A few weeks ago, a friend asked me what the problem might be when all browsers stop working. More specifically: Whenever they tried to launch either Microsoft Edge or Mozilla Firefox, they only got an error message saying they didn’t have sufficient permissions. It was a strange error – why would permissions suddenly be missing after years of trouble-free operation?

After some back-and-forth, we managed to install a new version of Mozilla Firefox with step-by-step phone instructions and finally installed AnyDesk using a working browser, which allowed me to take a look at the problem myself. That’s when a surprise came to light: The PC has been running Windows 8.1 Home, 64-bit, just as it always has!

Having gone without any updates for quite some time (end of life was on January 10, 2023!), the problem ultimately came down to expired root certificates. Continuing to run 8.1 wasn’t a viable option anyway, so I performed an in-place upgrade to Windows 10 22H2 remotely via AnyDesk. That took hours because the factory-installed 3.5″ WD hard drive with 1 TB capacity* is still the boot drive. The already very underpowered fourth-generation dual-core i3 is further severely hampered by the now very limited 4 GB of DDR3 RAM in the form of a single memory module. All in all, a significantly outdated system, but one with the potential to be upgraded very cost-effectively. That’s exactly what I did today, a few days later.

After a deep clean, everything looks so much brighter:

Upgrades

The PC is used solely for viewing photos and occasional web browsing. So there’s no reason to switch to the very latest and most expensive hardware. For a reasonable price, you can still get a lot out of this 2014 Acer Aspire, here’s a comparison of the current components and the planned upgrade:

The UEFI was already up to date; I checked that before taking anything apart. The reason is that newer processors within the same generation are usually only supported by updated microcode in the form of a newer BIOS or UEFI version. The planned i7-4790 is actually a refresh within Haswell-DT, so it’s even newer than the original i7 release models and incompatibilities cannot be ruled out.

Here are the components; everything except the CPU is from my own stash:

Processor replacement

This is incredibly easy thanks to the FC-LGA 1150 socket:

For thermal paste, I used the tried-and-true MX-4 from Arctic* and 99.9% isopropyl alcohol* to clean the components. I added the two new RAM modules, and lo and behold: POST was successful, and it’s running. Everything is recognized correctly in the UEFI. Now I just need to shrink the partitions, clone the hard drive to the SSD, and put everything back together.

Success?

For comparison purposes, I ran Cinebench R23 and the PassMark Performance Test v11.1 on the original hardware at the beginning and then again after the upgrade. Both tests were run on the same Windows 10 Home 64-bit system. This makes it easy to see the performance gains:

The performance gains are consistently very substantial and extremely noticeable. In particular, the integrated graphics unit – which is slightly more powerful in the i7 than in the i3 – benefits further from dual-channel operation with two memory modules (instead of just one as before). However, due to the patches addressing Spectre and Meltdown as well as other factors, the results do not quite match the reference scores. Here are the Cinebench screenshots:

Next, the PassMark results – though, due to a lack of internet connection at the time of testing, these are only available offline:

Last but not least, here is the HWiNFO overview of the respective configuration:

Der Beitrag Acer Aspire XC-605: A Late Upgrade in Times of Skyrocketing Storage Prices erschien zuerst auf flohs blog.

Once the basics are clear, RAM must be procured. ECC would be possible, but it is significantly more expensive and harder to find than normal, unbuffered RAM. So “normal” RAM is used.

The maximum expansion with four memory channels and eight slots would be eight DDR4 modules, each with a capacity of 32 GB. That would give a whopping 256 GB. Desirable, but not financially feasible at the time of my search.

Corsair Vengeance LPX

Because price/performance also plays a role, we opted for DDR4 modules with 16 GB each*, giving us a total of 128 GB for all eight. This means that the four memory channels are fully utilized and the memory controller is not overloaded.

⧉ Corsair

⧉ Corsair

⧉ Corsair

Power supply

I was able to purchase two HX1200s* at a bargain price from a miner in Bavaria. The second one, which has not been used until now, will now be used in this server (the other HX1200 is in “Tamy-PC”). Full accessories, very good condition.

⧉ Corsair

⧉ Corsair

⧉ Corsair

⧉ Corsair

⧉ Corsair

⧉ Corsair

Der Beitrag <h5>TR-Server #3: </h5><h3><b>Memory and PSU</b></h3> erschien zuerst auf flohs blog.

