Acer Predator (Vega 56+Ryzen 2) Helios 500

Well, the modules I have only run at 2400, xmp profile is 3000. So not sure what the xmp profile has to do with it.

 

Because your system doesn’t support XMP and is limited to 2666MHz max. No XMP = 2400MHz. My Corsairs are the same way, difference is my system supports XMP so they run at 3000MHz CL16. That’s why the Kingston modules are best for systems like yours, as they do 2666MHz CL15 without XMP.

 

IIRC if you put the NVMe drive in the slot that the factory SATA SSD came in then the BIOS should see it.

 

So, now we know that Kingston HyperX Impact (SO-DIMM DDR4 2666 MHz) should work:

https://www.kingston.com/dataSheets/HX426S15IB2_8.pdf

Why? Because these sticks have JEDEC PnP option.

But what with ADATA RAM? Look at this:

https://www.adata.com/pl/feature/484

These memories also have the JEDEC option. Of course they have higher CL, but maybe you can lower them. What do you think about it? Will it work?

 

Crucial has quite a long list of 2400/2666 memory compatible with the Acer Helios 500, Click See More - there's a bunch of compatible sku's:

https://www.crucial.com/usa/en/compatible-upgrade-for/Acer/predator-helios-500

The Acer specs I see reference a max of 2666mhz on SODIMM's supported for the Acer Helios 500

 

I guess we are still waiting for Zen 3 release date, however what's the chances you could upgrade to Zen 3 from the 2700?!

 

Neither Asus nor Acer have come up with a plan for upgrading CPU's in the Ryzen laptops, and I wouldn't count on them providing a BIOS update to support Zen 2.

BTW, it's Zen 2 not Zen 3...

Zen 1 was the 1xxx series CPU's, Zen + was also called Ryzen 2 the 2xxx series, and Zen 2 is Ryzen 3 the 3xxx CPU's.

 

Thanks for the help, That's a real shame :/

 

I know what you mean, but at the same time.. there really is no 'plan' to come up with.
All they need to do is write the needed microcode updates they already released into existing desktop MOBO BIOS-es that includes Zen 2 support (which would likely not require too much of a tinkering, if any because the mobos in the laptops are desktop based anyway, classified as B350 and B450) and case closed.
Not only that, but both Asus and Acer should release not only updated BIOS-es which support Zen 2 on GL702ZC and Helios 500, but also unlock them (so users have full access to the BIOS without restrictions).
Acer should be taken into higher account due to their messup with RAM support on Helios 500 and RAM speeds.

As a side note... I ordered the Helios 500 with Ryzen/Vega for myself just a few hours ago.
Should be here tomorrow.
It was £200 off on PCWorld website (previous price was £1700).

 

For a company to support user upgradeable CPU's, have resources allocated ahead of time to apply to BIOS updates - and any other software that needs updating to support CPU changes, requires adding those options to the overall plan for the product, with additional heads up to finance, marketing, documentation, customer support, engineering, and QA / QC + Validation Testing / Verification Testing / Regression Testing to make available supportable updates in those sections besides hacking a few bits in the BIOS

Acer staff resources are likely already thin on the ground, and to redirect and apply them to add support for CPU upgrades that will reduce laptop sales (models with the new CPU's) and increase costs due to resources requiring attention for the CPU upgrades, it's pretty much a "no brainer" that it won't happen.

The desktop motherboard expectations are different than for laptops, and so we rarely get full BIOS or CPU / GPU upgrade updates down the road.

It's sometimes difficult to remember that the awesome desktop features don't map into laptops - "even though they should", and it's been this way for a long time.

Encouraging people to open up their laptops to do a tear down to the point of pulling and replacing the CPU is likely a "thing of nightmares" for most of the laptop maker non-engineering staff.

 

OEM's really need to step up their game for laptops.
I'm sick of useless notion that releasing new BIOS updates that include support/functionality/features for new CPU's would reduce sales.... it doesn't hold ground because the same thing could be applied to desktops which are FAR more modular than laptops (and people tend to usually upgrade their destkop components, leaving the mobo as the 'last resort'). So why are desktop mobos getting new BIOS-es which support newest CPU's and give people the ability to upgrade the CPU while laptop mobos (even though they are pretty much identical to desktop ones) dont?
I mean, the principle is the same.
If OEM's are getting incentive from AMD to upgrade the BIOS, they would still do so if they use the same approach for laptops.

