Western Digital products are very popular not only among retail consumers and corporate clients, but also among modders. And today, truly unusual and interesting material awaits you: especially for Habr, we prepared an interview with the founder and head of the Tech MNEV (formerly Techbeard) team, specializing in creating custom cases for PCs, Sergey Mnev.
Hello, Sergey! Let's start the conversation a little further. There is a joke: “How to become a programmer? Study to be a philologist, doctor or lawyer. Start programming. Congratulations! Are you a programmer". Hence the question: what are you by education and profession? Was it originally a "technician" or a "humanist"?
The joke is completely true. I have two higher educations: "socio-cultural service and tourism" and "clinical psychology". At the same time, at one time I worked first in a private computer service in Bratsk, then, when I moved to Krasnoyarsk, I got a job in a company specializing in servicing the IT infrastructure of corporate clients. So I am a self-taught IT specialist and I think that this is absolutely normal. It seems to me that it is not crusts that speak about the professional qualities of a person, but practical skills.
Tell us more about your team. By the way, which is correct: Techbeard or Tech MNEV? How did your passion for modding start?
Initially, the project was called Techbeard (that is, "Technical Beard" - I think it's clear why), but recently I decided to rename it, so now we are everywhere listed as Tech MNEV. Our history began with the site Overclockers.ru. I liked everything related to the computer world, then the topic of modding caught on, I started writing , and off we go. There I also met a very talented 3D engineer Anton Osipov, and we started doing common projects.
By the way, why does Anton prefer to stay in the background? Where is the video of him? What are you hiding from us?
Everything is simple here. Firstly, Anton is a highly demanded specialist and he is sorely lacking time. And secondly, to be honest, he is not very good at acting as a presenter (in terms of the experiment, we tried to record several videos, but it didn’t work out very well), and he doesn’t like to shine in public.
Is modding for your team just a hobby or is there a commercial component as well?
To be honest, at one time we had plans to launch our own product line. They started small: they began to produce their own frames for mounting video cards and even sold them at one time. The next step should have been water cooling systems for the CPU, but here we are faced with the harsh truth of life. We went to state institutions, which, in theory, should help small businesses, but did not receive assistance as such. They tried to find partners in the face of manufacturing enterprises, but they put up some crazy price tags even for test samples. In total, it took a year and a half to “go through the throes” - and all to no avail. Unfortunately, Russia is not the country where you can build a business of this kind. What is the result? The developments have not gone away, and we still would like to implement them, but at this stage it is impossible simply because no one is tritely interested - neither investors nor consumers.
Ok, I understand that it is rather difficult to attract investments into such a project, but modding (albeit it is difficult to call it a mass sector) seems to be quite popular, if you look at the audience of the same Overclockers.ru and other specialized portals. And the videos on your YouTube channel still collect several thousand views. Why not the target audience?
Yes and no. The problem with modding is that the personal computer is more of a consumer topic than, for example, cars. In principle, the PC is utilitarian, with it you will not go out into the street to show off in front of others, there is no kind of party here, like the same street racers. The computer is first and foremost for yourself. The mass user either does not need this at all (he is only interested in performance, silence, compactness), or there are enough RGB fans on the front panel. And those who are in the subject usually make custom builds for themselves. That is, readers of Overclockers or viewers of our channel do not convert into customers: they come for inspiration and exchange of experience.
Well, it is futile to launch in Russia, there are not as many potential customers as it might seem at first glance. But here a logical question arises: what if we enter the international arena? Try to establish production through the same China, look for investors in Europe?
We are currently considering launching a crowdfunding campaign on Kickstarter. We have a new hull concept and a test sample will be ready soon. I can’t reveal all the cards yet, I can only say that this will be a completely different look at the PC case, what it should be and what it should do.
In general, for ourselves, we decided this: we do not want to do cheap things. We want to create truly well-thought-out cases made of high-quality metal (3-4 mm aluminum AMg6), with powder coating, thoughtful cooling, and convenient layout. But at the same time, we want to create custom cases that could become a full-fledged decor item. We began to treat modding as an art form, no matter how pathetic it may sound. Now this is all in its infancy, but who knows, maybe in the future we will become some kind of IT artists.
