Slashdot Mirror

It seems contrived and misleading of you to limit your comparison to current quad-core CPUs, as Intel hardly makes those anymore. If instead comparing the launch cost of the 2600K (~300 USD), then a current match would be the 8700.

For that you get approx. +80% multi-threaded performance and +35% single-threaded performance, for -30% power consumption.

It seems contrived and misleading of you to limit your comparison to current quad-core CPUs, as Intel hardly makes those anymore. If instead comparing the launch cost of the 2600K (~300 USD), then a current match would be the 8700.

For that you get approx. +80% multi-threaded performance and +35% single-threaded performance, for -30% power consumption.

The "best" CPU is always the best one for *your* workload. If that's max single-threaded performance and money doesn't matter, then that means Intel, and it likely will for awhile. If we're talking about a workload that can be processed massively in parallel, then AMD has earned a seat at the table. I like the "High End CPUs - Intel vs AMD" benchmarks at PassMark -- should enable plenty of dick-waving no matter who you are. Take the time to understand your workload in detail, set your budget, and choose. Whatever you buy will be obsolete in a few years anyway. ;-)

3rd party developers managed to hack PCID to do the same thing, yet the performance improvement (or rather reduction of mitigation's degradation) is not as big as using INVPCID. That's why Linux will now try to use INVPCID if available, then PCID if not or just suck up the mitigation cost.

3rd party developers, other than Microsoft. That's my whole point. They swallowed the Intel version, didn't care to investigate any possible other mitigations that'd work on earlier versions, and haven't adopted what everyone has done even after it became clear there's other approaches. I understand why Intel wouldn't work too hard on supporting older CPUs, but Microsoft trying to push itself as some sort of service should be focusing on trying to keep its customers happy.

It'd be different if Meltdown was going to be fixed meaningfully soon, but only the latest Intel CPUs now finally have a hardware fix and actual wide adoption is going to take years at the least. It'd also be different if the CPUs in question were actually substantially inferior to modern ones but their new this year i3-8300 is pretty comparable to the i5-3570 (what I have) released in 2012 yet right before the Haswell cut-off. That doesn't even include the more high end i7-4960X released in 2013 which is also before the cut-off and near twice the performance.

Microsoft probably had the patches ready from Intel way before the PCID method was even conceived. Remember that Intel was notified a long time before those vulnerabilities were made public, they even extended the deadlines for publication multiple times.

I get that, but the point is that they've had over 6 months to introduce the PCID method into Windows. They're announcing only the Retpoline Patch method and that's not coming until next year. If Intel had to wait for Microsoft to have done the patching themselves without Intel's help, would we still not know about Meltdown or Spectre?

3rd party developers managed to hack PCID to do the same thing, yet the performance improvement (or rather reduction of mitigation's degradation) is not as big as using INVPCID. That's why Linux will now try to use INVPCID if available, then PCID if not or just suck up the mitigation cost.

3rd party developers, other than Microsoft. That's my whole point. They swallowed the Intel version, didn't care to investigate any possible other mitigations that'd work on earlier versions, and haven't adopted what everyone has done even after it became clear there's other approaches. I understand why Intel wouldn't work too hard on supporting older CPUs, but Microsoft trying to push itself as some sort of service should be focusing on trying to keep its customers happy.

It'd be different if Meltdown was going to be fixed meaningfully soon, but only the latest Intel CPUs now finally have a hardware fix and actual wide adoption is going to take years at the least. It'd also be different if the CPUs in question were actually substantially inferior to modern ones but their new this year i3-8300 is pretty comparable to the i5-3570 (what I have) released in 2012 yet right before the Haswell cut-off. That doesn't even include the more high end i7-4960X released in 2013 which is also before the cut-off and near twice the performance.

Microsoft probably had the patches ready from Intel way before the PCID method was even conceived. Remember that Intel was notified a long time before those vulnerabilities were made public, they even extended the deadlines for publication multiple times.

I get that, but the point is that they've had over 6 months to introduce the PCID method into Windows. They're announcing only the Retpoline Patch method and that's not coming until next year. If Intel had to wait for Microsoft to have done the patching themselves without Intel's help, would we still not know about Meltdown or Spectre?

3rd party developers managed to hack PCID to do the same thing, yet the performance improvement (or rather reduction of mitigation's degradation) is not as big as using INVPCID. That's why Linux will now try to use INVPCID if available, then PCID if not or just suck up the mitigation cost.