Many people won't really open up their laptops to replace the CPU in laptops, they will sooner just go out and buy a new one... but, Acer and Asus could have offered a service that if the laptop is within the warranty period and you want to upgrade the CPU, they should technically do it for you (especially if you buy the CPU off their website and use 'upgrade' as a reason in RMA).

Devoting resources to writing the BIOS updates is not hard.
Asus and Acer released several BIOS updates for both GL702ZC and Helios 500 since they were released integrating better security and other features for 'improved stability' and 'improved performance' (as they claim).
So, obviously, the teams needed to do the work are ALREADY there... and given the fact that both laptops (with all AMD hardware) have in all probability been a huge success, they have more than enough revenue to give us an unlocked BIOS with proper RAM and newest (Zen 2) CPU support (which in all honesty, shouldn't be difficult - they already rolled out upgrades for different versions of MOBOs they have on the market... I doubt they need to rewrite the BIOS from scratch every time... its more like porting AGESA from AMD and integrating it into existing BIOS and then testing to see if it works... in all probability they standardized the process).

Laptops don't usually get same treatments probably because they intentionally don't want laptops with increased upgrade paths (even though they easily support this).
It's laziness and greed... not 'technical' or 'monetary' issues.

 

Well said! Sort it out Acer!

 

It's not Acers job to make laptop BIOS updates to support CPU configurations they never sold.

Desktop builds from vendors also don't get BIOS updates for CPU's that were never sold in configurations of those desktops.

If you are lucky the desktop builds that use "generic" off the shelf desktop motherboards "might" be able to use the BIOS directly from the motherboard maker, but often that's also not a straight forward update as the desktop build vendor will often put in their own format BIOS that isn't compatible to be updated with the "generic" off the shelf motherboard BIOS.

AFAIK there are Zero off the shelf "generic" laptop motherboards that get continuing new CPU BIOS updates through the lifetime of the socket, as we expect of the "generic" off the shelf desktop motherboards.

The fraction of overall sales that each custom laptop motherboard represents is too small to support the costs of continual BIOS updates past the configuration's sold for the laptop.

Until there is a pool of generic off the shelf "laptop" motherboards that can be used in a wide range of laptops - whose form factor would now be "locked in" to the form, size, and function over the life-time of a socket, we won't see many if any laptop makers spend the resources to keep releasing new BIO's to add new CPU's for configurations they never sold.

Maybe AMD could require vendors to do just that, by edict force the laptop makers using AMD socketed CPU's / APU's support the full generation of AMD CPU's available for a socket generation.

Other than that I can't think of anything that would motivate vendors to do it. Engineers lost the rule of the house long ago, good ideas and intentions are overruled by the bottom line.

 

My impression is that the AMD Helios 500 is a limited release version but one that Acer thought and perhaps think serves a small niche. If so maybe there is nothing pushing them to release a "Helios 600" (whatever) release with significant upgrades but enough interest to update the original 500 design with a Zen 2 processor, after all it would just require a BIOS update and a series of internal tests.
If not I guess we have to buy USB programmers and do the hard work ourselves, shouldn't be too hard if the required BIOS editing tools are out there (haven't checked).

 

There is a new motherboard chipset to take advantage of all of the Zen 2 features... the CPU is compatible with previous AM4 motherboards, but a new laptop would be more likely this round of CPU feature updates given the addition of a new motherboard chipset.

Besides PCIe 4.0, more PCIe lanes, it might be like the transition between x3xx and x4xx with x3xx not supporting all the new features of the auto-tune / boost features of Zen+. x5xx might have additional tuning / performance / boost features only available on the new chipset.