Here you are talking about Kickstarter and a new project. I think among the readers of Habr there are many who want to support you. Where can all this be tracked?
Tech MNEV Main Representative Offices − и . There is also a group on the VKontakte network, but I practically don’t deal with it, so all the news appears on the “pipe” and on instagram.
Listen, does modding itself bring any income?
Modding brings fabulous ... costs. Taking into account the time, materials, production of test models, some improvements, we always go into the red, since creating a custom case is not a cheap pleasure, to put it mildly. In order not to be unfounded: the budget of two Zeniths was 75 thousand rubles, 120 thousand went to the String Theory project, and 40 thousand to Assassin.
Hmm, to be honest, I figured it paid off somehow.
Ultimately, no. Well, of course, some projects are sponsored by component manufacturers, in some cases the same components are used several times (for example, iron from Apex later came in handy in the implementation of three other projects), and something is sold. But the end result is always a loss. Modding is not a plus, modding is a minus, it is a very costly hobby that does not generate income.
But, perhaps, the publication on Habré will fix this! When this material is released, it will be read by thousands of people. Surely someone will be interested in what you are doing and will write to you in Direct: they say, so and so, you are so cool, make me a cool assembly. Will you take on such a private order?
In fact, our subscribers have already written to us with similar proposals. We are completely open to cooperation and are always happy to work on an interesting project, but there is a nuance. It's one thing when a person comes to us and says: "Guys, here is such and such a budget, I need such and such a PC to be beautiful, functional, practical." There are no problems here: we make a 3D model, coordinate the details, and start production. Again, as an option, you can order something from us based on existing developments - we will also do it.
But very often we are approached with orders in the style of “I want this, I don’t know what”. We do not undertake such work in principle. I'll explain why. Designing a hull from scratch takes at least 3 days. I mean 72 hours of net working time. And it’s not yet a fact that the first time you get a thing suitable for further implementation: we, for example, have about a dozen dead projects that have not even gone into metal, since even at the initial stage it became obvious that they were not viable. And if the customer does not have a clear vision of what he wants to receive, then in principle we will not come to anything good. If in the middle of the work it begins “what if you do this, and what if you remove this, and what if you add it here”, then this project can be considered obviously unpromising: you can communicate for a month, six months, a year - and still do nothing.
Project Zenith: Threadripper and 8 NVMe SSD WD Black RAID
We talked about the team, it's time to go directly to the hero of the occasion - the Zenit project. How did it start and how did the idea of creating such a building come about?
I will not prevaricate: I am a longtime friend of Asus. More precisely, I am in very good relations with the people who work there (it all started again with the Overclockers portal and the overclocker party). How good? Well, I can call them and say: “Guys, you have a cool mother coming out soon. Can I take it for review? And they will send it to me, no problem at all. Actually, that's how I got the ASUS ROG Zenith Extreme Alpha X399 - by the way, the first in Russia. And as you can easily guess by the name, it was Asus products that inspired the Zenith project.
In general, a rather interesting story came out with this case. As I said, on average, it takes us 72 hours of pure time to design. However, I drew a sketch of Zenith on paper in just three hours: the day before the release, they sent me photos of the motherboard, and I was so inspired by this product that I immediately gave out the concept. As a result, the first version of the hull was built in just two weeks. But the second one took almost a year, but here the whole snag was polishing and finishing some details, which turned out to be quite laborious, since we set ourselves the goal of making a full-fledged, viable product out of Zenit.
Great! Ok, the Asus motherboard served as the basis and source of inspiration. How were the other components selected?
We tried to work with a variety of companies (I won’t talk with anyone, so that there would be no black PR), we wrote off the same Overclockers with someone, we got in touch with someone directly. And very often we got nothing but empty promises. It is not failures, but unfulfilled promises. That is, it was exactly like this: it seemed that everything was agreed on, it seemed that they told you: “OK, it’s not a question, we’ll do it, we’ll give it, we’ll send it.” And silence. A month or two later, no result. Considering how much time and effort goes into each project, these situations do not go unnoticed. Therefore, we fundamentally do not cooperate with such companies, since now we have partners with whom we can conduct business adequately.
