• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Anandtech.com is gone

OKC Yeakey Trentadue said:
Its still sad to see all the legends leave. Toms, Anand, Hardcop. Most have a max 2 minute attention span now so the click baiti short format videos are the only things profitable.
Click to expand...
I mean, you say that, but how many of you actually click on the GamerNexus 30+ minute long videos and watch them? Like others have said, nothing new. People just skip through to the conclusion. No one has the time for that.

Well, to be honest I read through some of them and sometimes watch those 30+ minute videos though, so I guess I'm a weirdo.

I miss how Anand used to have that area where you could easily compare CPUs and GPUs by comparing them across the same game, but I guess Techpowerup has sort of taken over that anyway.
 
1_rick said:
https://www.techspot.com/news/108967-anandtech-27-year-archive-has-vanished-but-someone.html

The website, and all the articles are gone; anandtech.com now redirects to the forum sub-site.

Someone uploaded a 74GB archive, link is in the article.

Future's assertion that "the site will stay up indefinitely" lasted less than a year.
Click to expand...
I'm sad but not hugely surprised.

IIRC Anandtech was never migrated from its homebrrew platform to one of it's corporate overlords consodlidated CMS platforms. That meant that even with the best of intentions, it was never more than one major technical issue from costing more to maintain than it was likely to generate in ongoing revenue.
 
DanNeely said:
I'm sad but not hugely surprised.

IIRC Anandtech was never migrated from its homebrrew platform to one of it's corporate overlords consodlidated CMS platforms. That meant that even with the best of intentions, it was never more than one major technical issue from costing more to maintain than it was likely to generate in ongoing revenue.
Click to expand...
Isn't it interesting, then, that Toms' announced a new paid area, and part of that will include a new benchmark section called "Bench"? Be interesting to see if that's using Anandtech's data.
 
What happened here? I've been waiting patiently for nine months. :(
2025 is another year of changes here at [H]. We are currently looking at doing some new things this year, some old things this year, and will be letting you know of those things coming up soon. Happy New Year! -Kyle Bennett
Click to expand...
 
  • Like
Reactions: TMCM
like this
DooKey said:
Rat Padz are coming back.
Click to expand...
This is the old things, I think, although it could be the new things I guess. Either way, I'm not sure what the other is, haven't heard anything else.
 
output (1).png
 
1_rick said:
Isn't it interesting, then, that Toms' announced a new paid area, and part of that will include a new benchmark section called "Bench"? Be interesting to see if that's using Anandtech's data.
Click to expand...
Since this thread's been necroed, has anyone paid Toms to see if they're using their own data, or just wrapped a paywall around something ripped from Anandtech's desecrated corpse?
 
DanNeely said:
Since this thread's been necroed, has anyone paid Toms to see if they're using their own data, or just wrapped a paywall around something ripped from Anandtech's desecrated corpse?
Click to expand...
There was a recent paid post on upcoming GPU landscape.
For AMD it referenced a techpowerup article which in turn referenced an outdated AT Forum post

They didn't even do the 'research' to check latest AT forums before publishing their paid article


Potentially using this, can expect a range from 96 CUs all the way down to 32 CUs for UDNA, with GDDR7 memory being the VRAM technology of choice.
Click to expand...

https://www.tomshardware.com/pc-com...nvidia-rubin-amd-udna-and-intel-xe3-celestial
 
Dan_D said:
The issue is return on investment for the time put in.
Click to expand...

Sorry, late to the party, but don't you have to do a lot of the grunt work to get to the conclusion and fancy graphs anyway? I don't write reviews for anything, but it seems like writing a few paragraphs detailing the experience during testing wouldn't take THAT much time compared to doing the actual testing and inputting the info into excel or whatever to generate the graphs. But then again, I don't write reviews, so don't shoot the guy just asking a question :D.
 
kirbyrj said:
Sorry, late to the party, but don't you have to do a lot of the grunt work to get to the conclusion and fancy graphs anyway? I don't write reviews for anything, but it seems like writing a few paragraphs detailing the experience during testing wouldn't take THAT much time compared to doing the actual testing and inputting the info into excel or whatever to generate the graphs. But then again, I don't write reviews, so don't shoot the guy just asking a question :D.
Click to expand...
For the reviews I did for HardOCP, the written part took about as much time as the physical testing did. Those articles were long. Ours were much longer and far more detailed than the stuff you get on the sites that are still running now. There were parts of the article that served as a template with verbage that was used talking about testing methodology and that kind of thing but that made up very little of the article. Everything else was written fresh each time. Writing is more of a creative endeavor than you think with going back and fourth and rewording sections and trying to reduce redundant text or words to be more concise. I tried to make the text more interesting rather than read like some AI chat bot did it and when you've written your 10th article about an ASUS, GIGABYTE or MSI board using the same chipset, well it gets harder and harder to do that. It can be quite a slog to write something like that.

Not to mention, the graphs were horrible to do and get right. We had to do them a certain way which often meant lining things up just the way we were told to do it. Things had to be in a certain order, etc. It wasn't difficult so much but it was extremely time consuming. Parts of it like the network testing, drive testing etc. were also really a pretty dry subject and sometimes parts like that were the hardest to write.

Articles were on average 25-30 hours worth of work and sometimes more if there were problems. At least 10-15 hours of that for me was spent on the written part. The testing was down to a science and was pretty easy to do. Though overclocking could be problematic back in the day as some boards were really hard to work with or just wouldn't give you good results.
 
Well looked at their forums that seem to be back up for a couple minutes and I strongly encourage them to take them down permanently...what a cesspool
 
Last edited:
The ongoing thread there about NVMe USB enclosure/controller reliability is still useful, once you filter out the noise. I think there's a similar thread on another forum but can't recall the site.
 