3rd party developers, other than Microsoft. That's my whole point. They swallowed the Intel version, didn't care to investigate any possible other mitigations that'd work on earlier versions, and haven't adopted what everyone has done even after it became clear there's other approaches. I understand why Intel wouldn't work too hard on supporting older CPUs, but Microsoft trying to push itself as some sort of service should be focusing on trying to keep its customers happy.

It'd be different if Meltdown was going to be fixed meaningfully soon, but only the latest Intel CPUs now finally have a hardware fix and actual wide adoption is going to take years at the least. It'd also be different if the CPUs in question were actually substantially inferior to modern ones but their new this year i3-8300 is pretty comparable to the i5-3570 (what I have) released in 2012 yet right before the Haswell cut-off. That doesn't even include the more high end i7-4960X released in 2013 which is also before the cut-off and near twice the performance.

Microsoft probably had the patches ready from Intel way before the PCID method was even conceived. Remember that Intel was notified a long time before those vulnerabilities were made public, they even extended the deadlines for publication multiple times.

I get that, but the point is that they've had over 6 months to introduce the PCID method into Windows. They're announcing only the Retpoline Patch method and that's not coming until next year. If Intel had to wait for Microsoft to have done the patching themselves without Intel's help, would we still not know about Meltdown or Spectre?

Passmark rates it as a high-end CPU. I would say it's a high-mid range CPU.

Processors have improved dramatically since 2006. I selected 3 chips that were all relatively high end for a desktop but reasonably affordable and popular chips (not extreme CPUs) from the stable of Intel corp. Namely: Q6600, I5-2500K, I5-8600K. The Core 2 Quads came out late 2006/early 2007, Sandy bridge in 2012 and Coffee Lake 2018, so a relatively even timeline distribution. Shortly after launch the Q6600 was $280, 2500K $220, 8600K $260.

Take a look at the benchmarks and performance scores, not to mention platform changes. We also went from no standard SATA SSDs during C2Q's reign to NVME SSDs for the 8600K. Just because you only use excel on small data sets which can still be done with a C2Q doesn't mean that there haven't been large gains.

https://www.cpubenchmark.net/c...

are you speculating on the gains, or are the gains all speculative? and if you remove speculation, are there any gains?

Processors have improved dramatically since 2006. I selected 3 chips that were all relatively high end for a desktop but reasonably affordable and popular chips (not extreme CPUs) from the stable of Intel corp. Namely: Q6600, I5-2500K, I5-8600K. The Core 2 Quads came out late 2006/early 2007, Sandy bridge in 2012 and Coffee Lake 2018, so a relatively even timeline distribution. Shortly after launch the Q6600 was $280, 2500K $220, 8600K $260.

Take a look at the benchmarks and performance scores, not to mention platform changes. We also went from no standard SATA SSDs during C2Q's reign to NVME SSDs for the 8600K. Just because you only use excel on small data sets which can still be done with a C2Q doesn't mean that there haven't been large gains.

https://www.cpubenchmark.net/c...

PassMark - CPU Mark Single Thread Performance
AMD Athlon XP 2800+ @ 2.25 GHz / rel. October 1, 2002 / Score: 627
Intel Core i7-8700K @ 3.70GHz (4.7GHz turbo) / rel. October 5, 2017 / Score: 2708 (highest-scoring processor for single thread performance as of June 5, 2018)

Single thread performance ratio: 4.01

**Single thread performance doubling time: 7.5 years**

Note: it's not clear what the best-performing processor was in the early 2000s, the performance doubling time may be even greater.

Moore's law ceased to hold for computing performance quite a while back. Lots of cores doesn't speed up sequential computing tasks (Amdahl's Law).

. I'm looking forward to trying out AMD for the first time,

You won't be disappointed with AMD this go around. Take a look at the specs for the new 2700X.

https://www.cpubenchmark.net/c...

That is a $329 chip and has better performance than the closest intel chip in that class, the 8700K. The 8700K is also $30 bucks more. Sure there are more powerful intel chips but those chips are in the $1000+ range.

  You can also find AMD chips in that range too but if you are going to do a bitch'n build and not break the bank the 2700X seems to be the way to go.

Don't forget that for that $329 you get 2700X and a really good CPU heatsink/cooler. Where as the 8700K costs $30 and comes with no cooler at all so you have to add another $50 for that.

. I'm looking forward to trying out AMD for the first time,

You won't be disappointed with AMD this go around. Take a look at the specs for the new 2700X.

https://www.cpubenchmark.net/c...