And if we talk about the choice between Intel and AMD ... I myself am not a supporter of the "blue" or "red" camp, these are completely different sides, both are very interesting, each has its own chips. You just need to understand why you need this or that hardware, what tasks are supposed to be solved on it, and then everything falls into place. I think this is the most correct approach. It's kind of strange to choose this or that platform based on fan feelings, especially since everyone has their shortcomings. For example, if we talk about the RAID from the WD Black SSD, which we made at Zenith, then Threadripper was the perfect fit here. However, I still have a very specific complaint about AMD: this technology is far from the end user. Yes, a groping person will do everything without any problems, but it will be difficult for a simple user without basic knowledge, although I think a fast RAID array of solid state drives would be very useful to everyone who works with content. In the end, such people are not required to understand computers, and it would be cool if AMD would simplify this point: you need RAID, you installed the program, launched it, and enjoy it.
Here you are talking about the fact that it was difficult to interact with many companies. What happened to Western Digital?
In terms of work, everything turned out to be very simple: I got in touch with them, told them about the project, offered to implement it - and implemented it. No expectations and games of silence, as often happens. Why WD? It can be said that this is an old love, from the time when I worked in a service center in Bratsk. It so happened that if the hard drive is, then it must be WD, and there have never been any special problems with these hard drives. Here's another point: thanks to the experience of PC service, I know very well the main sores of HDDs from different vendors. Almost every company at one time or another had frankly unsuccessful products or devices had weaknesses. Here, Western Digital did not have such noticeable jambs in principle. For comparison: the client has a low-quality power supply, the voltage jumps at 12 volts. If there is a screw from WD, then it has a maximum of SMART flies, which is fixable. But another well-known company (again, I will not name it so that there is no anti-advertising) in such a situation, the controller dies. That is, there is reliability.
I myself use WD, and have never noticed any problems. Here I have 12 hard drives from WD with different data: 8 pieces of "black" 2-3 terabytes each, a few more "greens", which are no longer being produced now. Some of them used to work in computers, but now they are used for archives and feel great. By the way, we are now opening a computer club, and WD Black 500s, M.2, are just standing there. Why were they chosen? Because in terms of price, reliability and performance, everything is more than satisfied (as for me, the most adequate offer now).
Is there really, absolutely no claims against Western Digital?
For all the time I have been working with this brand, I have only positive impressions, this is a personal experience. Of course, on the same Yandex.Market, a different picture emerges, but again, all reviews must be analyzed correctly. Ideally, when choosing an SSD or HDD, you need to do this: you take, say, four models from different companies that are in the same price category and compare. Whatever one may say, it is foolish to demand phenomenal speed from the budget line. Not to mention the fact that a mass product is mass-produced for that: more devices - more marriage. Plus, the curvature of users is added from above. And the same hard drives are quite delicate things. If these factors are taken into account, everything falls into place.
Although, in general, I have complaints about Western Digital. I think that they really lack really top-end, fashion solutions in the SSD segment. WD has top-end drives, top-end NAS, and it would be cool to see some more, shall we say, premium SSDs. I mean something like 970 Pro. Yes, such solutions are expensive and not everyone needs them. But I am sure: if Western Digital had created something similar, they would have easily squeezed Samsung into the market. It would also be great to see something interesting in terms of hybrid drives: at one time, WD developed this direction quite well, but now there are no new products to be seen.
Let's now move on from iron directly to Zenith. Tell us, what are the features of this platform and how does the second version differ from the first?
In terms of size, Zenith is a Midi-Tower, but the case itself is open-type with an inclined motherboard. It can install two 2,5-inch drives, four 3,5-inch drives, it supports the installation of 5,25 devices - in this regard, everything is standard. A thick 40 mm radiator can be placed on the front panel, and a 360 mm radiator (we used Aquacomputer Airplex Radical 2) for CPU water cooling is placed on top. In fact, everything with technical features.