470 vs 570 motherboards with zen 2
https://www.reddit.com/r/Amd/comments/ago4lc/470_vs_570_motherboards_with_zen_2/

aborto 2 months ago*
"As everyone else has mentioned the 500 series boards will get PCIe 4.0. But to add to that:

While the current boards may get updated to support it they will only have it on the first PCIe x16 slot, and given no current graphics cards see much of a noticeable degradation from running at even x8 it is unlikely to make a difference for years to come.

The main benefit I think we will see will be the doubling of bandwidth to the chipset (AM4 currently uses 4x PCIe 3.0 lanes to communicate to the chipset). I assume the new boards will change to PCIe 4.0 for the chipset link. This will double the rather constrained bandwidth available for all the SATA ports, networking, USB, secondary M.2 slots etc and should allow systems to run a lot more IO without conflict.

This should also double the bandwidth available for NVMe SSDs which will be important going forwards as current NVMe drives are starting to saturate the 4 PCIe lanes they are connected with."

 

I think there may be a lot of options out there for Intel MOBO's in regards to BIOS editing tools, but I'm afraid that AMD ones might be different due to the company's lower market share (still, it should be doable).
Plus, there hadn't really been too great of a need because at least on desktops, majority of OEM's did release BIOS upgrades for Zen 2 support already (even Asus apparently released newest BIOS update for B350 with AGESA version that corresponds to the one that should technically support Zen 2 - in addition to releasing BIOS-es for X370, B450 and X470).

I wonder though how safe it would be to install a modded bios with Zen 2 support on Helios 500. The prospect of bricking the laptop is not an appealing one (hence why I'd prefer Acer to release it and test it properly).

 

Its Acer's job to ensure customer satisfaction and provide us with a decent experience like any desktop user would have - we certainly paid them (and Asus) handsomly for the products we use (especially if you take into account that customer support services across many OEM's are usually ridiculous).

They both knew that AMD was adamant about future upgrade paths, so Asus unwillingness to implement Zen 2 support is a poor cop-out with no real substance (We still don't know what Acer will or will not do).

Asus by your own explanation shouldn't have released BIOS updates to support Zen 2 CPU's they never sold (which technically aren't out yet) for desktops motherboards either, but they still released those BIOS-es and support Zen+ and Zen 2 through majority of their mobos (even B350 if latest reports are accurate).

Doesn't really matter if laptops come with generic mobos or not- they are based on desktop mobos anyway, they are just locked out to prevent user level access tinkering.

Each laptop has its own BIOS releases from the OEM up until a certain date (the usual cutoff point is 6 months, even though this is unrealistically low because the warranty is usually for 2 years - so they should technically release BIOS updates for as long as the Warranty lasts).
Providing a user with an option to upgrade should be there... especially since both Asus and Acer used a socketed desktop CPU.

 

The tool required are not AMD specific but from the BIOS/UEFI maker so the tools used to unlock the Intel version should be perfectly fine, problem is if they are available just for a few BIOS hackers or not. If they aren't available one could perhaps ask someone with those tools for help but it is a much slower way to do it and requires cooperation. It is also my understanding that Acer have locked the BIOS so they have to be modified for each machine individually making things even slower.

One would extract the relevant modules from those other BIOSes and replace old ones in the Helios 500, should be mostly plug and play in theory. In practice? Perhaps some other tweaks would be needed to make a stable system.

That's why a DIY solution would require an external programmer plus an extra computer - at worst one just reprogram with a standard BIOS.

 

Interesting that the tools aren't Intel or AMD specific, but yes, if those tools are only available to few bios hackers, then its a problem.
Though, I don't understand why would each Helios 500 with Ryzen/Vega need individually modified bios despite them being locked?
They pretty much have same configurations (sans the SSD for example - in some US versions, there's only 256GB SSD and no HDD, in UK, there's the 256GB SSD and 1TB HDD... and of course there are more expensive ones with 512GB SSD and 1TB SSD)... but I don't think that the BIOS would need to be individually modified for that... would it?

That's how I figured it would be... an existing bios from Helios 500 B450 mobo is extracted, then the microcodes are dumped from say a similar B45 (same manufacturer) desktop mobo into ours.
Question is, would the updates for RAM stability/support translate in those transfers, or are you saying the editors would just dump the microcodes?
I mean, they might as well unlock the BIOS in that case and give users full control over RAM speeds and timings for instance.

