In December of 2020, a group of friends came together to form Chips & Cheese. It started out as a simple outlet for our in-depth investigations into hardware but quickly turned into a much bigger project altogether.
In 2021, we jokingly said that getting 10,000 visitors and 20,000 views that year would be ambitious. That goal was shattered with a total of 56,649 visitors and 97,084 views. The following year of 2022, we had an optimistic goal of 100,000 visitors and around 250,000 views. Once again those goals were completely and utterly decimated with a total of 242,267 visitors and 430,004 views. In 2023, this trend continued with a grand total of 531,682 visitors and 911,369 views in that year. Well beyond any expectations we could have had. We can’t thank everyone enough for their support.
This growth in the past three and a half years means we have decided to make some changes.
The Changes
The biggest change that has happened is that Chips and Cheese (Formally Chips and Cheese Technology Publications) is now a 509(a)(2) with 501(c)(3) status. Or in plain English, we are a tax-exempt non-profit charity which means that any donations that you send us are tax deductible.
Forming a nonprofit organization entails a lengthy and tedious process consisting of a lot of paperwork, a lengthy waiting period, and the formation of a Board of Directors to act as a guiding force for the non-profit in order to fulfill its mission statement.
The specific purposes for which this corporation is formed are to educate the public on technical topics, to build and maintain a set of in-house tools to test and evaluate technological products, to release a useful set of these tools as Open-Source Software to inspire curiosity in our audience.
Mission Statement of Chips and Cheese
As for the members of the board:
- George J. Cozma: President and Chairman
- Dr. Ian Cutress: Vice President
- Ryan Mull: Treasurer
- Tristiano Agostino: Secretary
- Chester Lam: Board Member
- Joshua Gregory: Board Member
- Zachary Monchamp: Board Member
- Felix LeClair: Advisor to the Board
- Philipp Munkes: Advisor to the Board
- Usman Liaqat: Advisor to the Board
In pursuit of that mission statement, there have been several projects that we have been working on in the background that we would like to bring y’all up to speed on.
A Call to Action
The first project we have been working on is a new microbenchmark framework. This new framework will hopefully allow for standardization between different tests to keep things as consistent as possible. In the short term, this will also allow folks other than the Chips and Cheese team to add to the Chips and Cheese test suite.
In the long term, we hope that this framework will allow for more tests to be written than the current Chips and Cheese team could ever write on its own, along with diversifying test authors. The vast majority of our current benchmarks were written by Clam and having a single person writing the majority of our tests is not sustainable long term. We hope that this will let Clam to move into sort of an Editor-in-Chief of the Chips and Cheese benchmarks as we grow and expand.
Elaborating further, our goal is to provide more tools for the community, developed by the community. Previously, the tools we provided on our Github were almost entirely code written by Clam. While these tools are not going away, they were written for Clam own direct use, not for the community. To that end, we are aiming for this new suite to be focused on community engagement. We are aiming for a 2 tier model: A framework that will standardize and abstract some of the common elements of the tests being written, as well as a curated set of tests distributed directly by us. Once the framework is released, we encourage the community to design and build their own tests. Additionally, we are actively seeking out contributors to help us build out the Chips and Cheese Github with more in-house code.
In a similar vein, Clam is also writing most of the articles on Chips and Cheese and that is also not sustainable long term. We hope that we can get people who are willing to write articles of high quality, if not the same length, as what we currently publish.
To that effect, this State of the Union will also double as a Call to Action. We are looking for new authors, code contributors, and other enthusiast community members to help shoulder the workload that keeps Chips and Cheese running smoothly. Anybody is welcome, so don’t feel discouraged. Our goal is to diversify the contributions going in, so that the community is as involved as they can be. The more of us that help, the stronger Chips and Cheese can become.
Feedback Needed
We’ve been considering moving over to Substack and want to know how readers feel about this before we make a decision. We want to be clear that a move to Substack would not involve pay-walling any of our current content output or future articles. There are upsides and downsides, here’s how it looks from our perspective:
Upsides:
- Integrated email newsletter makes it simpler and easier for folks to be notified of and read new articles
- Easier linking between our own articles and articles from other Substack outlets (e.g. More Than Moore by Dr. Ian Cutress)
- Better discoverability so that more folks can read our articles and learn from them
- Clean, responsive, cross-platform reading experience (a significant improvement over the current layout, but see downsides)
- Increased security versus current WordPress setup
- Decreased monthly cost compared to the current WordPress + Cloudflare setup
- More ways for readers to contribute beyond PayPal and Patreon (which are not available in some countries)
Downsides:
- Much less control over site design/layout and platform technical details
- Substack operates on a percentage-of-revenue fee model so may work out to be significantly more expensive in the future
- No automated external backups of content (an externally-hosted service is required to achieve this)
- No support for tables; would have to work around this with image or SVG tables
There are several alternative options that we are also exploring which include a significant redesign of the site to improve the user experience across different devices as well as looking into new solutions for improving our current newsletter service.
Minecraft Testing Suite
Alongside the new microbenchmark framework we have been working on, we have been also working on a new Minecraft testing suite created by TitanicFreak.