After high-capacity hard drives and subsequently SSDs and RAM of all kinds had risen sharply in price due to extreme demand from AI companies, this unexpected news came at the beginning of December:

SSDs und DRAM: Crucial ist am Ende

Anyone who has been involved in the hardware sector for some time knows that Micron and Crucial, in particular with their extremely popular (but unfortunately long since discontinued) MX500 SATA SSD, were a real force to be reckoned with and often the only real alternative to Samsung. It is a real shame that the consumer brand “Crucial” is now leaving the stage. Competition stimulates business, and now there is one less supplier.

This will only fuel the rapid rise in prices even more. On December 14, 2025, there was another piece of news that has since been denied, but which I find even more worrying:

Samsung stoppt Produktion von SATA-SSDs: Leaker warnt vor bis zu 18 Monaten Preisdruck bei SSDs

Samsung has stated that these rumors are false. We will see whether this is true or whether Samsung is just trying to control the narrative in the first half of 2026 at the earliest. Since Samsung has always been my number one choice in the SATA sector, it would be tragic if production were to end. There are also no usable NVMe to SATA adapters that could be used to “re-adapt” the problem.

To make another component that has been totally overpriced for a long time even more unaffordable, the following happened the day before yesterday:

GeForce-Grafikkarten: Nvidia will GPU-Produktion angeblich massiv senken

If this proves to be true, the demand for high-performance DRAM will also be the reason here. The “super” refresh models that were actually expected are also not going to be released. These are all unconfirmed rumors, of course.

Future prospects

There are basically two possibilities:

Either all AI companies prove to be highly profitable, as many investors hope, and prices remain extremely high or even worsen. Then the era of forced subscription-based cloud PCs, which many have been predicting for some time, will draw ever closer—running your own powerful hardware will simply be too expensive.

Or much of what is currently being hyped will prove to be an unprofitable business. This would result in falling prices and a richly filled second-hand market – which would be desirable.

It is very difficult to predict how all this will develop, especially since the global economic situation has been anything but stable for some time now.

Der Beitrag Thoughts on the current situation with data storage devices and RAM erschien zuerst auf flohs blog.

As a 32-bit operating system, Windows XP can only handle a maximum of 4 GB of RAM. So far, so clear – but, similar to the RetroBooks, the PC should ultimately be quad-bootable, and everything after XP can address significantly more RAM in the 64-bit version. Besides, 4 GB of RAM with quad channel seems a bit strange – four 1 GB sticks, or what?

The maximum configuration without ECC on this board is 8 DDR3 sticks with 8 GB capacity each – 64 GB in total*. Pretty impressive for a 12-year-old system.

Fast RAM, matching the CPU rarity

At the time of the search, RAM was very cheap, especially the “old” standards such as DDR, DDR2, and DDR3. So I ended up buying two 32 GB kits (CMY32GX3M4A2400C11) on eBay, each consisting of four brand new Corsair Vengeance Pro modules with 2400 MHz, CL11 Volt XMP profile in red, which came directly from China. I suspect overproduction or stock items, as no discrepancies could be found upon closer inspection.

I chose red so that everything would match the motherboard, which is also red and black. The red component is made of aluminum and can be easily removed if necessary.

⧉ Corsair

The exact configuration of these memory modules is:

• DDR3-1333 as JEDEC basis with 1.5 volts
• XMP v1.3 on DDR3-2400 with 1.65 volts
• Clock rates with XMP then 11 – 13 – 13 – 31 CR2

I paid around €140 for all eight bars, but now the same kits cost more than double that at the same retailer. I was lucky.

Der Beitrag <h5>Retro-XP-PC #4: </h5><h3><b>RAM?</b></h3> erschien zuerst auf flohs blog.

As with the previous processors, the documents from HP provide some assistance.

The original service and maintenance manual for both models can also be found here:
https://www.levnapc.cz/ProductsFiles/hp-elitebook-8540p-8540w-manual-en.pdf (archive.org)

On page 60, there is an overview of the memory modules that are supported:

It’s a bit strange that the smaller modules are listed as 1600 MHz, which, to my knowledge, is not supported by any of the processors that run on the motherboard/chipset. It doesn’t make any difference; faster RAM is simply throttled down. I suspect that this made it easier for HP in terms of hardware inventory management. 1600 MHz was standard for DDR3 in many devices at that time and especially in the years that followed.

The maximum RAM for the models would be as follows:

Digression: RAM designs

To briefly explain, as there is often little differentiation between them, there are generally three types of DDR RAM designs, which apply to DDR, DDR2, DDR3, DDR4, and DDR5. This will likely also apply to their successors.