I meant more in the sense that if a BIOS hacker modifies a bios for Helios 500 Ryzen/Vega and just uploads it for us to use.
We'd also need a BIOS recovery method just in case (though, I can never seem to find a procedure on how to do this).

 

That's what I did but it's not appearing in the BIOS.

 

I got that from "The bios firmware is unique to your computer. You can’t use a modded bios from another Helios 500 on your own":
https://octoperf.com/blog/2018/09/02/acer-predator-helios-500-mods/

That is about the Intel version but it should be the same in the AMD version.

That would be the logical first step: just enable the features that are in there but disabled.
My understanding is that installing the latest AGESA adds support for Ryzen 3xxx and any stability updates in one go.

I think an external programmer would be needed if the BIOS isn't an Acer original. But they aren't that expensive and AFAIK the ROM chip itself is easily accessible. In my limited experience BIOS recovery solutions on laptops are not reliable... Could just be bad luck.

 

But would that not apply only if the hardware configuration is different (if they come with different kind of GPU's or CPU's - components that are directly soldered to the mobo)?
AMD's Helios 500 comes with a replaceable 2700 CPU, and that baseline config is pretty much identical 2700 CPU with Vega 56 (apart from SSD/HDD combo, and possibly RAM amount - though 16GB seems fairly consistent).
The Intel versions differ more because they come with different CPU's and possibly GPU's... not to mention different components... different sound card on the mobo (as its a different motherboard for Intel CPU's and who knows, maybe Acer uses different kind of mobos for different CPUs on the Intel/NV version).

AMD is far less complicated as they allow/insist on CPU compatibility on the same chipset with most/all Zen CPU's... whereas Intel aren't similarly minded (and correct me if I'm wrong, but Helios 500 with Intel has soldered CPU's do they not?).
Wouldn't things be easier with non-soldered CPU's, because, that only leaves the variations in GPU hardware... and Ryzen/Vega comes with V56 in every config (while the CPU is removable - and other components sans the HDD/SSD combo should be identical).

I don't think the replaceable components such as the CPU, RAM, SSD and HDD would really play a part.
But, laptop BIOS-es are different in respect that their UEFI's are programmed with the Windows activation key... that's the only individual thing that differs (which I can think of) laptop to laptop with AMD hardware (as everything else would be pretty much the same - sans of course SSD and HDD)... but even if you remove that Win activation key, I don't think it would break the computer or prevent it from running... it would just give you a 'clean slate' like what you get with a desktop mobo, so your Windows would no longer activate on its own (you'd need to get in touch with acer and request the key from them or from Microsoft?).

My thinking the same... it might actually get rid of that ridiculous RAM bottleneck when running RAM in quad channel as Zen 2 might get quad channel support and much higher RAM speed support (in which case, it would be great... but Acer DID mess it up by placing 2 RAM slots beneath the keyboard, necessitating disassembly which is entirely inconvenient for a laptop of this caliber - Acer could have probably put all 4 RAM sticks on the easily accessible end or just stick with only 2 slots).
Still, since AMD reworked the Infinity Fabric by turning it into an I/O, from their own words, this was done to remove the Infinity Fabric dependence on RAM speed... so, hypothetically, Zen 2 performance wouldn't really care much if you have faster RAM with lower latencies anymore to improve performance in that kind of software.

Well, people have installed custom BIOS-es on laptops before, and even managed to recover a laptop BIOS... so its doable... my problem with the whole idea is that we'd need some kind of solution/software to test the custom BIOS functionality before one tries to install it, and we need a more elegant solution for BIOS recovery on laptops in case it doesn't work... I mean, OEM's are able to do it... but I REALLY detest doing things manually like that... ideally a stick should be used that does this automatically - or the BIOS should have a backup reinstated if a faulty BIOS that bricked the laptop was installed.

 

I've been toying with MSI afterburner on my Helios 500... and unfortunately, the clocks section is wonky at best.
Please have a look at the pic for more information to see what's happening.