That is a $329 chip and has better performance than the closest intel chip in that class, the 8700K. The 8700K is also $30 bucks more. Sure there are more powerful intel chips but those chips are in the $1000+ range.

You can also find AMD chips in that range too but if you are going to do a bitch'n build and not break the bank the 2700X seems to be the way to go.

You could be right, or maybe not...
I believe most of average home power consumption is poorly designed appliances and choice... It's so easy not to think, and just plug anything in, ignore idle power consumptions...

Take fridges for one, they can be built to use 0.1KWH.day and be large enough for a family; and they wouldn't cost much more to make if they were mass-produced in quantity. Compared to most residential fridges use ~1KWh/day+.

Take my journey in computers... Desktop/monitor ~100-200W(3.6KWh/day), then went to laptop ~60Watt(1.44Kwh), then power efficient ~14" laptop ~15W(0.36KWh). Then I discovered intels power efficient chips, and run a tablet ~2-6W(0.096Kwh) connected to a 40" power efficient monitor(~18W ~16hrs/day= 0.288KWh)
So laptop(0.36KWh/day) vs tablet+40"(0.384KWh/day)

The user experience with a 40" screen, wireless kb/mouse is soo much better.

https://www.cpubenchmark.net/p...

Most of my chinese security cameras take ~7W vs Panasonic's 1.1W... This all adds up. Available products and choice... The consumers are mostly to blame, because they choose not to think about power/cost ratios.

Analog microwave 0KWh/day idle
Digital microwave: ~0.2KWh/day (73Kwh/year ; $11-22/year+- depending on location; New analog microwave ~$40)

My digital washing machine takes more power on idle than a load of laundry in a day(I unplug after every use, and I only need to do a load once a week)

People don't even need to be off-grid to make this stuff pay for itself. Run the long term numbers of devices and you can often get new power efficient devices/toys for the power savings from the grid alone. Again they choose to be wasteful/not think about it.

Yes, or more specifically:
https://www.cpubenchmark.net/c...
https://www.videocardbenchmark...

How about we demand a 10% refund on our chips? I wonder how that would fly. I think replacement would be a better offer though.

I don't think a 10% refund covers it. Depending on how cutting edge your processor was when you bought it, you may have paid a pretty steep premium to get an extra 10% performance vs lower speed processors. If you look through the chart, small performance differences can have huge costs associated with them.

https://www.cpubenchmark.net/h...

I think you'd be hard pressed to convince US politicians of either party to go full on trustbusters on Intel. Especially as they'll claim they're not a monopoly.

Though they're definitely turning into one

https://www.cpubenchmark.net/m...

In B) they want price per watt. Intel has done much better than AMD on price per watt. AMD hasn't had a server chip in 6 years.

Don't compare the consumer lines to the professional. They're designed by different groups. My wife's Inspiron required a near complete teardown to get to the hard drive.

http://www.dell.com/support/ma...

I'm considering a new laptop this Black Friday

Unless you need the latest and greatest the old ones work fine. I have a M6700 from 2012. The i7-3940xm still benchmarks fairly well. You can get one used laptop with CPU on Ebay for under $1k.

My current laptop benchmarks better than that and it's from 2012.

While a dual Xeon 256GB machine is not the "standard" desktop, nor even the "standard" workstation (maybe the dual CPU part), it is definitely worth pointing out that you can fit way more metal into a box, more practically than before.

But also- the summary is a lie. Here's Microsoft's page for the Surface Book 2, which the summary touts as "desktop brains":

https://www.microsoft.com/en-u...

Their most powerful option- the 15 inch, with 1 TB storage, 16 GB RAM, and a 3,300 dollar price tag- offers "i7 quad core" and "GTX 1060". nvidia has been engaging in a new type of shenanigans with their mobile cards, implying that they are the same as desktop cards- and of course, they are not. Arguably, they are shafting their desktop users by even making them close. Meanwhile "i7 quad core" applies to the i7-8650U, I think:

https://www.cpubenchmark.net/c...

What's the "i7" desktop equivalent? Well, it's got 6 cores, and I think:

https://www.cpubenchmark.net/c...

With like triple the whatever-goodness-numbers is appropriate.

Anyway, the takeaway is that the 3000+ version of this thing uses a laptop CPU that isn't really close to its desktop equivalent, and nowhere near close to what you can shove into a desktop, and a mobile version of a card that is close to, but not surpassing, the desktop version.

While a dual Xeon 256GB machine is not the "standard" desktop, nor even the "standard" workstation (maybe the dual CPU part), it is definitely worth pointing out that you can fit way more metal into a box, more practically than before.

