AMD's Ryzen CPUs (Ryzen/TR/Epyc) & Vega/Polaris/Navi GPUs

For those who're interested in getting the 1st gaming Ryzen set for evaluation and experience...
The Legion 5(AMD) is now available in Canada, US, and selected Asian Lenovo websites as customizable.

However do take note that the 2nd Storage Drive is by default, un-removable as there isn't removal option available even if you chose just to have the 1st Storage Drive installed or removed.

Note that the 1st Storage Drive, NVMESSD have to be chosen to get the battery upsized to 80whr type.

If you choose to have a mechanical HDD installed into the 1st Storage Drive, the default battery size sticks to the initial default package of 60whr.
If you choose to have the 1st Storage Drive stays vacant for future self upgrade, the battery size sticks to the initial default package of 60whr.

Well... the title of this clip is very misleading...

...as there isn't a commercially available RX5600m for evaluation yet...
except desktop's RX5600x.
And the reviewer clearly states that the RX5600m/Ryzen 4800H evaluation is still in PLANNED!
Hmmm... What's gonna that make then?​

New evaluation and benchmarking...

...with RX5700x!​

 

So GN further discussing the AMD bios situation:



Sent from my SM-G975U1 using Tapatalk

 

The MSI Bravo's are indeed using RX 5500M.

 

Perhaps my poor grammar had mislead you. What I meant was that the benchmark picture identifies the GPU as generic, hence it can be either 5500m or any other such as the Vega graphics in the CPU. Yet, the benchmark pic is used to identify that the GPU is 5600m which may perform differently.

 

Actually, I don't think those are accurate desktop GPU scores because the pre-bios update 5600XT scores the following:
https://www.techradar.com/uk/reviews/amd-radeon-rx-5600-xt

And post bios update:
https://www.techspot.com/news/83616-radeon-rx-5600-xt-new-vbios-boost-shows.html

There are clear differences/discrepancies between what WccfTech posted and actual results we got during release... but then again Wccftech isn't the most reliable source to begin with.

Also, its way too early to tell either way because 5600M still isn't released so we actually don't have a full suite of benchmarks and gaming tests to put the GPU through its paces, whereas you've taken a GPU that's been out in the wild for a longer time frame and underwent a full battery of tests.

Also, performance of RTX 2060 (mobile) varies from system to system depending on the thermal solution in place (as you also noted from the fact you mentioned that WccfTech posted results from a cheap notebook using 2060).
We need to wait for DELL to release the machine so we can see where the actual performance lies (not just in FireStrike and similar benchmarks but actual games) and see if the unit is thermal throttling.

For all we know, those preliminary scores Wccftech posted (if they are accurate to begin with) are a result of thermal throttling or early tests with an unfinished GPU.
So, we need similar thermal conditions to test the GPU's accurately.

Plus the WccfTech article is from 2nd January.
We're past mid May... so 4 and a half months later.
How about we wait for the actual release of the GPU to test it fully before jumping to conclusions which scores higher and what is a product of thermal throttling or not?

 

...with a large, replaceable battery, the full complement of 4 DIMM slots (user upgradeable), plenty of storage bays, and a wide gamut IPS screen.

 

Good video. I do firmware work on Atmel / Microchip / Espressif microcontrollers at my job and very often there are technical limitations that you just can't get around. It is also very true that firmware work usually has a huge backlog of work and is very understaffed with people who have the skill / desire to work on it. I try to have a junior dev work with me on projects and they are overwhelmed with how complicated things are compared to scripting languages, Arduino or even limitations compared to C++/Java in desktop apps, or having to hook up a scope to debug some strange bug in code because something does not match up with what is said in the datasheet.

 

There are two new laptops with Ryzen 4800H from Eluktronics, they have been announced just today: Introducing the Matrix RP-15 & RP-17. AMD Ryzen 4800H paired with a FULL performance NVIDIA RTX 2060 GPU. No blocked vents, proper thermals, color accurate 144Hz display, Covert Gamer design.
Both are based on TongFang chassis:
Eluktronics RP-15 15 inch: Tongfang GK5NRFO / NR0O
Eluktronics RP-17 17 inch: Tongfang GK7NPFR / NR0R
Here is a review of the RP-15 by Bob of All Trades

Edit: It seems that these laptops are also coming to Europe through Schenker/XMG as the updated CORE 15/17, check the announcement on reddit.