Every time I try to set the clocks to say 1450 MhZ and apply that, the slider just goes back to minimum for some reason.
However, I CAN increase the VRAM frequency to 900 MhZ and it sticks to that after applying it (voltage modifications also stick after applying).

Its the clocks slider that goes all the way to the left after trying ti change that one.
I could try installing an older version of MSI afterburner if it helps?
Otherwise, I might have done something different in MSI Afterburner settings that I'm not aware of?

Btw... I updated the AMD drivers from outdated Acer ones using the Express option just like on GL702ZC and preserved FreeSync functionality.

Wattman did come up in the driver after tinkering with MSI afterburner, but its very weird.

Any advice on these issues?

P.S. I played Darksiders III for a while last night and the game was running beautifully maxed out.
The laptop was also very quiet... barely any more audible than when using it for Internet, writing, 3d Studio Max and videos

 

Figured out the problem.
The Frequency bit in Wattman was set to Dynamic... whereas it needs to be activated (the nodge needs to be facing right side).
I still maintain that the Frequency and Voltage areas in Wattman look wonky at best.
Aren't they supposed to be represented as graphs with default values showing up so we can edit them?

I was able to increase the core and memory frequency of the GPU in MSI Afterburner, as well as undervolt it all at the same time.
I think I reached 1500 MhZ on core and 900 MhZ on VRAM, with undervolt set to -90 I think... but, I may just drop the core frequency to 1450 instead and keep the VRAM at 900 Mhz.
Setting the core UV to -100 crashed the system while running Furmark (which showed about 136 FPS... but the GPU wasn't maxed out and the temps were quite low).

I might have to try testing the GPU using something different that will stress it properly/fully where I could control MSI Afterburner voltages on the go.
That way, I could setup the GPU to desired frequencies on the core and VRAM with an undervolt.

The machine really has some serious headroom.
Obviously, I don't want to push it unnecessarily high...

 

OK... Ran Superposition benchmark.
Got 3114 with default values.
But I noticed the GPU VRAM in Superposition benchmark under those settings did not exceed 500MhZ.
In MSI Afterburner, the VRAM is set to 800 MhZ but the GPU doesn't achieve it.
Also, the core frequencies don't go up to 1301 MhZ as displayed in stock MSI Afterburner... they hold around 1100MhZ in Superposition

Is this regular behavior?

 

After some tinkering with MSI Afterburner and following settings:
Core voltage: -75mV,
Core frequency: 1440 MhZ
Memory clock: 900 MhZ

I got 3683 score in Superposition Benchmark under 1080p Extreme setting.
I supplied a screenshot.

Core frequency during superposition run goes up to over 1200MhZ, but not coming close to 1440MhZ.
And of course, the memory clock is switching between 700 and 900 (only the 7 and 9 keep interchanging, the 00 stay like that).

I hope I hadn't gotten another bad GPU iteration.

 

I had something similar happening to my machine when I tried to overclock it and test it out in RE2. I found that even if I would leave overclocking unlocked in MSI the whole system would become unstable. I think it's probably due to higher power consumption, I was hitting the max as soon as I changed anything in the settings. Vega 56 by itself is rather power hungry and it seems there's just not enough left in the tank without increasing the power limit.

Sent from my SM-G955F using Tapatalk

 

Most people dont bother messing with core clock, as you get much more out of overclocking the memory anyways.

 

Congrats on the new laptop!

 

Actually, Vega 56 is quite efficient, especially when undervolted.
I think the main problem here is that its been TDP constricted by a bit too much, so for that reason, it might not be able to reach its stock clocks for some reason and needs undervolting... but the funny thing is the temperatures are extremely low with stock or overclocks.
And for some reason, I cannot change the power limit at all (its locked) - otherwise, I'd put it to +25% and run the GPU with that.

I did just that on another Superposition run, and got 3567 Score (1080p Extreme).
Basically, I undervolted the GPU to -81mv on the core, left the core clocks on stock and increased the memory to 900MhZ.
I could try pushing the memory more, but not sure how stable it would be.