But also- the summary is a lie. Here's Microsoft's page for the Surface Book 2, which the summary touts as "desktop brains":

https://www.microsoft.com/en-u...

Their most powerful option- the 15 inch, with 1 TB storage, 16 GB RAM, and a 3,300 dollar price tag- offers "i7 quad core" and "GTX 1060". nvidia has been engaging in a new type of shenanigans with their mobile cards, implying that they are the same as desktop cards- and of course, they are not. Arguably, they are shafting their desktop users by even making them close. Meanwhile "i7 quad core" applies to the i7-8650U, I think:

https://www.cpubenchmark.net/c...

What's the "i7" desktop equivalent? Well, it's got 6 cores, and I think:

https://www.cpubenchmark.net/c...

With like triple the whatever-goodness-numbers is appropriate.

Anyway, the takeaway is that the 3000+ version of this thing uses a laptop CPU that isn't really close to its desktop equivalent, and nowhere near close to what you can shove into a desktop, and a mobile version of a card that is close to, but not surpassing, the desktop version.

a couple years, anyway.
http://cpubenchmark.net/market...
this isn't based on sales, but rather activity in passmark's testing software. still painfully (for amd fans) obvious exactly when 'core' came out and how long intel's dominance has been.

Your scores are questionable. Passmark shows a ~20% better overall performance from the 1400 for a CPU that costs $4 more.

https://www.cpubenchmark.net/c...

A chart comparing price vs performance. Note that it's mostly AMD on the "good" side of the graph.

Really?
https://www.cpubenchmark.net/c...

And the date on that is... today! The Ryzen 1700 has better performance but the FX-8300 still has a better price/performance ratio.

I think what's happening is they do get twice the performance per area, but it's also twice as expensive per area. Who cares about Moore's law if "performance per dollar" stays the same? Heck, the best value per dollar for a processor scoring at least 10000 on PassMark is a processor from 2012.

E3-1505M v6

I hope AMD gets Intel to stop twiddling their thumbs. The E3-1505M v6 benchmarks at 9798 / 2166 single threaded.

I'm typing this on a 4 year old M6700 with a Intel Core i7-3940XM that benchmarks at 9324 / 2009 single threaded. It cost me all of ~$800 last year. Room for 4 hard drives, 32 GB of RAM, 17" screen. Thunderbolt and USB_C really don't seem like they're worth the $2k price tag.

And it runs Linux and BSD just fine.

Of course, for firewall any modern Celeron will do. I guess that the Atom based ones even have an edge because they have more cores. Depending on what deamons you run, that might turn out better than Core based ones that are pretty much all Dual Core without Hyperthreading.

And I have a haswell.

It beats the crap out of my of older sandy bridge during compile. This is using a compiler that does not rely on Ram, but on CPU only.

That's a meaningless statement. You didn't say which Haswell or Sandy Bridge. I have a Core 2 Extreme chip that will kick the shit out of Haswell Haswell.

And I have a haswell.

It beats the crap out of my of older sandy bridge during compile. This is using a compiler that does not rely on Ram, but on CPU only.

That's a meaningless statement. You didn't say which Haswell or Sandy Bridge. I have a Core 2 Extreme chip that will kick the shit out of Haswell Haswell.

still all a generation behind PC hardware.

It's not like Intel has that much differentiating their generations.

The MacPro comes with a E5-2697 v2 that is still competitive.

The reason they're seeing a decline in sales is PC has plateaued performance wise. 3-4 old generation Intel chips are still competitive. I'll *never* realize the energy savings with how much I use my desktop. My laptop has a 3940XM that is still very competitive speed wise and is 4 years old.

In the same amount of time I've upgraded my GPU on my desktop 3 times because of advances in CUDA.

I can't be the only one that hates Intel's numbering.

AMD is pretty damn obvious. Get an FX (soon Zen), and pick the highest number you can afford in your cost/benefit calculations.

Meanwhile, Intel is "i3/i5/i7" with a number afterward (and sometimes a T, sometimes an ML, MQ, HQ, Q, QM, QE suffix) , plus Mobile editions, plus Xeon editions. 'X' Xeons, 'D' Xeons, 'E' Xeons. Plus VERSION NUMBERS (E Xeon v3, E Xeon v4).

Good God, Intel, get your shit together.

Take a look for yourself and tell me whether any particular entry (without looking it up for details) is better-or-worse based solely on the model number.

https://www.cpubenchmark.net/h...

