Slickdeals is community-supported.  We may get paid by brands for deals, including promoted items.
Missed This Deal? Get alerts when similar deals are posted.
expired
Suryasis 112K Views
Aug 20, 2021 4:09 PM

Minisforum EliteMini HX90 PC: Ryzen 95900HX, 2xDDR4 SODIMM, 2x 2.5" SATA Bay

$649

$750

13% off
+ Free Shippingat Minisforum
Visit Minisforum

Why are you downvoting this deal?

Good Deal?
215Comments
Save
Share
Deal Details
Minisforum.com has Minisforum EliteMini HX90 Mini PC Barebone Pre-Order on sale for $649. Shipping is free.

Thanks community member Suryasis for sharing this deal

Specs:
  • Ryzen 9 5900HX 3.3 GHz (8C/16T, 4.6 GHz Boost, 20MB Cache)
  • 2 x SODIMM DDR4 Slot (64GB Max)
  • 1x M.2 2280 PCIe Gen 3 Slot
  • 2x 2.5" SATA 3.0 6 Gbps Bay
  • 1x M.2 2230 with Intel AX200 Wi-Fi 6 + BT 5.0
  • Ports:
    • 5x USB 3.2 Gen 1 Type-A
    • 1x USB 3.2 Gen 1 Type-C
    • 2x HDMI 2.0 (4k@60Hz)
    • 2x DislayPort (4k@60Hz)
    • 2x Line Out
    • 1x RJ45 Gigabit Ethernet
    • 1x Clear CMOS
Includes:
  • Power Adapter DC 19V / 6.3A
  • 1x HDMI Cable
  • 1x USB Cable
  • 1x Mounting Bracket
  • 1x Computer Stand

Editor's Notes

Written by slickdewmaster | Staff
  • Deal Expert Suryasis recommends that you watch Review Video posted below for deciding whether to go with various Ram + Solid State Drive option.
  • Review: Head on over to TechTablet for their In-Depth review of this PC.
  • Warranty: Includes 12-Month Warranty which starts at date of purchase (details).
  • Minisforum Return Policy: Undamaged products may be returned within 7-Days (details). -slickdewmaster

Original Post

Written by Suryasis
Community Notes
About the Poster
Deal Details
Community Notes
About the Poster
Minisforum.com has Minisforum EliteMini HX90 Mini PC Barebone Pre-Order on sale for $649. Shipping is free.

Thanks community member Suryasis for sharing this deal

Specs:
  • Ryzen 9 5900HX 3.3 GHz (8C/16T, 4.6 GHz Boost, 20MB Cache)
  • 2 x SODIMM DDR4 Slot (64GB Max)
  • 1x M.2 2280 PCIe Gen 3 Slot
  • 2x 2.5" SATA 3.0 6 Gbps Bay
  • 1x M.2 2230 with Intel AX200 Wi-Fi 6 + BT 5.0
  • Ports:
    • 5x USB 3.2 Gen 1 Type-A
    • 1x USB 3.2 Gen 1 Type-C
    • 2x HDMI 2.0 (4k@60Hz)
    • 2x DislayPort (4k@60Hz)
    • 2x Line Out
    • 1x RJ45 Gigabit Ethernet
    • 1x Clear CMOS
Includes:
  • Power Adapter DC 19V / 6.3A
  • 1x HDMI Cable
  • 1x USB Cable
  • 1x Mounting Bracket
  • 1x Computer Stand

Editor's Notes

Written by slickdewmaster | Staff
  • Deal Expert Suryasis recommends that you watch Review Video posted below for deciding whether to go with various Ram + Solid State Drive option.
  • Review: Head on over to TechTablet for their In-Depth review of this PC.
  • Warranty: Includes 12-Month Warranty which starts at date of purchase (details).
  • Minisforum Return Policy: Undamaged products may be returned within 7-Days (details). -slickdewmaster

Original Post

Written by Suryasis

Community Voting

Deal Score
+58
Good Deal
Bad Deal

Why are you downvoting this deal?

Top Comments

Join The Conversation

Share your experience with the Slickdeals community

Share information with the community. Please follow our Community Guidelines and be kind!

215 Comments

Sign up for a Slickdeals account to remove this ad.

