Hi. So, if AMD Renoir a better choice, why is it that AMD laptops are only paired with a lower AMD graphics chips or an RTX 2060 Max Q, and I have not seen any Renoir platform paired with RTX 2070 or higher from any manufacturers. I was kinda looking/hoping for Ryzen 4800H or4900H with a RX5700XTmobile or an RTX2070 Super. I'm looking for 1.
Just my observation.
Have a nice day.

Every AMD Ryzen 7 4800H laptop we've tested thus far have been outperforming the Intel Core i7-10875H and Core i9-10980HK, but there's a catch

I'm in AMD's court. My current desktop is Ryzen, I hope to have Ryzen in my next notebook. As a developer, multiple cores are important, but very often there are parts of a pipeline that come down to a single-core task, including presentation to the operator, especially in rapid change-compile-tests-interact cycles. So this is a clear advantage for Intel. If a task scales well to more cores, Intel can add more cores with decent IPC and they're still ahead, until AMD has better single core performance.

Another article had claimed that Dell's G5 SE 4800h is the fastest and the closest to AMD in hardware design.  ???

Quote from: anon on May 31, 2020, 04:00:24
I definitely want 4900HS in my next laptop but I'm not going to pretend that hashing isn't a sequential algorithm. It has to be by design to prevent preimage attacks. Hash Tab is one of the programs (Windows shell extension, really) that I install on every fresh Windows install--I hash something practically every day or two, to check for duplicates, for changes, etc, especially for OS install images which are large and take several minutes. Every time you use virustotal.com the first thing it does is hash the file, as well.

A single hash might be sequential, multiple ones are not (as in, you can hash multiple files at the same time, not as simple for OS iso's but how often do you do those anyway?)

4600H do outperform anything below i9. Why even bother to compare?

Why on Earth would you isolate Intel single core as a relevant factor?

In fact, the majority of benchmarks and actual programs tested on both AMD and Intel suggest AMD has SURPASSED Intel in single threaded performance with Zen 2.
That in fact, Zen 2 has higher IPC than Intel. Intel mainly 'compensates' for this by being able to clock their parts higher than AMD... but this in turn doesn't give them much of an edge (and it comes with efficiency losses).

In such circumstances (at least with laptops), Intel either matches AMD or slightly exceeds it by maybe 5% (in single threaded performance). Oh and that's only on the first run... subsequent tests push the CPU harder and because Intel needs more power to run, its performance tend to drop below levels where it could sustain its advertised clocks (AMD manages to keep its clocks about 200MhZ above baseline when all cores are stressed - and this is expected behavior when stressing all cores on either laptops or desktops).

In desktops, the single core performance is more pronounced (by maybe 10% in some scenarios) because Intel can clock their cores even more aggressively.

However, bearing in mind that this increase in clocks comes with much higher power use.

For example, even in laptop, Intel runs away from its TDP by a much larger margin compared to AMD (which stays at or close to its TDP on both laptops or desktops).

Efficiency wise (aka performance per watt), AMD is much more efficient.

To use single core as an 'excuse' to make Intel appear 'relevant' seems like unnecessary pandering to a corporation.
The article might not have even be paid by Intel (although it very easily could be), but various reviewers don't want to lose out on getting future free samples from Intel for testing/reviews.

If you had any journalistic integrity, and corporations actually cared, they would either give you a product to test for a few weeks before having to return it, or give it for free.

Unfortunately, for the rest of us, we don't get such fancy deals, so we need to eek out full prices and are at the mercy of OEM's (sometimes awful) pricing.

This is a pretty well established website... so I doubt finances would be an issue (which probably wouldn't prevent you from buying new technology as it comes out).

So if you artificially set things so that Intel wins, Intel wins

Mindblowing.

Boasting only about "single threaded" performance is like bragging about single-legged hopping in the world of bipeds. Do we need to talk about "Saharan silver ant speed" wrt body length?

All that matters

Efficiency: performance / watt
better price tag: performance / price
effective cooling MB & chassis design
an OEM who offers more configurable options

I definitely want 4900HS in my next laptop but I'm not going to pretend that hashing isn't a sequential algorithm. It has to be by design to prevent preimage attacks. Hash Tab is one of the programs (Windows shell extension, really) that I install on every fresh Windows install--I hash something practically every day or two, to check for duplicates, for changes, etc, especially for OS install images which are large and take several minutes. Every time you use virustotal.com the first thing it does is hash the file, as well.

Single core performance isn'T relevant anymore in 2020.

Just stop with intel horsesh**. They've been slowing things down in their monopoly and deserve every negative coming their way. It's little the consequences of their greedy choices.

Agreed.  My first thought on reading it was "How much did Intel Pay to get this spin printed?".    Sure, the majority of programs ever made run on a single core.  So what?  Dual and quad core processors have been out for well over a decade.   No programmer of any modern application ignores the extra cores.   And now we are leaping from 8 to 16 to 64 cores.  People won't take you seriously if you try to talk about single core performance comparisons.

AMD clearly wins this round, and it is about damn time.   We needed some performance competition to force Intel to start offering competitive pricing.   Well done AMD, and thanks.

no offense but the single core argument sound like a inter pr machine, i used to like intel until they started lying starting with that 56 core cpu that ran at 5ghz all core but then we found out they were using a 1000 watt one horsepower cooler to keep it cool.

comeonnow, yes singe core performance is relevant if the program itself only used a single core and you had no choice in the matter and the extra cores didnt matter but the point i was trying to make is that they purposely took a program that would use more cores redistricted it to a single core and tested it and said it was 15% percent faster like anyone would do that in the real world if that was the case we should all save money and buy a dual core celeron and enjoy that 15% speed boost by restricting all programs to single core hek if they made a single core processor still that would be even better!