My question is, how do my stock scores compare?
I mean, without an undervolt or anything else tweaked, the Superposition benchmark got 3114 score... whereas the reviewers got 6.5% higher score.
So I'm wondering if that's 'normal' for this laptop or not.

 

Last user that post feedback was on reddit I think ? Actually overvolted to get much higher memory clocks

 

During your second run did the overclocked memory stick or was it all over the place again? I tried that too and no luck.

Sent from my SM-G955F using Tapatalk

 

In Superposition benchmark 1080p Extreme run, it seems that my memory clock is holding steady at 900 MhZ with a core undervolt at -81mv.
Core clocks are stock (1301MhZ).

These values seem to stick for me repeatedly and I got a score of 3585 (13% higher than my stock score).
Try those values and see if it works.

I've also uploaded the screenshot so you can copy the settings directly.
If it doesn't work, try reducing the voltage beyond -81mv (your GPU might have higher tolerance).

I just don't know if my score is considered good with that undervolt and memory oc.

 

Thanks! I'll try these settings probably tomorrow or the day after and I'll post what I got here, stock results included.

Sent from my SM-G955F using Tapatalk

 

Here's my TimeSpy score (at stock values).
-81mV Undervolt doesn't work there if the memory is overclocked to 900MhZ.
The benchmark hungs even at around -60mV... so I think I'd need to go lower than that for TimeSpy (possibly to -50mV or -25mV) - if I want to preserve the 900 MhZ memory OC.

It seems that TimeSpy is more capable of stressing the GPU properly since its end results report proper stock values in terms of core and memory frequencies. But I don't know why Superposition doesn't stress the GPU similarly in that case?

Anyway... I'll play around these tomorrow as its 3am here and need to sleep as I'm gonna help someone with work at 10am.
Got carried away.

 

Helios 500 on ebay for 750 new in box

 

Item location:
Taiwan, China
Posts to:
United States


I don't think that its a smart thing getting a new laptop like this without any warranty.
If something goes wrong... what are you supposed to do?
Ship it back?
The import duties on getting a laptop overseas are quite large... and even if they shipped it to UK, the customs alone will slap the tax and import duties on top of it.
Since its for now available for shipment to USA, I'd imagine import and tax duties would put it close to the original selling price in USA anyway, if not, then maybe up to $1000... or $1300?
Which is certainly cheaper, but then you have to ask yourself if a lack of warranty is... well, warranted?

 

Never had to pay duties yet, so i wouldn't know. Last product I got was 300 and paid no duties.

Also, warranties are merely a veil these days, as you know. And can careless about them when I just tell ebay it's broke and it goes back no questions asked unlike when dealing with companies where you have to threaten lawsuits for them to fulfill their obligations let alone a refund.

The bigger and better question is are replacement parts source-able. If not then it's a problem overall warranty or not. If so, then I don't care about the warranty, as I can buy a part myself and have it up in a week at most instead of having a paperweight for 1-6 months.

I understand that my opinion is not representative of all, but that's my point of view on the matter. Lastly you can get 2-3 year protection plans on ebay if my memory serves me correctly.

 

You do make a good point about returns being a lot better handled by ebay, but that raises the question... what if the product fails months or a year after purchase?
Doesn't seem like it would be applicable for refunds in such an event.

As for replacement parts... lol, they wouldn't be source-able in all probability... and even if they were, they would probably be cost prohibitive.

 

Your first point is addressed in the quoted post. You can buy 3rd party warranties for most items.

Replacement parts usually comes down how large the OEM is, its one of the reasons why I tend to stick to Dell, despite the issues I take with their practice.

 

So I'm wondering if this thing is going support Zen 2 right around the corner since it's a regular desktop socket. Anyone know? Can I just put a new Zen 2 cpu in this thing when the time comes. I wonder if Acer would create a bios to make it work if it was possible...

 

Nobody knows, simple as that. Notebook manufacturers tend to not support upgrading their machines as it adds complications when it comes to service and warranty plus spending money on development and testing without any immediate payback.
I hope Acer will keep selling this model upgraded with a Zen 2 processor but if they don't it should be possible to modify the BIOS as the hardware should be compatible with a software update, how hard that would be is another question.