Aug 30, 2021 9:05 PM
637 Comments
Joined Sep 2011
duc135Aug 30, 2021 9:05 PM
637 Comments
Quote from ChrisP8282 View quoted comment :
Not the person you're responding too but I had a spreadsheet from a client that would take 6 hours to run one iteration on an i7 9700k with 32GB of RAM in 64 bit Office running on a Samsung 960 NVME 500 GB drive. I had to write a loader that would cycle through each iteration, copy the results out and then move to the next iteration. Took an entire weekend of running to finish the process with only ~30 of 200 desired nodes completed (we took the 30 most populous nodes as a sample.)

It was a million rows by about 25 columns wide with each column utilizing a previous set of columns to generate an output. I quoted them a week to convert it to SQL Server but they needed it faster than they had time to even stand up a SQL Server much less refactor the model.
You are obviously a moron like the rest of us. According to the poster you quoted, any PC processor built in the last two decades can easily perform any Excel function in 30s or less no matter how much data is in the spreadsheet. Haven't you seen the synthetic benchmark test results? Get out of here with your "real world" use cases. Those mean nothing.

Oh, also.

1. Did you check to see your commit charge numbers in Task Manager? Most likely you're out of memory and it has nothing to do with your CPU.
2. Did you check CPU utilization in Task Manager?
3. Pics or it didn't happen.
4. The root of the issue is that Excel is not written with special CUDA instructions to take advantage of GPUs with high refresh rates. Did the PC running the Excel spreadsheet have a powerful GPU with high refresh rates? If so, you need to open a ticket to Microsoft to have them rewrite Excel to take advantage of the CUDA instruction sets and then get the design engineers at AMD to design their new GPUs to also includes the NVIDIA CUDA instructions so that Excel will run in the exact same amount of time as you have experienced because Excel doesn't even run in the GPU.
Aug 30, 2021 9:17 PM
819 Comments
Joined Jan 2013
ChrisP8282Aug 30, 2021 9:17 PM
819 Comments
[QUOTE=MaroonPanther3033;149380105]You're really barking up the wrong tree here; I've forgotten more about IT systems than you will ever learn. As the former 1st VP of Branch Infrastructure for the largest mortgage underwriter in the world, and as the former Director of Midrange Systems for the market data company that processes 90%+ of all market transactions in North America, not to mention my past contracts as an intelligence contractor for the DoD, I can safely say that I've worked with not only some of the largest spreadsheets in the nation but indeed the world. Elitebooks are pretty crummy laptops and indeed loaded with slow processors and other subpar quality components, as are Inspirons (although the Inspiron line has MANY different models, with some being close to workstation powerhouses). That being said, if they are locking up for ANY reason, it isn't because the files are too large UNLESS you're overwhelming either your addressable memory space OR your system or application page pool, which can be confirmed by checking if you see any event 2020 [or 2021] errors in the system logs. The application may appear to be locked up while processing data but if the system locks up then it is for another reason and NOT because your processor is slow.



What is the commit charge? If you don't know then you are making unsubstantiated remarks. The fact that you refer to commit charge as being out of memory already tells me that you aren't qualified to make this statement because they two are related but not the same. Also, you haven't a clue what it means to be "out of memory" because you don't understand how the system kernel allocates those resources. If the system is performing correctly when loading huge files, you should see your RAM being maxed out up to whatever percentage the kernel decides should be reserved for other application or system processes. You say you've 32GB of RAM but that really means NOTHING. If you have 32GB, Excel might allocate (depending on its configuration and that of the kernel's page pool allocation [and NO, the system paged pool and non-paged pool is NOT the same thing as paged RAM, even if they both use the word "page"] for example 28GB for Excel to use, cut up into the non-paged memory pool which then purges data from the pool when more is added. Typically for examples such as yours you would set it to 0xFFFFFFFF in the registry to overcome depleted memory pools. Again, you can check for event ID 2020 or 2021 in the log files to see if this is becoming extreme enough to cause system instability issues.

If you're experiencing extended delays when working with spreadsheets that are that "massive, I pontificate that you are accessing them over a local network, most likely from a typical DFS server or cluster. That network traffic? Yeah, that's what's slowing the process down. It might not even be a bandwidth issue; it's probably something related to your TCP window size and the well-documented issues with TCP autotuning for SMB 2.0+ systems. That issue is often further exacerbated by accessing large databases or other datasets (such as large spreadsheets) over multiple segments and your Network engineering team should work on cross-router window sizes and traffic prioritization.


Again, you're desperately clinging to this fantasy that this system is good for working on large spreadsheets. Perhaps, but that doesn't change the fact that this system is generally an off-balance system that serves little to no purpose except for marketing to people like you who don't really have a firm grasp of the technology you employ. Yet again you're trying to argue a specific example versus a general idea. It's the debate method those with little to offer in terms of actual knowledge typically resort to.