Yes there are significantly better Xeons around: look here. The best CPU you can buy for the mac pro is number 27 on that list. The top Intel proc has 20 cores compared to 12 and runs 50% faster.

The question is, how many PCI-e lanes do they support? I didn't see that on the list you linked; and don't have time to run-down a bunch of CPU product pages on the Intel site. Perhaps you know...

From what I have read, we are only JUST now, LATER THIS YEAR (2017) going to have a "Skylake"-based Xeon with enough PCI-e lanes to make TB3 worthwhile on the Mac Pro. And if the Mac Pro can't enter the world of USB-C/TB3 yet, it isn't worth upgrading... yet .

So, let's hope Apple has gotten some Engineering Samples of the new "Skylake" Xeons, and is busily incorporating them into an updated Mac Pro...

Take a look at this: https://www.cpubenchmark.net/c... It's faster than even the latest i5-6600, and cheaper. Only the i7's are faster. And don't tell me about single core performance, it's 2017, any program I care about is multithreaded. Now if you talk about energy consumption, you may have a point...

Yes there are significantly better Xeons around: look here. The best CPU you can buy for the mac pro is number 27 on that list. The top Intel proc has 20 cores compared to 12 and runs 50% faster.

They don't have to do much work if any to 'make them fit'. It's about power efficiency(which has a relation to size). Microsoft partnering to emulate x86 is a step backwards, it will only limit options. Native/intel REAL hardware exists they just need to stick it in the right case.

https://www.cpubenchmark.net/p...

Phones I linked were just as close as they came to developing a working x86 phone, as far as I can tell noone found a way to actually natively install any other OS's on them tho. Like Linux On Android, they could get Win7 running parallel connected through VNC, no sound...

I'd be tempted to try a kickstarter if one could find a chinese manuf. that could throw a 4-5" intel tablet together to spec.

If you're looking for a small bump in the mean time you can get a i7-3940XM. It shares the rPGA988B socket with the 3630QM. They're ~$300 on eBay.

In Benchmarks it's ~20% faster than what you have currently.

• PassMark 3630QM: 7599/Single Thread: 1686
• PassMark 3940XM: 9372/Single Thread: 2025

It's still faster than the 6820HQ (8765/1891) or E3-1505M (8699/1887) CPUs that Dell ship in their newest 7000 Precision lines.

If you're looking for a small bump in the mean time you can get a i7-3940XM. It shares the rPGA988B socket with the 3630QM. They're ~$300 on eBay.

In Benchmarks it's ~20% faster than what you have currently.

• PassMark 3630QM: 7599/Single Thread: 1686
• PassMark 3940XM: 9372/Single Thread: 2025

It's still faster than the 6820HQ (8765/1891) or E3-1505M (8699/1887) CPUs that Dell ship in their newest 7000 Precision lines.

AMD Phenom II X4 945 is a processor model.

Surface Pro4's top CPU: i7-6650U

Average CPU Mark: 4889
Single Thread Rating: 1821

My laptop's CPU: i7 3940XM

Average CPU Mark: 9378
Single Thread Rating: 2025

My desktop's CPU: i7 4770K

Average CPU Mark: 10121
Single Thread Rating: 2255

[Side by side benchmarks of the above]

I have no idea what you're using your machine for but it certainly won't fit a lot of people's use cases.

Surface Pro4's top CPU: i7-6650U

Average CPU Mark: 4889
Single Thread Rating: 1821

My laptop's CPU: i7 3940XM

Average CPU Mark: 9378
Single Thread Rating: 2025

My desktop's CPU: i7 4770K

Average CPU Mark: 10121
Single Thread Rating: 2255

[Side by side benchmarks of the above]

I have no idea what you're using your machine for but it certainly won't fit a lot of people's use cases.

Surface Pro4's top CPU: i7-6650U

Average CPU Mark: 4889
Single Thread Rating: 1821

My laptop's CPU: i7 3940XM

Average CPU Mark: 9378
Single Thread Rating: 2025

My desktop's CPU: i7 4770K

Average CPU Mark: 10121
Single Thread Rating: 2255

[Side by side benchmarks of the above]

I have no idea what you're using your machine for but it certainly won't fit a lot of people's use cases.

Someone that needs/wants a 10" ultra portable isn't going to be happy with a 17" mobile workstation.

I like my Dell M6700 with a i7-3940XM. 32 GB of RAM, 4 hard drives and space for 2x wifi cards. 17" screen. Full keyboard, with number pad. Trackpad and clit mouse (if you're into that). I only wish I could get a higher resolution screen.

When Apple stopped making their 17" laptops I jumped to Dell.