Although no, there are still chips. Firstly, protective glass on permanent magnets, such a mount itself is our know-how. Secondly, we implemented passive cooling of installed hard drives. Heat is removed from the drives to the case itself through thermal pads (we used Thermal Grizzly 3 mm thick). We checked it on WD Red Pro and Black: on “eds” it turned out 5-7 degrees lower than under air cooling, and on “blacks” - by all 10. But the most important thing here is good cooling of the controller and cache. There is no throttling, due to which a stable speed is ensured.
But, in general, Zenit is not only about performance characteristics. It is primarily about design and quality. We don't use cheap materials, we have a solid 3 mm thick aluminum frame, which can be lifted with one hand without any problems. We have high-quality Black Silk powder coating (by the way, we repainted the case already 4 times, because such paint does not fit well in bends, so we had to remove defective layers by sandblasting, grind and reapply), we also have chrome-plated copper tubes, not acrylic. In general, Zenith is about aesthetics. This is a show project, which with all this can be a home computer. Well, it's like with expensive disks for a wheelbarrow: it seems that it's not clear why they are, but, damn it, cool!
And the famous "beauty requires sacrifice" - is this not about Zenit? My point is that often when case or finished PC manufacturers try to make a design thing, it turns out to be terribly impractical. Without a hammer and a file, you can’t install a motherboard, or shove a disk, it makes noise and stuff like that.
No, this is not about Zenit at all. Technically, it is ready to be assembled even by a schoolboy. Of course, instructions for it should be made ... and then you can immediately start mass production. On the other hand, the production of "Zenith" is a separate story: there is a lot of carving, a lot of soldering, in general, a lot of manual work. But if we had an order for a batch, I think that I could optimize the design precisely in terms of modularity.
Noise: the configuration we made turned out to be very quiet. We set the turntables to Coolermaster at 1500 rpm, the pump - Watercool HEATKILLER D5-TOP. All of this worked great with the Threadripper overclocked to 4 GHz, and the noise level was quite comfortable even for an apartment.
Tell us more about the RAID itself. Of course, we are not going to make a guide to setting up an array now, but describe it in a nutshell so that our readers understand how difficult it is (or vice versa).
In fact, it is more difficult to assemble a RAID from hard drives on a SATA controller than on solid state drives. The point is pretty simple. We used 8 NVMe SSD WD Black. Each drive uses 4 PCI Express lanes for a total of 32. Threadripper has 32 lanes on each side. Accordingly, you need to correctly use 16 lines on one side and 16 on the other (or 8 and 8, for example, if there are fewer drives). The main thing is that there is no skew, full mirroring is needed: if you put 8 on one side and 4 on the other, then there will be a very strong drop in performance. All this is done in the BIOS. And then you launch the operating system, launch AMD RAIDXpert2, create the desired array - and voila, you're done! It turns out very reliable, and most importantly - very fast storage.
That is, no pitfalls and dances with a tambourine? Any more or less advanced user can handle it without problems?
Yes, anyone who understands what an M.2 drive is will be able to set up such a RAID. But you still need to understand a little about the topic. As I said, this is precisely the drawback of AMD software - they do not have a purely consumer solution in the style of "clicked - and it all worked." The only problem I had was that Windows 10 didn't want to pull the driver, and because of that, the array couldn't be used as a system drive. But these are already revision jambs: I ran into problems on assembly 1803, and this was fixed for 1909 - the necessary firewood is pulled up automatically.
Do you plan to somehow further develop Zenith? Can we expect MKIII with even crazier stuff?
Zenit is a very cool project, one of the most successful and quickly implemented by us. I consider this case almost perfect and fully accomplished both as a show project and as a consumer PC. In addition, it has become a valuable base for us in terms of design, metalworking, painting, layout, cooling - that's it. And I would really like to make this project serial. For this, in general, everything is there. But it so happened that no one wants it. "Zenith" - cool, but not massive.
For us as a team, he was left behind. We are moving forward, participating in international modding competitions, developing new cases. In light of this, I do not see much point in reanimating and somehow rethinking Zenith. It is a thing of the past, now we have cooler and more interesting concepts that are worth trying to implement.
Source: habr.com