So you're a grunt; I got it. That further substantiates my assertion that YOU really haven't a clue what you're talking about and are in fact just some bean counter without nothing more than a very basic understanding of systems.



That's true, but mathematical operations are actually FASTER than sorting in almost every case. CISC architecture, which is what Core-iX processors employ includes many integrated instruction sets in the microcode to accelerate complicated mathematical expressions.


It has a CPU installed that is way out of balance with the rest of the system. With the exception of an INCREDIBLY RARE set of examples, this system would in fact be SLOWER than some systems that strike a balance between its components. Anyone worth their salt as a systems integrator, administrator, or solution architect will tell you exactly the same thing, although I doubt that you would listen, as you're the type of wannabe who likes to delude himself into thinking he's right, even when what he's saying is wrong.


LOL, NO. Again you demonstrate just how ignorant of the subject matter you actually are. "Refresh rate" refers to the speed in Megahertz that a component operates at; in this case the CPU. A fast GPU contributes nearly nothing to Excel as there is no CUDA acceleration employed the software... YET. I doubt it ever will, either since Microsoft wants to move to a subscription-based service that leverages cloud-computing. Excel 2019 is believed to be the last standalone version of the software that will ever get released, and even in that case you can't use the software without Windows 10 unless you go through an elaborate hacking process to get it working on previous versions of Windows.


Nobody is arguing the convenience of using spreadsheets versus databases. You're just confusing the conversation in an effort to look less of a fool. I stated that this system is not particularly well-engineered as configured. The number of people that would find this a better option over similarly-priced alternatives is INCREDIBLY small. You and your buddy's crying about Microsoft Excel is not only unfounded but frankly even if it was accurate, it is a moot point because no enterprise company would deploy an unsupported no-name system like this regardless.

Quote from Meribela_S View quoted comment :
8. Who the Fck said anything about using this system as an enterprise work machine? Are you fcking retarded.
Are you suggesting that SOHOs would be using massive Excel spreadsheets with complicated formulas and reporting? Hmmm


I've already established that I have worked with many large spreadsheets in enterprise corporate environments, not to mention had many direct reports over my career that are bean counter grunts just like you.


My first comment was that this system is pretty odd and won't be particularly useful for the vast majority of end users. That comment is accurate. You only consider my comments absurd in that they dumpster anything of the opinions you've given thus far... using facts.


I know you're tired from being buried repeatedly but if you don't like hearing TRUTH then YOU are the one that needs to leave. Also, Neither Jesus nor God would have anything to do with you. You are not a godly person, and are in fact bigoted in your adamant yet [mostly] untrue statements, fascist in that you don't want to hear that you're wrong so you demand people to be silent, and evil in that you exhibit these evil qualities while simultaneously using both Jesus' name and God's formal address in vain.

*mic drop*
What the heck does "out of balance" even mean?
1
Aug 30, 2021 9:19 PM
819 Comments
Joined Jan 2013
ChrisP8282Aug 30, 2021 9:19 PM
819 Comments
Quote from duc135 View quoted comment :
You are obviously a moron like the rest of us. According to the poster you quoted, any PC processor built in the last two decades can easily perform any Excel function in 30s or less no matter how much data is in the spreadsheet. Haven't you seen the synthetic benchmark test results? Get out of here with your "real world" use cases. Those mean nothing.

Oh, also.

1. Did you check to see your commit charge numbers in Task Manager? Most likely you're out of memory and it has nothing to do with your CPU.
2. Did you check CPU utilization in Task Manager?
3. Pics or it didn't happen.
4. The root of the issue is that Excel is not written with special CUDA instructions to take advantage of GPUs with high refresh rates. Did the PC running the Excel spreadsheet have a powerful GPU with high refresh rates? If so, you need to open a ticket to Microsoft to have them rewrite Excel to take advantage of the CUDA instruction sets and then get the design engineers at AMD to design their new GPUs to also includes the NVIDIA CUDA instructions so that Excel will run in the exact same amount of time as you have experienced because Excel doesn't even run in the GPU.
This is such a bizarre thread. Even if what he was saying was true (it's not) sometimes I just want more muscle to throw a spreadsheet at instead of refactoring it.
Aug 30, 2021 9:48 PM
637 Comments
Joined Sep 2011
duc135Aug 30, 2021 9:48 PM
637 Comments
Quote from ChrisP8282 View quoted comment :
This is such a bizarre thread. Even if what he was saying was true (it's not) sometimes I just want more muscle to throw a spreadsheet at instead of refactoring it.
How can what he said not be true? Please remember that he was a "1st VP of Branch Infrastructure for the largest mortgage underwriter in the world... the former Director of Midrange Systems for the market data company that processes 90%+ of all market transactions in North America" and has had "past contracts as an intelligence contractor for the DoD". But above all else, he has seen the synthetic, benchmark test results that show Excel calculations can be completed in 30s or less.