My current machine is a M6700 with a 3940XM, 32GB of RAM, 4 hard drives, 2 wifi cards, IEEE1394, 5x USB, eSata, Display Port, VGA, and HDMI. I've tested it with Windows, Linux and BSD and all work just fine.

Then again it's the antithesis of what most Apple laptops are, the battery life sucks but it's a mobile workstation and I need that.

With all those specs, new, it cost ~$5k.

I'm not familiar with the HP Stream 11 but a quick web search reveals the i5-2520M in the X220 is 4x faster than the N3050 in the HP Stream 11.

X220 w/i5-2520M @ 2.5 Ghz: https://www.cpubenchmark.net/c...
Stream 11 w/N3050 @ 1.6 Ghz: http://www.cpubenchmark.net/cp...

I'm not familiar with the HP Stream 11 but a quick web search reveals the i5-2520M in the X220 is 4x faster than the N3050 in the HP Stream 11.

X220 w/i5-2520M @ 2.5 Ghz: https://www.cpubenchmark.net/c...
Stream 11 w/N3050 @ 1.6 Ghz: http://www.cpubenchmark.net/cp...

These are pros. Not airs for consumers

[sarcasm]And pros never buy during the Christmas season. And consumers never buy the MacBook Pro.[/sarcasm]

Yes the newer GPUs use Samsung 14 nm processes compared to the 28 nm from previous generation. Big boast in performance.

[Citation Needed]. My information says that the Radeon Pro 455 which is Radeon Arctic Islands architecture and is manufactured on TSMC 16nm FinFET process not a 28nm process.

These are supposedly for professionals. Not capable of any real work dealing with VMware fusion, Adobe premiere, compiling code, or anything else a professional would use.

Let's look at this argument. You are saying that pros would benefit greatly enough from using Kaby Lake over Skylake. The fastest mobile Kaby Lake is the Core i7 7500u(cpu score: 5381) vs a Core i7 6700T (cpu score:8971) which Apple used instead the MacBook Pro. The mobile CPU Apple used clobbers the best Kaby Lake mobile CPU. Do you want to guess why? Because Intel hasn't released a Quadcore mobile Kaby Lake CPU yet. Intel isn't expected to release the Quadcore versions until December which means Apple might have them ready for consumers in April at the earliest.

At best pros will have to wait six to nine months for the next MacBook Pro. Or Apple could release a MacBook Pro with the processors it has now.

These are pros. Not airs for consumers

[sarcasm]And pros never buy during the Christmas season. And consumers never buy the MacBook Pro.[/sarcasm]

Yes the newer GPUs use Samsung 14 nm processes compared to the 28 nm from previous generation. Big boast in performance.

[Citation Needed]. My information says that the Radeon Pro 455 which is Radeon Arctic Islands architecture and is manufactured on TSMC 16nm FinFET process not a 28nm process.

These are supposedly for professionals. Not capable of any real work dealing with VMware fusion, Adobe premiere, compiling code, or anything else a professional would use.

Let's look at this argument. You are saying that pros would benefit greatly enough from using Kaby Lake over Skylake. The fastest mobile Kaby Lake is the Core i7 7500u(cpu score: 5381) vs a Core i7 6700T (cpu score:8971) which Apple used instead the MacBook Pro. The mobile CPU Apple used clobbers the best Kaby Lake mobile CPU. Do you want to guess why? Because Intel hasn't released a Quadcore mobile Kaby Lake CPU yet. Intel isn't expected to release the Quadcore versions until December which means Apple might have them ready for consumers in April at the earliest.

At best pros will have to wait six to nine months for the next MacBook Pro. Or Apple could release a MacBook Pro with the processors it has now.

Not sure if you just cherry picked those two particular CPUs as they are almost identical in performance:
http://www.cpubenchmark.net/co...

But i agree, upgrading from a 4 digit (2nd gen+ core i7 series cpu), to another core series i7 CPU is not going to give you any sort of performance increase. However, the newer chipset with 4x m2 support can drive a high end SSD over 2000MB/s (some over 3k now!). So upgrading from a core 2 quad was a good move for me. Even though the quad really did work fine.

I did manage to skip a whole upgrade to core series around 2012 and waited it out till q1 2016. But the machine i would have bought at the time, a 2600k is still selling for a premium on ebay. CPU is still worth $150 USD last time i looked.

it may be that people need to revise their upgrade schedule to be more like 8 years as opposed to the old model of a 3-4 year cycle, but that doesn't bother me one bit.