 

Ryzen 7 4800H beats the AMD Ryzen 9 4900H in latest 3DMark Time Spy leak: Dell laptop with AMD RX 5600M is 10 percent slower and no match for the Intel i9-10980HK

https://www.notebookcheck.net/Ryzen...-match-for-the-Intel-i9-10980HK.465925.0.html

"Tipster @_rogame recently spotted a 3DMark Time Spy benchmark listing for a Dell system powered by the AMD Ryzen 9 4900H and an AMD RX 5600M GPU. The 4.4 Ghz, eight-core processor with a peak boost frequency of 4.4 GHz scored only 8388 in Time Spy's CPU test. This disappointing result is actually 10 percent slower than benchmark results for the lower-end Ryzen 7 4800HS and far behind the Intel Comet Lake S i9-10980HK.
On the GPU side of things, however, the RX 5600M powers past the NVIDIA RTX 2060 Max-Q and the NVIDIA GTX 1660 Ti in Time Spy's GPU test, with a score of 6098.

How do we explain the seemingly low CPU test results for the Ryzen 9 4900H? Not all OEMs deploy top-tier cooling, especially on mid range laptops. Moreover, the Ryzen 9 4900H and the Ryzen 7 4800HS are both fundamentally the same 8-core, 16-thread Zen 2 chips, albeit binned differently.

A particularly poor cooling and power setup on the Dell could result in its Ryzen 9 4900H losing out to the lower-tier Ryzen 7 4800HS. A properly cooled AMD Ryzen 9 4900H paired with the AMD RX 5600M should be capable of delivering substantially better results. We will on the lookout for more Ryzen 9 4900H results to see how they compare."


It looks like the 5600M graphics score in TimeSpy of 6098 is actually quite capable and is in fact only 4% from matching MSI GP65 RTX 2060 graphics score (that Papusan previously claimed to be 17% higher) of 6340.
So, the 5600M does NOT seem to be an 'inferior' GPU.

Its also possible that the poor cooling that's affecting the 4900H in the DELL unit might be limiting the 5600M as well (and as the article suggests, the actual scores could be substantially higher for both the CPU and GPU than what we're seeing here if the cooling was in fact better).

It seems unlikely that DELL will improve the cooling for the unit before its released (who knows), and also, the cooling itself might be fine but the results are actually a product of badly applied thermal paste.

Dell's 'experiment' does seem to be gimped indeed a little more on the cooling side. Mind you, the machine still seems capable so lets hope the reviewers replace the thermal paste themselves with something of higher quality like GC Gelid Extreme or Kryonaut and see if that might improve these results further.

Thoughts and opinions?

EDIT: I also seem to remember that 4900H didn't score particularly well in preliminary tests compared to 4800H... however, at the time I thought it was due to the APU being an engineering sample and clocked relatively low.

People choosing this DELL machine might be better off selecting the 4800H with RX 5600M and 144HZ display (which I hope is Freesync capable) - so let's hope the reviews cover the performance of those options with better thermal paste (if that is indeed what's 'choking' the CPU [and some of GPU] performance).

 

If the performance of 5600m falls below 5500m, AMD's ATI team is really shooting their own foot!
Although a "slight" improvement over 5500m, they deserves the kudos! As 5500m age is not a year yet... Right?
Looking forward to 5700m!

 

Looks like a good laptop and pretty good cooling solution (coupled with lower fans noise).
Its sad Eluktronics didn't consider 5600M/5700M in this thing as it seems it would THRIVE under that cooling.

 

How is the performance of 5600M 'falling below 5500m'?
Those newer benchmark scores I posted today (roughly 50 mins ago) put the 5600M only 4% behind the MSI GP65 RTX 2060 (and quite possibly because of the cooling in the DELL which limits CPU performance and could be interfering with the GPU as well).

Doesn't that put the 5600M ahead of the 5500M?

EDIT: MSI Bravo 17 with 4800H and 5600M scores 12734 in Firestrike Graphics according to the following test:


From the most recent tests, we only have a TimeSpy score for the DELL unit which puts 5600M only 4% behind (properly cooled and fully boosted) RTX 2060 in a laptop.