Design	Description	Application	Modular / interchangeable?	Types
DIMM	Dual Inline Memory Module	Desktop, Server	Yes	UDIMM, RDIMM, NV-DIMM, etc.
SO-DIMM	Small Outline – Dual Inline Memory Module	Mobile	Yes	UDIMM, RDIMM
LP	Low Power	Ultra-Mobile, Smartphones, Tablets, etc.	No	unbuffered or registered possible

More detailed information here: DIMM

Provided that the RAM has not been soldered (which is unfortunately increasingly the case with modern devices), ONLY SO-DIMM modules fit into a notebook; in the case of the 8540p, this is the DDR3 variant.

If the notebook is to be used exclusively for Windows XP, a total of 4 GB RAM is optimal, preferably in the form of two modules with a capacity of 2 GB each (dual channel operation). As a 32-bit operating system, XP cannot address more than 4 GiB, regardless of how much memory is installed.

In our case, I am utilizing the maximum amount of RAM that fits into an 8540p motherboard, namely two DDR3L SO-DIMM modules, each with 8 GB, 1600 MHz* from the manufacturer SK Hynix, which adds up to a total of 16 GB.
After all, I’m planning a multi-boot setup, more on that later.

By choosing an i7-840QM as my CPU, I can achieve clock speeds of up to 1333 MHz – so my memory modules are throttled, but fortunately in exchange for slightly better latencies:

It is important to note that DDR SDRAM (DDR stands for Double Data Rate) specifies twice the clock rate, as data is transferred on both the rising and falling edges of each clock cycle.

However, the screenshot above shows the “real” clock rates – 666.7 is actually 666.6667 period and corresponds to 1333.3334 period MHz clock. This therefore matches the maximum capacity of the CPU (“1333 MHz”). I’m glad that these memory modules seem to have these latencies etc. stored in the SPD chip ex works and that the appropriate values for the clock are applied automatically (probably by the BIOS). Lucky me.

There are no settings for this in the BIOS of the 8540p, so it’s all the better that everything is detected and set fully automatically. Cheap no-name modules will probably not have these functionalities, or at least not completely, so be careful!

In my case, I had another, slightly newer notebook that was irreparably defective (I destroyed it myself…PROCHOT Override, burned out – another story), from which four of these bars were “left over”—ideal, two for each of the two 8540ps. I had previously purchased eight of these bars at auction for a good price (at the time) – according to the description at the time, they had been removed from Apple Mac Minis.

SK Hynix, Samsung, Micron (Crucial), and Kingston are all reputable brands; I usually tend to go with Samsung.
There are many model names, but not many memory manufacturers, so it’s best to do your research carefully.

Digression: Types of DDR3 SO-DIMMs

It is important to mention that there are two widely used versions of DDR3 as SO-DIMM in the mobile sector and one less common version:

I have never seen the “U” variant in the wild, but DDR3L is very common in devices that came onto the market later in the DDR3 life cycle. It is also backward compatible with “normal” DDR3 and can therefore be used as a direct replacement for DDR3 in most cases (as is the case with the 8540p).

But be careful:

The early CPUs of the first generation Core i, and especially those before that, i.e. Core 2 and earlier, are in exceptional cases very sensitive to RAM, which, like the processors themselves, is known to consist of silicon dies manufactured using different lithography processes.
Old processors can sometimes only handle the “older,” “coarser” RAM modules. This problem was particularly pronounced in the Core 2 Duo/Core 2 Quad era, when the basic structure of a motherboard/chipset was still somewhat different.

This also has something to do with the fact that these older processors usually had to cope with two generations of RAM (e.g., DDR2 or DDR3) and therefore simply seem to be more sensitive.

Finally, on the subject of RAM, it should be noted that there are general differences in the configuration of the individual modules, but these are not relevant in this upgrade project, as the memory controller is located in the CPU and, except for overclockable Intel processors, determines the maximum clock speed (1333 MHz for the i7-840QM, which is fixed).

For example, there are SS and DS modules, i.e., single-sided and double-sided. There are also ranks that do not necessarily follow the configuration: single rank, dual rank, and quad rank.
The whole thing is definitely a science in itself, just as RAM overclocking is a complex topic.

We continue with the most important of all upgrades, provided that a hard disk drive (HDD) was still installed.

Der Beitrag <h5>RetroBook 8540p #4: </h5><h3><b>Upgrade – RAM</b></h3> erschien zuerst auf flohs blog.

The following procedure has proven to work well, regardless of the newly built system:
First, check whether there are any BIOS/UEFI updates available from the motherboard manufacturer, in my case ASUS—there was a newer version (v1501). I installed this using a FAT32-formatted USB stick. Even if there had been no newer version, continue with:

Set everything in UEFI to factory settings and then to minimum settings, i.e.:

• RAM at minimum clock speed, i.e., simply leave the existing setting (presumably 2133 MHz for DDR4) unchanged.
• Processor settings all “original”
• Prepare Windows—disable CSM/Legacy, enable TPM, boot order, enable Secure Boot if desired, etc.