 

New BIOS 1.08 available, no AGESA update.

https://www.acer.com/ac/en/US/content/support-product/7737?b=1

 

Sorry for snaching the thread but AMD fans might be interested in this new budget experiment:
http://forum.notebookreview.com/threads/asus-tuf-fx505du-mini-review.829041/

 

Not bad. But with MSI putting out another mainstream laptops with the high end support for a socketed CPU, I'm looking for if BIOS support will allow this laptop to support the upcoming Ryzen Zen 2 based chips, or if Asus or another company (or Acer) will put out a Ryzen gen 3 laptop supporting up to the 12-core variant (the 12-core TDP matches the 8-core Intel chip, but is a multi-threaded beast).

 

500 more nets you Vega 56 and ryzen 2700

But it's a good budget option for those that don't want Intel cpu

 

See, the GPU is where I felt the Asus laptop failed, by pairing it with Polaris. For the Vega 56, at least that is around 1060 Ti to 2060 if memory serves (desktop, I haven't looked at notebook GPUs in awhile).

For the CPU, the TDP matches that of the 3700X. If that is able to be dropped in, talk about a nice performance boost!

But, Nvidia still has more powerful GPUs overall, so not allowing a pairing with their 2070 or 2080 may be a mistake. AMD could lose sales simply due to not having the choice on GPU, if being honest.

 

At stock the Vega 56 is below the 1070 but can punch up above it when oc which is irrelevant for most people don't touch that sort of thing.

Either way though it's actually good timing on Asus / AMD's part considering the latest exploits in addition to spectre/meltdown that were discovered. Some of the market may at least consider it an option where as it was written off before.

The thing that makes no sense to me is that cpu running 92c at sub 20w its like these damn odm/oem forget that you have to actually cool these chips.

 

I definitely hear you on the Vega 56, which is why that made this Acer viable, whereas the 580 on the Asus laptop the year before, I really couldn't recommend it. Now, with Navi, if you pair something with the power of the 2070, considering the castration of the mobile Nvidia cards, with the new Zen 2 chips, I'd argue you have a really competitive gaming laptop.

There are still questions out there and we will get more information on June 10th, then with the release in July.

On the new exploits, there was a similar exploit talked about last August or September, near the time that PortSmash was discovered (the L1TF vulnerability, IIRC), that also postulated the current exploits found with Zombieload and company, with that researcher saying that HT on Intel was a security risk and everyone should turn it off. Nine months later, we find out that he was absolutely correct.

I do think those vulnerabilities are precisely why Dell last year said they thought Rome would not do much in the server market and they wouldn't expand their offerings, then this year, they are tripling their offerings from 3 to 9 servers based on AMD designs, out of 50 designs, roughly (so around 20% of their offerings for servers are now based on AMD).

As to cooling, that is something everyone here has complained about. In the search for creating thin and lights, companies have, instead of creating robust yet compact cooling solutions, pursued gutting the cooling system so long as the chips run within their stated spec, even if pushing chips to their limits. This also shortens the life of products, in theory, which is another way of forcing planned obsolescence into the product stack.

With that said, I am happy to see "gaming" laptops now shipping with liquid metal from the company. Also, MSI's socketed laptop is using a robust heat pipe system which may work well. So hopefully they will start working on cooling innovation.

 

The GL702ZC had a bad cooling design, period.
I don't think however it was particularly thin. I've opened mine before I returned it for a full refund and I saw it had plenty of space.
Asus just never executed it properly.

They certainly could have installed higher quality heat-sinks for instance, as well as better/quieter fans.
As for pairing 1700 with Polaris... that wasn't a particularly bad decision... I was ok with the performance (even though I think they could have put higher clocks on the RX 580 while still keeping a lower stable voltage), but I also don't think the price of the laptop was representative of its value. All in all, I think Asus should have charged much less for it... round about £1000/£1000... maybe £1250/$1250.

And of course don't get me started on Asus lack of Zen+ or Zen 2 support.