We need the reviewers to test the DeLL with 4800H/4900H and 5600M in full and possibly replace the thermal paste with something far more decent to see if that fixes the thermal issues.

And if the unit is throttling and reviewers don't do anything to acknowledge or 'try and fix' the problem themselves, we won't know what the 4900H and 5600M can do in a 'properly cooled' laptop - as such we should not be making any 'final statements' about 5600M performance if its limited by bad cooling from the start.

 

Awww... Dude!
U missed that "if" in that beginning...

 

I haven't missed it...
I'm just saying that if those most recent TimeSpy scores are accurate, then a thermally limited laptop with 5600M is already scoring only 4% behind fully 'unhinged' and 'properly cooled' RTX 2060 in MSI.

Mind you, DELL needs to get their act together in the cooling department and stop 'experimenting' with AMD and just make a decent all AMD laptop with proper cooling.

Oh btw, according to this:
https://www.notebookcheck.net/AMD-Radeon-RX-5500M-GPU.436556.0.html

5500M in Bravo 17 (with 4800H) seems to get a Graphics score in TimeSpy of 4738 (vs 6098 for 5600M in DELL).

That implies 28.5% better performance for 5600M (which is hampered by the cooling in DELL).

And the 4800H (along with the 5500M) in Bravo 17 is not hampered by cooling

 

Looks like AMD has backtracked and will provide Zen 3 support for B450/X470 motherboards.

 

I feel this is the right choice with this being the last CPU on AM4. I really do not expect X670 boards with the Zen 3 Ryzen 4000 chips considering B550 just launched. And those that didn't feel pCIe 4.0 was worth getting the X570 boards yet did get X470 with expectations.

So I feel for the bios teams' work ahead, but this is the right move.

Sent from my SM-G975U1 using Tapatalk

 

The problem was there were so few TR x399 owners to cow tow too. AMD stomped all over us and almost no one cared. Now they failed to do it to the main stream enthusiasts.

 

I'm not a builder nor a desktop builder. But the way i look at B450, is like a "transitional vessel" to take/accept gen3 Ryzen 4000series as the last qualified passengers, but not further more... By the time 5000series comes, B550 should be stable enough to handle future Ryzens until the 6000series, and 7000series will be like the current 4000series, the last qualified passengers to hop onto the B550 vessel.

This is of course my own vision on how generations of components are bring to another stage...

 

I think you are confused. AMD has AM4 ending this year with Zen 3/Ryzen 4000. This has been known since AM4 was released. So Zen 4/Ryzen 5000 was NEVER going to be compatible with X570/B550 boards! EVER!

Instead, Zen 4 brings DDR5, but for mainstream no support for PCIe 5.0. 5.0 has spec published in 2019, and usually takes 18-24 months to make it to products, but the feature sets of it are not really needed by consumers and 4.0 is so new and hasn't fully been utilized, it would increase costs without much benefit.

I'm hoping Zen 4 Epyc supports PCIe 5.0, primarily due to Intel's CXL and cache coherence advancements, gen Z node connections, etc. But that may still be a ways off for inclusion. Don't know.

The GN video made me wonder if AMD is going to do the nightmare Intel did with the DDR3 to DDR4 switch. We'll find out.

Yeah, they learned potential lost sales with us, then applied that to mainstream, plus goodwill lost, etc. They figure HEDT will still buy the best for their needs no matter what (which they are right).

But imagine what would have happened had we had more support from the mainstream community...

Either way, I foresee two gens max moving forward. Imagine if they kept the 32 MB and used it for mem oc tests or overclocking stability test (something to at least check heat output, etc. ). Just thinking of EVGA's system on Intel chips.

Things like that would be interesting, although the CPU part is less needed with boost algorithm, but...

Sent from my SM-G975U1 using Tapatalk

 

Wouldn't it be better for AMD to continue maintaining future platform support for as long as its 'technically feasible' but say that they might not be able to guarantee support on very old motherboards if the CPU's become more advanced than predicted?

I mean, such advancement would likely necessitate a motherboard and possibly socket change, but if its using the same socket (such as AM4), then you may have to sacrifice a new feature or two instead which may not exist on older motherboards - but as we've seen with Zen 2, it seems to run with no performance losses on B350/X370 mobos (and Zen 3 would likely run the same), whereas PCI-E 4.0 is not such a big deal of a feature (yet).