Being our most requested title for us to benchmark while also being very customizable and expandable, Minecraft is the perfect fit for our benchmarking purposes.
We will be targeting the server software side in particular because that’s where most of the heavy lifting in Minecraft is located. Even in single player, you are still playing on an internal server which has to simulate the world 20 times a second.
We do have some overall goals of this benchmark suite, which are:
- Create a wide array of tests targeting different parts of the game, two examples are
- Time it takes to generate a world from scratch to a 2,500 block radius world
- The milliseconds per tick (MSPT) it takes the server to simulate a late game base (any higher than 50 MSPT results in some degree of ‘lag’)
- Make all tests scalable to show off architecture and cache differences better
- Every test will scale until the server takes 1000 milliseconds per tick (MSPT) to simulate everything or the CPU is completely saturated across all threads. This will show us the capabilities of a particular CPU while also showing how much faster or slower relatively it is with smaller test sizes.
- Make the benchmark suite portable across Windows, macOS, and popular Linux distributions.
- Bake in expandability from the start so that we cover popular Minecraft modding APIs such as Paper, Fabric, and Forge in the future.
- Provide ample documentation to allow for easy setup and configuration
- For various important reasons we will not be redistributing the Minecraft server software and any code that isn’t our own. We will have links to acquire all of the software you will need to get setup however. Hence the big focus on documentation.
All of this will take substantial time for us to create. But as we complete individual tests we will have articles discussing it so everyone can follow along the development process to completion.
Since we are Chips and Cheese, we are going include some preliminary data on chunk generation speed comparing the AMD Ryzen 9 7950X and Intel Core i9 14900K. The data was collected with the following configuration:
- Stock power limits for both Intel and AMD CPUs
- 14900K: 253W PL1 / 253W PL2 / 307a Iccmax
- 7950X: 170W TDP / 230W PPT
- The following memory
- 4x32GB of ECC DDR5 running @ 3600MT/s JEDEC
- The following versions
- Java: Temurin-17.0.11+9
- Paper: 1.20.4-b446
- Chunky: 1.3.146
- Debian 12.5 (kernel 6.1.0-20)
- The following configuration
- Aikar’s Flags for the bootup JVM flags
- Seed: 503
- Radius: 1000 block radius from coordinates X: 0 Z: 0 with the square shape
Thanks to Paper having a scalable and configurable chunk generation system that far exceeds the capabilities of the vanilla system, we are able to see how the different architectures scale when you give more the server software more cores to work with. In this case it shows that the heterogenous approach of the 14900K’s core configuration has put it at no disadvantage compared to the 7950X. In the end the 14900K is actually faster with a peak generation speed of 509 chunks per second vs 474 chunks per second on the 7950X.
This is just one example of what we are trying to accomplish with our Minecraft Benchmark Suite and we can’t wait to show off more in the future. But it’s time to wrap this up now.
Conclusion
There have been 1,008,797 visitors and 1,760,828 views as of the time writing this article.
All we can say is wow. When we started this site we could never imagine getting over 25,000 visitors and 50,000 views let alone over a million each. That is all thanks to the help of you, our wonderful readers, and the valuable members of the Chips and Cheese community. Our goal with all of these changes has been to set up Chips and Cheese for a stable, long term future, so that it can continue to enrich the tech community with deep architectural insight, analysis, and data.
Expect the articles to continue uninterrupted, but with new names, faces, and topics as we engage and bring in our community. Thank you for the continued support, and back to our regularly scheduled…Chips and Cheese.
It’s definitely a bad idea to move to a platform like Substack when you already have a successful independent website. There’s much more to lose than there would be to gain.
definitely
Please fire whoever suggested Substack
I would caution you against moving completely to substack if you have already built a successful site.
I have seen enough predatory behaviour from various large user-generated content sites that I would be very wary. They can change anything at any time and you are forced to live with their bad decisions because changing sites would mean losing a majority of readers.
They can change anything at any time and you are forced to live with their bad decisions because changing sites would mean losing a majority of readers.
“Better discoverability” can be a double-edged sword. I think it is not unlikely that in the future the number of readers on the site will be largely determined algorithmically. For example, one day Youtube stopped sending alerts for all your subscriptions. Now it is usually algorithmically determined unless you specifically change it. Look at the titles and thumbnails of almost every successful channel. They are forced to use clickbait or lose a lot of viewers.
I wouldn’t call substack a “Clean, responsive, cross-platform reading experience” either. The site is technically sub-par. The first thing I see when I go to substack.com is layout problems:
https://i.postimg.cc/vH7LjYVd/pic.png
The English version is better, but it also has an element that is cut off. The bigger problem is that normal posts are laggy on my phone. If a website cannot display text or simple images without significant frame rate drops, then the technical basis must be an absolute mess.
I’ve learned a lot reading Chips and Cheese articles. Many thanks to all those who have contributed to writing them!
I find the current website model works well for me. My experiences with substack have been largely negative.
Firstly, congrats on the new organization! Second, thanks for the insights and in-depth analyses, always much appreciated. Third, I agree with Mark and Chri – your website is working well, at least from my perspective as a reader. Not sure if moving to Substack would be an improvement, and there’s risk it’d be a downgrade.
