Bored of laptops? Love 200Gb/s interconnects? Then you're going to hate today's Intel news Intel today waved its arms around in the air to remind us it has another family of products coming later this year – after quietly swinging the axe on another one. Chipzilla on Thursday teased a few more details about 11 of its forthcoming 10nm 10th-generation Core processors code-named Ice Lake, which are due to land in …
> eight hardware threads, four of which you may want to turn off for security reasons
Funny nobody talks about that anymore... People have gotten used to the idea that their CPUs are inherently unsafe, and that you need to buy 5 sizes larger if you need the computing power. The human mind is a wonderful thing (says marketing, happy they don't have to spend money fixing stuff...).
the mind has adaptable IO...
pity Intel who has to ratchet up their Cache size to achieve good IO
at least they have put the USB-C on chip so that at least means good IO speed to the monitor
personally I wish they would put a radio receiver that is capable of being programmed for AM/FM/DVB-T/AIS/GPS i.e. a decent SDR front end now that would be KILLER !
> Intel has baked in hardware fixes to the vulnerabilities by now
Why on earth would they do that? To throw money out of the window?
It's not like it would affect sales in any way, people keep buying like before, and will keep doing so no matter what happens. Have you heard of any bigger potential client saying they won't buy any hardware until that security mess has been reliably fixed?
The problems they have are:
1) There are a stack of known security problems with their chips
2) To protect against these issues you need microcode patches and to turn off SMT (hyperthreading)
3) Even before you do (2) above they have a competitor in the form of AMD that produces a range of CPUs that are close in single core performance, but much better in multi core, are less prone to the security flaws and are cheaper for the same level of performance
4) They can't match demand for chips because they don't have enough 14nm fabs and they have yet to bring the 10nm kit online in any meaningful way.
5) AMD can fill the demand with their own parts, so any capacity shortage is a lost sale.
So yes, they damned well will fix the security holes to prevent AMD getting too much of a foot in.
Those "Security Holes" are a joke.
I know it, You know it and so does any normie that read "Must be local to the device with Admin Privs".
Almost too convenient to have the dullards panic and the rest of us waiting on Bonuses.
But I digress. Tech ended when we let it become a consumer product.
When the security holes can be exploited by a script on a malicious web page (and at least one of them can) or by another virtual machine in a public cloud then they are no joke. It’s not even that hard to get users to download and run software that can use the holes for privilege escalations to compromise the whole machine, even if the code is supposed to be running in a sand box.
recently put together a Ryzen-based system. Hadn't upgraded hardware in a DECADE. Intel was ok on the old system, but I like the newer AMD stuff a WHOLE lot better!
Intel needs to up their game in the face of AMD. That INCLUDES fixing Spectre/Meltdown
Are you sure?
I'm hoping mGig gets some real traction - while it is available, it's a little too expensive at present, but 2.5/5Gbps looks like a nice balance between power/cabling/cost/requirements for desktops, small servers and wifi that need more than 1Gbps.
While 10Gbps/25Gbps and higher are required for high performance environments, the gap between teamed 1Gbps and 10Gbps provides a big hole for more cost sensitive applications.
"Bored of" is a neologism, but now in common usage, and generally perfectly acceptable.
Here's a five-year old grammar blog discussing the usage (and how it has changed from five years prior to that):
http://www.macmillandictionaryblog.com/real-grammar-quiz-question-3-bored-with-it-or-bored-of-it
It's probably fair to assume that "bored of" could well have taken over from "bored with" by now.
There is no confusion of meaning when using it, so go ahead, I say! The preposition is largely moot, you could say "bored of", "bored with", "bored from", "bored by", "bored at", and probably several others, and still be understood (and be syntactically sound, unlike the uterrings of the real grammar criminals who say "should of" rather than "should have").
Language evolves, or we'd all be speaking Middle English like Chaucer.
Language evolves, or we'd all be speaking Middle English like Chaucer.
Whan that Aprille with his shoures soote the droghte of Marche hath perced to the roote, and bathed every veyne in swich licour, af which vertu engendred is the flour;
Whan Zephirus eek with his swete breeth inspired hath in every holt and heeth the tendre croppes, and the yonge sonne hath in the Ram his halfe cours y-ronne, and smale fowles maken melodye, that slepen al the night with open ye, so priketh hem nature in hir corages, than longen folk to goon on pilgrimages.
There are quite a few people in Cambridge's English department who still do, and I've still got a smattering from my O level days.