If they do go down this road, it might prompt mobo OEM's to cooperate with AMD and create 'as future proof BIOS ROM chips as possible' (including laptop mobo manufacturers - though this may take more effort)... maybe chill out with new motherboard releases as well. Stick with one mobo release for a bit (I don't think we need a new mobo year after year [would also cut back on resources - maybe shuffle more into BIOS/UEFI development/upgrades/streamlining] - even if AMD maintains releasing future Zen cpu's year after year).

I realize that preparing for the future can be difficult and some unexpected changes may happen (which is understandable), however, if it is the same socket, then I'm sure we can live with losing a feature or two which don't impact the CPU performance (a lot or at all).

It does make sense to include older motherboards since they all use AM4. There are no 'crucial' features that even B350/X370 couldn't do without. And if its a question of VRM's, then why not limit support on B350 for example (and maybe X370) to CPU's with certain TDP's or just leave it to the motherboard OEM's to integrate support if they think the mobos could handle the hw in question?

 

I trust nothing moving forward as it apparently is all at AMDs whim and their current word can no longer be trusted source. I say we can only build for current generation compliance, do not count on further compliance!

Edit as a side note this is exactly what I said they should do. Make a bios only compliant with the new generation CPU. As a second note they could have left TR4 pinout maps alone (with some mod for the 64core CPU) and made it where a new bios would accept the 3960x only for x399 boards. That would have been nice, but NO!

 

It seems you mix up things.

I have never said or claimed RTX 2060 to be 17% higher in TIME SPY.
http://forum.notebookreview.com/thr...laris-navi-gpus.799348/page-764#post-11014039

Do the math.... AMD RX5600M leaked 3DM FS graphics score of 14609 in this post vs. 17119 for RTX 2060 will be +17%.

-------------------------------------------------------------

But back to Time spy... From what have seen here been posted from 5600M, I can't say they are amazing....
El cheapo Walmartbook i7-9750H + 2060

One thing for sure... The Mobile version will provide lower scores.

Old or new vBios from AMD won't change that.

 

Look, my problem with your post lies in the premise that you took a very old FS score from early January and deduced (using only those 'leaked results') that 5600M will underperform by 17% in general compared to the RTX 2060 in MSI.

You did not seem to take into account that those results could be 'early' (more to the point 'LEAKED') or could suffer thermal throttling, and that therefore could change.

I then posted a NEW article (today, from literally TODAY) which paints a different picture of RX 5600M in TimeSpy from here:
https://www.notebookcheck.net/Ryzen...-match-for-the-Intel-i9-10980HK.465925.0.html

Now, why do you think the FireStrike score from early January (if its even accurate) holds any merit?
If this new TimeSpy score of 5600M is accurate and puts it within 4% of MSI RTX 2060 in at least THAT benchmark, what makes you think the newest FireStrike scores (which we don't have at the moment) wouldn't be higher as well?

The (latest) TimeSpy benchmark seems to have a BIT more merit than an outdated FS score from 4 months ago (but again, even this new TimeSpy score is not final because we don't have an actual review of the final unit), and also, the new article mentions the possibility that DELL is suffering from poor cooling design which is hampering its overall CPU and GPU scores - which I don't see you acknowledging or mention as a possibility anywhere, but did in fact confirm that RTX 2060 varies in performance based on the chassis its in and the kind of RTX 2060 is used [something I noted as well] - so I'm wondering, why do you acknowledge these possibilities for NV but not AMD?

Furthermore, the 4900H is supposed to operate at 3.3GhZ base clocks (and yet the TimeSpy recorded 3GhZ) ... which raises the possibility that this score could be a product of an (unoptimized) CPU sample which is not performing as it should (Dell may have been in the process of testing the CPU during optimizations) and is hampering GPU performance.

So, why should we take any of these scores as final and not subject to change?

 

Wait! U mean 5600M is now available? Commercially? Awww... that'll be interesting!!! Which make will that be? I've googling these few day to see if any Chinese makers had employed this...

 

Yes, AMD experienced delays in releasing the 5600M and 5700M (that's commercially available too) due to the Coronavirus.
Both were supposed to have been out by now.

