarily due to factors like power consumption and area constraints, with Intel focusing on maximizing performance within those limitations by allocating more resources to the cores themselvesDMR have SMT.
But SMT exists in P Core in server loss of SMT is a big thing in Server in client it doesn't matter much if you compensate the lack for it.Intel Darkmont, which is an E-core architecture used in Intel's recent CPUs, particularly in Xeon processors like Clearwater Forest, does not feature Hyper-Threading. Intel has made a strategic shift away from Hyper-Threading in some of its newer architectures, including Arrow Lake and Lunar Lake. This change is primarily due to factors like power consumption and area constraints, with Intel focusing on maximizing performance within those limitations by allocating more resources to the cores themselves
More like they are monumentally embarrassed that AMD have managed to significantly overtake them in SMT gains.Intel has made a strategic shift away from Hyper-Threading in some of its newer architectures, including Arrow Lake and Lunar Lake. This change is primarily due to factors like power consumption and area constraints, with Intel focusing on maximizing performance within those limitations by allocating more resources to the cores themselves
The issue is that going Big Little benefit Qualcomm and ARM the most. Intel will go to the normal non SMT only Big core screwing the rest, the issue? If Windows goes that way, allows the competition to be stronger.More like they are monumentally embarrassed that AMD have managed to significantly overtake them in SMT gains.
Doubtless they are trying to use their walk back from SMT to convince MS to abandon SMT focused code in the Windows CPU scheduler in order to handicap AMD.
(yes I absolutely think Intel are that diabolical)
I'd be surprised if the ARM64 and x86/64 schedulers weren't pretty separate.The issue is that going Big Little benefit Qualcomm and ARM the most
That might be the case, the issue is how developed is Big Little. Intel is not that good. Uses the small cores not as good as expecting.I'd be surprised if the ARM64 and x86/64 schedulers weren't pretty separate.
Intel's Little Cores are fine they are infact very good the problem is the Big Cores they need to fix their Big Cores.That might be the case, the issue is how developed is Big Little. Intel is not that good. Uses the small cores not as good as expecting.
It's a stretch to say they are fine.Intel's Little Cores are fine they are infact very good the problem is the Big Cores they need to fix their Big Cores.
They are lol you are underselling it not to mention some of their ingenious design choices.It's a stretch to say they are fine.
They can be performant, but not particular efficient beyond per mm².
Oh really? Skymont is a solid core wait till you see Arctic Wolf and get blown away 🙂.At this point *mont cores seem to exist solely to boost core counts for a given area of silicon, which is ironically what people were lambasting AMD for a few years back
As I said, performant yes in perf/mm² metric.They are lol you are underselling it not to mention some of their ingenious design choices.
Oh really? Skymont is a solid core wait till you see Arctic Wolf and get blown away 🙂.
They are lol you are underselling it not to mention some of their ingenious design choices.
Oh really? Skymont is a solid core wait till you see Arctic Wolf and get blown away 🙂.
Not for ArctSurely you have a source?
You do realise that Skymont is an insane upgrade over Gracemont?As I said, performant yes in perf/mm² metric.
But in perf/watt they are significantly worse than the *Cove cores.
While the *mont cores were originally targeted for a more balanced PPA type efficiency along the lines of Cortex A7x/7xx, since Gracemont they are targeted at adding compute density per mm² specifically in order to increase core counts per package.
As to underselling it I've actually used Gracemont laptop only a couple of years ago as a work computer and it was miserable even doing basic Google apps, so I've got good reason to be unimpressed as the presentation for the 6 wide Tremont sounded very impressive too.
The reality we get with Tremont's next generation successor Gracemont feels oddly reminiscent of Samsung's 6 wide Mongoose core.
hmm https://chipsandcheese.com/p/skymont-in-desktop-form-atom-unleashed ?I haven't found good one yet on ring but I expect worse from both LNC/Skymont on ARL due to how broked ARLs implementation was in PnP characteristics.