But in all seriousness, I'm right there with you. When people spend a large portion of their workday analyzing/manipulating large amounts of data, any additional help in reducing wait times equates to more productivity. But try telling that to the bean counters who control the budget and only source the cheapest parts that can barely support current environments, let alone future resource loads on the hardware. I work in a government regulated business and the amount of background processes to log/audit and enforce security policies is insane.
Aug 31, 2021 11:57 AM
222 Comments
Joined Apr 2009
moro310Aug 31, 2021 11:57 AM
222 Comments
So I cancelled my preorder; just FYI they charge a 2% 'handling fee' to cancel, which is a pain
Aug 31, 2021 2:42 PM
2.7K Comments
Joined Jul 2016
Meribela_SAug 31, 2021 2:42 PM
2.7K Comments
Quote from duc135 View quoted comment :
You are obviously a moron like the rest of us. According to the poster you quoted, any PC processor built in the last two decades can easily perform any Excel function in 30s or less no matter how much data is in the spreadsheet. Haven't you seen the synthetic benchmark test results? Get out of here with your "real world" use cases. Those mean nothing.

Oh, also.

1. Did you check to see your commit charge numbers in Task Manager? Most likely you're out of memory and it has nothing to do with your CPU.
2. Did you check CPU utilization in Task Manager?
3. Pics or it didn't happen.
4. The root of the issue is that Excel is not written with special CUDA instructions to take advantage of GPUs with high refresh rates. Did the PC running the Excel spreadsheet have a powerful GPU with high refresh rates? If so, you need to open a ticket to Microsoft to have them rewrite Excel to take advantage of the CUDA instruction sets and then get the design engineers at AMD to design their new GPUs to also includes the NVIDIA CUDA instructions so that Excel will run in the exact same amount of time as you have experienced because Excel doesn't even run in the GPU.
Beautiful assessment and response! You get a rep for it Smilie
Aug 31, 2021 3:01 PM
2.7K Comments
Joined Jul 2016
Meribela_SAug 31, 2021 3:01 PM
2.7K Comments
Quote from ChrisP8282 View quoted comment :
Not the person you're responding too but I had a spreadsheet from a client that would take 6 hours to run one iteration on an i7 9700k with 32GB of RAM in 64 bit Office running on a Samsung 960 NVME 500 GB drive. I had to write a loader that would cycle through each iteration, copy the results out and then move to the next iteration. Took an entire weekend of running to finish the process with only ~30 of 200 desired nodes completed (we took the 30 most populous nodes as a sample.)

It was a million rows by about 25 columns wide with each column utilizing a previous set of columns to generate an output. I quoted them a week to convert it to SQL Server but they needed it faster than they had time to even stand up a SQL Server much less refactor the model.
Our data report was so huge that I had to limit them to around 150MB or so because the company laptops would take too long and crash. Like your report we had millions of rows BUT around 50 columns of data fields to collect. People who has never worked on such large data don't know lol. A simple vlookup can take hours to complete depending on the power of the laptop.

The thing is this amount of data is peanuts to the shear amount of data that I've collected when modeling wind turbines...100+ channels collecting data every second in real-time.

Such BS statements like 30s can only be said by an ignorant person who clearly has no such experience and rightfully should be ignore.

Sign up for a Slickdeals account to remove this ad.