Like the others, I think the website is good as is. It is on its way to becoming what Anandtech used to be, a great source for deep dives and more technical information rather than just benchmarks.
Congratulations on the new setup, and thanks for all the great work you’ve done so far!
I’ll continue to pay regardless of where you are, but as others have mentioned, your website is working well, and it’d be a shame to have yet another independent site move to a “platform”. It’d be a downgrade in many ways – a monoculture of personality being one of them, platform dictated workflows being another. I’ve found myself using content on Substack lesser and lesser.
I’m very much against moving to Substack, partly because I disagree with the views of its owners, but mostly because your site is much more usable for me than theirs is, and has been ever since I’ve known about you. If you do need to move to a better platform than WordPress, there are lots of options that lets you retain (or even improve) control, instead of relinquishing it.
This is one of relatively few sites i visit regularly and work perfectly accross half a dozen devices, and even more screens. The only thing that occasionally gives me any trouble is trying to scroll past those double-graphs with the slider, on mobile devices.
And even then, it’s the minorest of issues. I just need to get better at scrolling on touchscreens.
If there were any features that could use improving… Maybe a slightly better search feaure? Other than that, i immensely enjoy how easy it is to read this website. There is no BS above, below or around the info i’m looking for.
I’ve been coming to C&C since early 2021, and a paying supporter since late 2021. Seeing it grow into what it is now is amazing! Congratulations to everyone involved.
I’ll keep supporting regardless of the possible future changes. With that said:
(1) Please don’t go to substack. I am slightly bummed by the political discourse regarding substack, but I guess every larger company will have problems like this. My issue is mostly that it is another large publisher that will try to shove other content down my throat at some point and integrate C&C into that algorithmic pool (I was supporting a creator there at some point and it took me quite some time to get rid of all the “suggested creators” and “call to action” emails). I understand that it is likely nostalgia speaking, but C&C being independent does give it a certain vibe.
(2) If you want to revamp the existing website, may I suggest tapping into the existing community of IT-adjacent professionals that you have already built? You are proposing to build the future of C&C on open source. Why not do the same with the website? Define some general guidelines regarding where the website should be headed in the future, and start an open call for designs, or make a repository where people can suggest updates to the WP theme.
(3) Finally, if you want to become more “mainstream” (which the proposed move to substack suggests to me), I think the content will need to become more appealing to more junior or non-technical people. For example, I teach software development at a university and I’m not even sure how many of our students would understand (or care about) most of the articles that appear on C&C. I’m not saying this is a demographic that you necessarily have to go after, just that maybe some series like “intro to CPU architecture”, “intro to benchmarking”, “intro to supercomputers” would be very helpful to a lot of young people and could bring in more mainstream/junior people to C&C.
Hey, please don’t move to substack. We love the site as-is. I don’t want to see one of the best technical sites around to fall to the “subscription based, blog like service actually built to serve annoying popups”.
I find myself instinctively not clicking articles hosted on substack. If it’s really enticing, I try to read it with reader mode, but that’s very rare.
I wouldn’t say that Substack’s design is any better than the current one, and viewing/typing this on a smartphone now there is nothing in my eyes that is lacking from the experience.
Do people even use email to learn what articles got published? I’ve been reading dozens of newsletters and websites for 20 years and RSS and my own dang memory have proven good enough for me. I personally despise the obnoxious “Give us your data!” popups on SS.
Congratulations on your well-deserved success. I’m excited by the move to open benchmarks, and the idea that everyone could test and add data to your database. I’d definitely love seeing a database with information on a variety of chips.
Regarding changing the platform, I don’t have a firm opinion about that, but I do think that the current website is too simple. It’s fine for just reading the occasional article (and you don’t publish that many currently, so it’s fine), but it’s not easy to reference past content, and considering the in-depth technical nature, referencing past content is a thing that will be useful.
Some sorting by topic would be good for a start: desktop GPUs, mobile GPUs, desktop CPUs, mobile CPUs. As I mentioned above, a database, or list by devices (or architectures), would be very nice.
In terms of expansion of content, I’m not that much of a fan of the idea. I have no real interest in Minecraft, and what I really want is just more deep dives into GPUs and CPUs that provide information not available elsewhere, with a focus on the newest architectures. For example, the RDNA 4 and 3.5 changes articles based on LLVM were really to my liking, but also the “Correction on Qualcomm iGPUs”, because it included information about GPU structure that’s otherwise not available.
I wish you a lot of good luck, and keep up the good work.
Congratulations on the success! I had no idea so many read this stuff, I would’ve estimated your audience more in the thousands 🙂
I also wanted to second what most commenters have been saying about substack. In particular, I think the issue of “enshittification” is overlooked. I also very much like your current site which just gives me your content with no added BS. And even if one trusts the current substack owners&management, these things change. It’s a lot easier to walk into the golden cage than to break out of it.
That being said, I also receive a few substack newsletters.
Congratulations for the success! Keep it upwards!
As most people say, please don’t move to Substack.
One suggestion for content. Some entrance level post or references to courses/books to get the most of your work.