First, install a fresh copy of Windows, if you don’t already have one, to rule out software problems later on and avoid unnecessary background processes and legacy issues that could interfere with overclocking. I installed Windows 11 Pro 24H2, bypassing the “incompatibilities.”

Required software

Once Windows is installed, you can let the system install the drivers itself via a working Internet connection—that’s the easiest way. Then install the updates offered and restart the whole thing. Next, check in Device Manager to see if drivers have been installed for everything relevant—usually they have.

However, I always take the “old-school” approach:
Before I even install Windows, I download the latest drivers for the components I am using from the manufacturer’s websites (in this case, ASUS, Nvidia, and Intel). Then I install Windows offline and, once that is complete, I install the drivers I downloaded earlier, also offline.

This has the advantage of usually resulting in newer driver versions than those offered by Windows Update. Sometimes the manufacturer’s websites also contain additional information, such as notes on current or past problems, etc., which you would never see if you used the automatic method.

I used the following software to overclock the CPU and GPU:

• HWiNFO – For monitoring
• Prime95 – For load generation, optionally with or without AVX / AVX2
• Cinebench R23 – For generating faults and testing differences
• ASUS ROG RealBench – To generate load in order to test stability
• Unigine Heaven – To generate load on the GPU in order to test stability
• MSI Afterburner – For overclocking the GPU
• MemTest86 – To generate faults on the RAM and memory controller in order to test stability

My goal is to achieve stability in ASUS ROG RealBench, as I believe it reflects moderate to heavy everyday use equally well.

Processor overclocking

Actually, the whole thing isn’t complicated, but you need to have a lot of time and patience.

I first overclocked the i7-7700K statically, ensured stability, and repeatedly pushed it to its limits (voltage, temperatures). Since my goal was 5.00 GHz on all four cores with hyperthreading enabled, I started with that—the delidding I had done beforehand should have made this feasible in terms of temperature.

So I simply set the core voltage to 1.30 volts in the BIOS and set the “CPU Core Ratio” for “All Cores” to “50”. Then I saved, booted Windows, and started HWiNFO / Prime95 with AVX and AVX2 and let it run.

First observation: It runs stably for at least 1-2 hours, but gets quite warm (around 80-85 °C).
Then back into the UEFI and down with the voltage, now to 1.25 volts – then tested again.
Temperatures have dropped to just under 80 °C, still everything stable.

I did this until I reached 1.18 volts, at which point I had my first “quick” blue screen.
So I increased it again to 1.22 volts. It then ran for many hours, stable with acceptable temperatures (still ~ 80 °C). Without delidding, even that would have been either barely possible or not possible at all; I’m at the limit.

Next, I ensured stability with ROG RealBench, which is a realistic everyday measure for me; it has to be stable in the end. It was, it runs smoothly and, of course, cooler; I saw a maximum of 75 °C on the hottest core.

Great, I would be satisfied with that in principle.
But now, of course, I want to know what will limit me in the end:
The cooling or the voltage?

How far will it go?

Then I dared to increase the clock speed, setting the multiplier to “51,” i.e., 5.10 GHz. I also adjusted the voltage to 1.30 volts to match. Lo and behold: Prime95 with AVX quickly forces the cooler to its knees thermally, the first thermal throttling kicks in, and later the first blue screen appears (well over 100 °C!).

So I continued with 1.28 volts. Same result, thermal throttling—but later. No blue screen.
But you can already see the limit with AVX, 5.00–5.10 GHz. Then the air cooler, which otherwise does a really good job and is also tolerably quiet, reaches its limit.

After only a few minutes of ROG RealBench, a blue screen appears.
The journey with AVX seems to be over here, from now on I’ll optimize the system for “stable everyday use with a little AVX.”

And I’ve still come a long way:

The final result:

First of all: I made it very easy for myself with the RAM. I simply loaded the stored XMP profile for 2400 MHz 16-16-16-39 at 1.20 volts, slightly increased the DRAM voltage to 1.2540 volts, and set the memory frequency to 3000 MHz, nothing else!

The whole thing ran for days with Memtest86 without any errors. That was a 25% performance boost for free.
Above 3000 MHz, errors occurred that would have required very time-consuming fine-tuning to fix, but that wasn’t my focus.

For the processor, I switched from static overclocking at 5.20 GHz with over 1.35 volts to adaptive overclocking. 5.30 GHz on all cores got too hot – it took a few minutes, but throttling kicked in.