So far, only DELL is using 5600M in the G5 15 5505, and MSI is using the 5500M in Bravo 17.

I hadn't seen/heard any other laptop OEM's mentioning they will be using Navi mobile GPU's though...
We may see more laptops with 5600M and 5700M coming out later in the year though.

 

5700M is going to be available on the Alienware 51m R2 as a DGFF module. Sadly though its paired with 10th Gen Intel desktop CPU. I was really hoping Dell might have made the 51m R2 with an X570 / Zen2 variant. Maybe Clevo will later this year.

 

This is where I was scratching my head when Papusan claims that the TimeSpy benchmark was from a valid 5600m... We have not seen any commercially available G5 in the market yet. How and Where does that benchmark comes from? Desktop 5600?

 

Well; https://technosports.co.in/2020/05/...rx-5600m-graphics-spotted-on-3dmark-time-spy/

 

Would it be better compare the FS results with the one from the Walmart-machine? 17% transformed into 24% lead. Put it this way... With newer results from 5600M who may show 7% higher scores, it would still lagg behind with 17%

 

AMD "Renoir" Desktop APU Could Lack PCIe gen 4.0, Hints BIOSTAR B550 Motherboard Product Page techpowerup.com | Today

AMD's 4th generation Ryzen "Renoir" desktop APUs, based on the "Zen 2" microarchitecture, could lack PCI-Express gen 4.0, hints the product page of an upcoming AMD B550 chipset motherboard by BIOSTAR. AMD already declared that the B550 lacks support for "Picasso," which means the "Ryzen with Radeon Vega Graphics" processors referenced in the BIOSTAR product page have to be "Renoir." On the mobile platform, Ryzen 4000H and 4000U series processors do lack PCIe gen 4.0, but it was expected that AMD will enable gen 4.0 for the desktop socket AM4 platform.

The lack of gen 4.0 support has implications for "Renoir." For starters, the APU, like its predecessors, spares only 8 PCIe lanes toward PEG (PCI-Express discrete graphics, or the main x16 slot you typically use for graphics cards). If these lanes are gen 3.0, then even the newer graphics cards, such as AMD's "Navi" RX 5700 XT, or next-gen GeForce "Ampere," would have to make do with a PCI-Express 3.0 x8 interface, despite being gen 4.0 x16-capable. We will test just how much of a bottleneck this poses, when the next-gen graphics cards come out.

 

Dell G5 15 SE Laptop Review: All-AMD Gaming With SmartShift
https://hothardware.com/reviews/dell-g5-15-se-review?page=1

Solid performance from the CPU and GPU.
Battery life seems quite poor (but then again this was only tested under gaming, and battery is of smaller capacity in the DELL, whereas 4900HS is also a 35W part with probably lower power states)... but otherwise, GPU seems to be performing between RTX 2060 and RTX 2070Max-q.

It could also be that DELL didn't necessarily optimize 5600M for battery life performance, or its more power hungry than 2060 and 2070Max-Q.
The same is apparent on the MSI Bravo 17 with 5500m. You'd think that even disabling the dPGU would improve battery life, but nope... maybe the OEMs' need to work more on their AMD dGPU power optimizations when programming the vBIOS (would be interesting to see if users can modify the voltages separately though).

Temperatures seem quite good, but that screenshot of the CPU temperature being at 105 degrees Celsius seems... too high (even though the article says both the CPU and GPU hovered around 82-84 degrees C maximum)?
Could be the junction temperature... might be worthwhile for users to repaste the thermal compound and possibly try out Liquid Metal?

Also, if GPU-Z is reading those voltages accurately, WHY on earth is the GPU ramping up to 1.3V?
That's even higher than the desktop 5600XT which (with slightly higher frequencies) is maxing out at 1.2V.
Very odd.
If that's the voltage of the GPU, it could explain the low battery life. We should hope Dell or AMD allows for voltage modding on the GPU in some fashion (without potentially flashing a vBIOS).

What's even odder is that the GPU chip is recording 35W (under full load)? That seems awfully low since 5600M is rated at much more than that.

I think GPU-Z may be incorrectly reporting both the voltages (which seems too high for a mobile GPU) and GPU power draw (as I don't think it would be THAT low).