Intel clearly left something out of their press release. The correct quote is "Now that AMD is making products that can easily compete with ours, 10th Gen Intel Core processors are foundational to Intel’s journey in enabling uncompromising and workload-optimized PC platforms with performance leadership across all vectors of computing. 7th and 8th Gen Intel Core processors were foundational to Intel's ultimate goal in enabling us to charge large amounts for stagnant performance."
"The 10th-gen family range, therefore, from the top-end four-core 2.3GHz (max 4.1GHz) 28W Core i7-1068G7 with built-in Iris Plus graphics, down to the dual-core 1.1GHz (max 3.2GHz) 9W i3-1000G1 with built-in UHD graphics. Each supports Wi-Fi 6 and Thunderbolt 3."
So the move from 14nm+++ (i7-8559U) to 10nm costs you 400MHz for base and boost frequencies?
I can see why Intel have skipped 10nm for a few years - the real question is can they get to their 7nm node quickly enough?
At the end of the day what matters most of the time is the speed from the CPU to ram and bus - MOST of the time that's still only running at a couple of hundred MHz (parallel equivalent) or with latencies around 40ns for random access - which is ~half what it used to be, but still a major bottleneck compared with how fast everything else has sped up.
time-of-flight matters too. 1ns per foot, more or less.
"At the end of the day what matters most of the time is the speed from the CPU to ram and bus"
It depends on the workload - if your workload is predominantly IO then maybe, but I would expect larger caches, wider memory buses or faster IO architectures to be the focus or performance increases (all are available from other vendors) if required for what is a general purpose CPU.
For compute based workloads, your performance depends on clock speed, IPC and how much time you can keep the CPU busy - caches and longer instruction pipelines will largely hide the actual speed of memory or IO from the processor. Again there will be exceptions that expose worst case behavior, but that's where software optimization plays a part.
For two generations of the same architecture from the same manufacturer, clock speed and IPC will tell a significant part of the story.
Given the process node advantage of the newer generation chip, you would expect either a significant jump in IPC to justify the clock speed drop or more cache/cores to boost performance with all those extra transistors.
Um, if there's little I and O, then you can compute Mandelbot sets, but not do much useful work.
You can't fix all that with cache - it slows down due to address computation the larger it becomes, which is why there are multiple (each one slower) layers already - for that very reason.
The glory of a general purpose computer is basically if() - it can do something different based on the input. That's why there's branch prediction and translation lookaside and all that - and nope, if we use a computer for a computer, and not just a glorified single purpose machine (where all the programming is hardware, like a totally pipelined FPGA perhaps) then that cache slowdown with size issue comes back, and as the original poster (Alan) said, it's all down to bandwidth outside the CPU being the limit - which has been true for quite a long time now. And trying to overcome that by keeping more inside the CPU with predictions and multiple choice pipelines to precompute just in case IS THE VERY REASON FOR THE SECURITY ISSUES. And would never have been bothered with if the latency to main ram and the other speed parameters weren't the limit.
Duh, as Drs mentioned, words do mean something.
What's the use of a computer that produces no output?
Go watch, or better yet, edit a video...it's nearly ALL IO other than encoding/decoding, which in all modern cpus or gpus have their own bit of logic, and in Intel - and NVidia - have embarrassingly not actually gotten faster in a long time. A gtx 1050 is as fast at that job as the newest shiny that costs many times as much...
Are you sure you are doing your video encoding/decoding right if you can't go faster than a 1050? It has a single NVENC vs the 2 or 3 in other models:
https://developer.nvidia.com/video-encode-decode-gpu-support-matrix
And as there are hardware limitations caused by offloading to specialist hardware, I'm not sure that matches the "general purpose compute" argument.
"If you make up the MHz in IPC, then it does not matter."
And the IPC difference between 8th gen and 10th gen is? With most Intel chips the IPC changes are around 5% per generation and you've lost close to 10% in clock speed.
While I agree with your comment in principle, the reality is that Intel have spent 4 years with 10nm struggling to match last years product on the previous process node.
"They claim the IPC uplift is 18%. The graphics is a lot better too though."
The 18% uplift is between Skylake (6th gen, first 14nm) CPU's and Ice Lake (10th gen, first "commercial" 10nm) CPU's. Skylake also offered 4,6GHz parts.
Reviews showing comparative performance between the 14nm and 10nm will be interesting because I suspect, after 4 years of tweaking, there isn't any additional clock speed headroom on 10nm.
The reviews will likely be 2-4 weeks after launch as Intel launched Ice Lake prior to sending out review samples...
Intel should wait until they have a seriously new technology, then introduce the Groom Lake range of CPUs. The UFO nutstheorists would go wild.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