In the end, of course, I spent hours experimenting and days testing to see what worked best and how close I could get to the limits.

The CPU now runs as follows:
Load on 1–2 cores = 5.30 GHz at LLC 6 with adaptive auto voltage, slightly above 1.42 volts in HWiNFO under load.
Load on 3–4 cores = 5.20 GHz at LLC 6, also adaptive, then 1.385 volts under load.

Cache = 4.90 GHz, above that it quickly became very unstable.

Without AVX, I am therefore at both the temperature and voltage limits; it would not be possible to go any further without entering even more dangerous voltage regions. I find 1.42 volts to be borderline; my limit was actually 1.40 volts.

I couldn’t use an AVX offset without the turbo and power-saving mechanisms only working halfway afterwards, or the single-core clock speeds never being reached, as Windows 11 seems to use AVX instructions quite often. I wanted to achieve a dynamic overclock that is also capable of downclocking in idle mode to save heat and power – hence the use of an adaptive “auto” offset in the first place.

Conclusion:

This inexpensive cooler packs a punch – without AVX, I can push the CPU to almost its maximum capacity. At the same time, the board’s power supply is so good that I have never experienced any problems in this regard (I didn’t have any additional airflow over the VRM coolers!).

I would describe the clock speeds as very high, achieved at the expense of equally high voltage. Here is a comparison of the changes:

Cinebench and PassMark results:

First, Cinebench R23 with single-core overview – 1423 points.
Then with multi-core overview – 7112 points.

Finally, Cinebench R24 – 86 points or 419 in multi-core.

When compared to other CPU models, it is easy to see how far the i7-7700K has risen, especially in terms of single-core performance. Temperatures were consistently around 90 °C while Cinebench was running.

Click here for the PassMark PerformanceTest v11.1 Baseline.

What about the graphics output?
We’ll continue with that in the next post.

Der Beitrag <h5>i7-7700K OC #7: </h5><h3><b>Now the overclocking can finally begin</b></h3> erschien zuerst auf flohs blog.

The following components have already been defined:

• Processor, Intel Core i7-7700K
• Processor cooler, Thermalright Peerless Assassin 120 SE
• Motherboard, ASUS ROG Strix Z270E Gaming

The following parts are still missing:

• RAM
• Possibly a graphics card with HDMI 2.0
• SSD / HDD
• Power supply
• Case, but since the whole thing will be overclocked on a benchtable and later built into a drawer, this is negligible.

RAM:

For RAM, I used a 16 GB kit consisting of two 8 GB modules – Crucial Ballistix (DDR4-2400)*. I already had them on hand, left over from a previous upgrade project:

⧉ Crucial

Graphics card:

As a dedicated GPU, even though Intel’s integrated GPU would have also enabled 4K @ 60 Hz via DisplayPort 1.4, I used an existing and unused Nvidia GeForce GT 1030 OC* from Gigabyte. Of course, in terms of connectivity (PCIe 3.0 x4!), it’s not a particularly good card, but it can do one thing better than the integrated one: 4K @ 60 Hz via HDMI 2.0. As an additional positive side effect, more RAM is available for the system as a whole, which would otherwise have been used by the integrated GPU.

⧉ Gigabyte

⧉ Gigabyte

⧉ Gigabyte

SSD / HDD:

I plan to install two data storage devices—a very fast NVMe boot drive, also a leftover Samsung 960 Evo with 250 GB of storage space (my first NVMe, newer model*), and a 3.5″ SATA hard drive with 1 TB of storage space* as a slow data grave for downloads, cache, etc.

There are SSDs with significantly more performance – I consider the read rates for everyday use of this system to be by far the most relevant, and even the smallest 960 Evo with 250 GB at the time can shine here, with a maximum of 3200 MB/s. The fastest PCIe 3.0 x4 Samsung SSD in the consumer sector, the 970 Evo Plus (the successor to the 970 Evo), only manages marginally more, namely a maximum of 3500 MB/s – which is the maximum for the interface anyway.

The 960 Evo is therefore a good fit for this system because even it almost maxes out the interface when it comes to important read operations. The NVMe also gets a passive heat sink from JEYI – I’ve installed it many times before – it works well, looks good, and is very affordable:

⧉ Samsung

⧉ JEYI

⧉ WD

Power supply:

Like everything else… left over. I usually use power supplies from Corsair*, and in exceptional cases from EVGA.
Here’s a picture:

⧉ CoolerMaster

Now all the components are together—assemble them and start overclocking!

Der Beitrag <h5>i7-7700K OC #6: </h5><h3><b>Remaining hardware</b></h3> erschien zuerst auf flohs blog.