 

I think battery size plays a little prominent part on how long it lasts per use per single charge.

The battery size is obviously small among all other Renoir-H based lappies, even though it "shares" the same body built as the INTEL variant that may comes with 68whr battery than the basic 51whr battery.

https://www.dell.com/en-us/shop/laptops/new-15/spd/g-series-15-5500-laptop


https://www.dell.com/en-us/shop/del...ion-gaming-laptop/spd/g-series-15-5505-laptop
Even if one willing to pay more to the highest configuration, you won't get the highest 68whr battery as the INTEL based one do. Weird? What kind of conspiracy theory is there to make AMD based variant look so ugly? Hmmm...

Lenovo offers a minimum of 60whr battery with upgradable option to 80whr.
​

 

This is rather odd to me too since the HotHardware article stated their version came with 68whr battery - although, the HotHardware video on the DELL 5505 mentioned that in idle tasks etc, the Ryzen/Navi had some of the longest run-times... nearly 8 hours if I'm not mistaken (but if this laptop comes with 51Whr and not 68Whr, then how come the HotHardware got the bigger capacity battery and is it possible they misread the battery capacity?).

Adding on top of that no apparent SATA port for an internal HDD at all. I thought the laptop not coming with HDD meant DELL decided to simply not include a HDD, but it looks like they went all the way and removed the SATA port entirely - unless the SATA port exists and is hidden somewhere (that battery looks awfully huge for 51Whr capacity in the teardown pic, but the hinges which hold the battery into place may be obstructing the SATA port (hiding it) on the far left of the teardown pic - the indent looks barely enough for a portable hdd).

I suppose that even if the hdd bay doesn't exist on this laptop, one could always use a portable drive... but its not exactly ideal since high capacity SSD M.2's are quite a bit more expensive than a 2.5" hdd and having everything inside the actual laptop is a bit more economical.

As for the design of the laptop... I'm not overtly bothered by that.
What's more important for me is the hw inside along with the cooling and overall quality...
You can have a great looking laptop which thermally throttles and is nowhere near capable of maintaining advertised performance (such as the Razer Blade series).

But to be fair, there's nothing stopping OEM's from designing a great looking all AMD laptop.

Dell is 'experimenting' with AMD... and I guess we are lucky to get even that, but we shouldn't really accept whatever OEM's throw at us either.

 

I couldn't find any high resolution picture of the teardown of G5 5500 15inch model so far. So as the G5SE 5505.
However, i do find a chinese site teardown video of the G5 INTEL 5500
https://www.bilibili.com/s/video/BV1754y1D795/
...where the commentator claims that the internal is nearly identical to G3 3590. Battery shown here is clearly 51whr, occupies about 2/3rd of the width of the interior.

 

Eluktronics has managed to increase the battery size from 46.7 to 62 WHr for both RP-15 and RP-17 models. It was announced yesterday and the spec sheet is updated. No news on this from XMG yet.

Battery: 4-Cell 62.36WHr Polymer Battery Now Included!

 

Yeah, and in the AMD version, the teardown pic is clearly showing 68Whr written on the battery (which begs the question, why is the DELL website not offering that option?).
Anyway, I checked out the teardown of the Intel version, and the HDD can fit in it, but in the AMD version, there doesn't seem to be enough space... there's an SSD in the way if I'm not mistaken.
Either that, or someone who gets the machine can have a look if there's an extra SATA cable/slot that could integrate a HDD underneath that battery on the lower left.

Quick question... would a 68Whr battery be that large?

 

Well... For evaluation purpose, especially gaining attentions from social medias, i would not be surprise that a "better spec'd" sample be sent out for reviews. What do they stand to gain, no clear idea. But making AMD device looks bad when it hit the market... Consider that as a fulfillment of promise by taking incentives from others...

As for the size of battery, is it that large, yeah, looking at that 51whr battery size, 68whr battery width sure kicks the 2.5" drive (HDD/SSD) off.

 

That's a bit weird, since the Zephyrus G15 comes with 76Whr battery, and there's room for at least 2x M.2 ssd's (which might be the case here too, but I don't see the second M.2 slot.