Aug 31, 2021 5:10 PM
637 Comments
Joined Sep 2011
duc135Aug 31, 2021 5:10 PM
637 Comments
Quote from CoolDolphin604 View quoted comment :
Use is for a couple of senior citizens that pay their bills online, and occasionally watch youtube. I just want something snappy for them with the latest version of windows.
Here's one that is currently on sale.

https://www.amazon.com/Beelink-Br...B08R78QWWR
Sep 9, 2021 5:41 PM
186 Comments
Joined Dec 2008
traveraSep 9, 2021 5:41 PM
186 Comments
Anyone has any idea when this will be shipped? It was said mid Sep but not sure if there will be any delays.
Sep 10, 2021 5:11 PM
11 Comments
Joined Nov 2015
thehoodedoneSep 10, 2021 5:11 PM
11 Comments
Quote from travera View quoted comment :
Anyone has any idea when this will be shipped? It was said mid Sep but not sure if there will be any delays.
I just checked and this is what the website says:
Orders before 8.25th are in transit.
Orders from 8.26th~8.31th will be shipped at End-Sep.
Orders after 9.1st will be shipped at End-Oct.
Sep 10, 2021 6:38 PM
186 Comments
Joined Dec 2008
traveraSep 10, 2021 6:38 PM
186 Comments
Quote from thehoodedone View quoted comment :
I just checked and this is what the website says:
Orders before 8.25th are in transit.
Orders from 8.26th~8.31th will be shipped at End-Sep.
Orders after 9.1st will be shipped at End-Oct.
Oh shoot, I ordered on 8/27...
Sep 11, 2021 7:53 PM
18 Comments
Joined Nov 2020
astromasterSep 11, 2021 7:53 PM
18 Comments
https://www.youtube.com/watch?v=V...amersNexus
Comically bad product. Gamers Nexus review says liquid metal was splattered all over the internal component. And reviews on the site are 1.9 stars. Would stay far away from this.
1
Sep 12, 2021 7:36 PM
11 Comments
Joined Dec 2018
Jcanales1Sep 12, 2021 7:36 PM
11 Comments
Sep 13, 2021 1:00 AM
186 Comments
Joined Dec 2008
traveraSep 13, 2021 1:00 AM
186 Comments
Quote from astromaster View quoted comment :
https://www.youtube.com/watch?v=V...amersNexus
Comically bad product. Gamers Nexus review says liquid metal was splattered all over the internal component. And reviews on the site are 1.9 stars. Would stay far away from this.
Omg, this is scary, can somebody received the unit confirm?

Can I cancel the order?

Sign up for a Slickdeals account to remove this ad.

Sep 13, 2021 2:31 PM
186 Comments
Joined Dec 2008
traveraSep 13, 2021 2:31 PM
186 Comments
reply from the company
------------------------------------
Hello,


Recently, some Youtube reviewers have questions regarding our product and the way we market it.
First of all, we would like to apologize for the confusion that we cause to reviewers and our customers and would like to make some explanation regarding the questions that you may have:

1.Liquid metal splitting problem:
We tested our product in several ways and one of them is a drop test. HX90 passed the 60cm test.
But later when we did a 80cm test, we found a problem. Liquid metal leaked.
We want to improve it and raise the standard so the HX90 can survive an 80cm drop.
But because we had limited samples, we removed the liquid metal and used thermal paste in these review samples.

2. Is the product liquid metal applied?
The actual product is definitely liquid metal applied. There will be no problem with sealing. HX90 can survive 80cm drop tests.

In mid August, we were informed that the sealing technique was good enough and we can produce samples with liquid metal applied so we decided to send the second unit to reviewers because it would be better if they can test the sample which can represent the actual product.
The reason why reviewers got confused is that there were misunderstandings in our company's internal communication.
Our factory mistakenly sent the same sample as the first unit which was still using thermal paste.
There is no reason in the world that we would send a false product to reviewers to show our problems, especially reviewers like Linus Tech Tips, Jayztwocents and Gamers Nexus.
As soon as we found out we sent the wrong sample we wrote emails to all reviewers who received the second unit to try to clarify the mistake.
But damage had been caused and we are truly sorry for the time they spent testing the sample.

3. The inconvenient screws:
This is another difference from the actual product. The actual product will use Philip Screw.
As shown in the picture, this was designed before August.

We sincerely apologize for the confusion we caused to reviewers and customers. We will never lie to our customers and our products will be the same as the way we market them. We will upload a video filming how we apply liquid metal.


But regardless, please accept our humble apologies.


Thanks & Regards

The MINISFORUM TEAM


BESSTAR TECH LIMITED

Tel: (00852)30506835

Address: 19H Maxgrand Plaza, No.3 Tai Yau Street, San Po Kong, Kowloon Hong Kong

Official website | Official store | Official forum

Follow MINISFORUM on Social

Facebook:MinisforumClub|Twitter:@Hi_MINISFORUM
Last edited by travera September 13, 2021 at 08:02 AM.

Join The Conversation

Share your experience with the Slickdeals community

Share information with the community. Please follow our Community Guidelines and be kind!