Its possible Dell might add the 68Whr battery option to their website as it really doesn't make too much sense for the review unit to come with 68Whr battery only for the rest of the units to not have the option (and if I'm not mistaken, the technical specs did mention 68Whr battery is an option with AMD).

 

The 68 wHr battery for the G5 SE is listed in its Specifications file, so it'll very likely be an option to choose in the future.

 

yes... It is listed in the specification files for sure. Because it shares the same build and structures of it's elder blue blooded kin.
But... Does it get the fair treatment of healthy innard after birth since both comes from same family?

 

The review unit for Hothardware got the 68Whr battery, so the possibility is there that others will too.

 

Base on what i come across in Dell's service manual,
https://www.dell.com/support/manual...8de7b7-879f-45a4-88e0-732155904029&lang=en-us
... there seems to be 2 system builds that could affect storage installation options as well as battery size options.
If you choose to hv a 2.5" storage bay, there won't be a 2nd M2 slot for further upgrades...

...the battery size will shrunk as well to 51whr.
If you opt for 2 M2 installation system, there will be a M2 slot soldered in replacement to the connection for 2.5" storage(HDD/SSD)...

...the battery size will be beefed up to 68whr.
Notice the difference between the 2? Here's a better focus...
​

 

Hm... looks like slots for 2x. M.2 SSD's (or NVme's).
That isn't too bad obviously... but again, HDD's are still cheaper than M.2. SSD's

 

Look at the 51whr configuration closely, in the picture, u'll see a white ribbon connector next to the wifi card, which i assume for 2.5" drives installation.
In the 61whr configuration picture, u'll see that the ribbon connector replaced by a NVMe connector.

I guess we'll hv to wait for the first review from first few buyer who placed the order for one for self upgrade...
:htttp://forum.notebookreview.com/threads/waiting-for-news-on-the-dell-amd-g5-se.832576

 

@Ed. Yang I would not trust those photos from the service manual. If you look at the one of the motherboard removal it looks like a Nvidia GPU and Intel CPU.

 

I have to agree.
The teardown pic which has 51Whr battery looks like its from a unit with Intel/NV hw.

As far as I know, the teardown pic of Ryzen/Navi version of DELL never showcased the 51Whr battery (unless someone has a different source pic)?

 

As much as i would like to shun away from the DELL Service Manual material as reference, knowing that there will be tendencies that makers(not only DELL) will tend to "recycle" pictures from other models that may 90% close in designs and layouts of another model... For the moment, as of today, there seems to be no "explicit" view on the internals from different reviews as well as video clips.
Hence, DELL's Service Manual shall be my reference material for time being.
Since DELL clearly specifies 68whr battery upgrade is available, why it's not available as an option now in the 3 markets, namely U.S, Canada, as well as Australia? While the INTEL variant have that? Fair Game? Hmmm...

Can i hv that link to that system you saw? It's really interesting to learn about what's to be expected and what's not to...

 

https://hothardware.com/news/amd-ryzen-4000-zen-3-cpu-tsmc-5nm-wild-rumor

If true...

Sent from my SM-G975U1 using Tapatalk

 

I read fro that mainline 5nm will be end of 2021. There is no way they will wait or that on the 4000 series.

 

Rumor | AMD's upcoming Zen 3 Ryzen 4000 desktop CPUs may actually use TSMC's 5 nm enhanced production process, still scheduled to launch in Q4 2020
https://www.notebookcheck.net/AMD-s...-scheduled-to-launch-in-Q4-2020.466975.0.html

Just a rumour.

To be fair, AMD maintained that Zen 3 will be released on 'enhanced 7nm' (although it was initially supposed to be 7nm+ EUV... now its just 'enhanced').

The only thing that may give credibility to this 5nm rumor is that yields on 5nm were apparently exceptionally high... achieving over 50% last year... which apparently went up to 80% if Anandtech is accurate at all:

https://www.tomshardware.com/uk/new...-zen-4-reportedly-already-at-50-percent-yield
https://www.anandtech.com/show/15219/early-tsmc-5nm-test-chip-yields-80-hvm-coming-in-h1-2020

I'm confused.
Some articles (a larger portion of them) seem to indicate 50% yields, whereas Anandtech reports 80%.

Yields are dependent on die sizes... the larger the die, the lower the yields... so what's the real picture here?