I actually miss the days of kicking back with a coffee and reading a good tech review. In reading a review, I could go back and forth quickly and easily to correlate descriptions with results, and life was good. In comparison, tech tube video's where the commentator talks like he's OD'd on caffeine are painful to follow and you can't easily go back and forth to correlate descriptions with results - As a result I avoid them entirely.
 
Dan_D said:
All the network testing, drive controller testing, DPC latency testing and things like that were the longest and hardest content to write in the motherboard reviews and no one read it. We used the same format initially on TheFPSReview. Guess what? No one read that stuff.
Click to expand...
And that's really too bad. DPC Latency is still a thing and most people don't even realize that their computers have high DPC latency when their computers are stuttering.
 
Last edited:
Zarathustra[H] said:
They apparently did some testing on articles, and it turned out everyone skipped over all of the detailed content, and went straight for the performance charts and conclusions.
Click to expand...
Dan_D said:
All the network testing, drive controller testing, DPC latency testing and things like that were the longest and hardest content to write in the motherboard reviews and no one read it. We used the same format initially on TheFPSReview. Guess what? No one read that stuff.
Click to expand...
WHAAAAAAAAAAAAAAAAAAAAAAAAT
Dang that sucks to hear. I didn't know it was like that, wow.
 
DrezKill said:
WHAAAAAAAAAAAAAAAAAAAAAAAAT
Dang that sucks to hear. I didn't know it was like that, wow.
Click to expand...
We are but a small minority. Not a lot of people are interested in the technical "how's" or "why's" anymore. Which is to our detriment I think, as a society. Understanding how and why it works is fundamental to understanding how to recreate or fix it.
 
Dan_D said:
All the network testing, drive controller testing, DPC latency testing and things like that were the longest and hardest content to write in the motherboard reviews and no one read it. We used the same format initially on TheFPSReview. Guess what? No one read that stuff.
Click to expand...
I mean, I skipped over that stuff but that was only because I never use any on board devices, so the testing of them never mattered to me.

Only on board devices I ever use are USB ports, and I don't use that many of them. I haven used any on board sound, networking or even SATA ports in over a decade at this point.

I really did appreciate the testing of PCIe latency, especially the comparisons of using NVMe drives in direct to CPU vs Chipset slots though. I found that to be really valuable.
 
Last edited:
jbltecnicspro said:
We are but a small minority. Not a lot of people are interested in the technical "how's" or "why's" anymore. Which is to our detriment I think, as a society. Understanding how and why it works is fundamental to understanding how to recreate or fix it.
Click to expand...
They never had that. That's why it was so infuriating. Nerds fighting to the death about a couple of numbers and they had no idea what those numbers actually meant, or how to interpret the statistics behind those numbers (or lack thereof as stuff like standard error is rarely reported)
 
billabong132 said:
They never had that. That's why it was so infuriating. Nerds fighting to the death about a couple of numbers and they had no idea what those numbers actually meant, or how to interpret the statistics behind those numbers (or lack thereof as stuff like standard error is rarely reported)
Click to expand...
I hate to necro this but I just read these two pages and what you said is something I've always struggled with. For example, I rebuilt my system when Intel 11th gen came out, and everyone told me I bought the worst Intel generation of all time. I'm still using it, bought it in May of 2021. I knew it wasn't a world beater. But the FPS differences between this platform and my four prior Ryzen chips (1700, 1800X, 2700X and 5700X) doesn't matter at all when your USB ports are disconnecting on any CPU load and other issues I fought for years on X470/X570. What is true is that 11th gen is the quirkiest of the Intel platforms released in the last 20 years. All of the issues I had on 11th gen were because I was too cheap and purposefully deluded myself to avoid replacing an old Corsair SF600 PSU until I moved to an ATX 3.1 SF1000.

There really is no place to go to get a balanced perspective or any sort of nuance. People back in the day, just looking at the FPS charts, do the same thing now.

But even articles like on Anandtech and HardOCP never tried to explain concepts that you're not buying a CPU, you're buying or investing in a platform. When you choose between say AMD and Intel. It always stayed within the lane of "trying to be fair" when that's not the whole story. So in a way, they led their audience to this thought pattern in the end.
 
housecat said:
I hate to necro this but I just read these two pages and what you said is something I've always struggled with. For example, I rebuilt my system when Intel 11th gen came out, and everyone told me I bought the worst Intel generation of all time. I'm still using it, bought it in May of 2021. I knew it wasn't a world beater. But the FPS differences between this platform and my four prior Ryzen chips (1700, 1800X, 2700X and 5700X) doesn't matter at all when your USB ports are disconnecting on any CPU load and other issues I fought for years on X470/X570. What is true is that 11th gen is the quirkiest of the Intel platforms released in the last 20 years. All of the issues I had on 11th gen were because I was too cheap and purposefully deluded myself to avoid replacing an old Corsair SF600 PSU until I moved to an ATX 3.1 SF1000.

There really is no place to go to get a balanced perspective or any sort of nuance. People back in the day, just looking at the FPS charts, do the same thing now.

But even articles like on Anandtech and HardOCP never tried to explain concepts that you're not buying a CPU, you're buying or investing in a platform. When you choose between say AMD and Intel. It always stayed within the lane of "trying to be fair" when that's not the whole story. So in a way, they led their audience to this thought pattern in the end.
Click to expand...
That's because what people were saying at the time was absolute rubbish. The only difference between an 8th gen i7 and a 10th gen i5 is clock speed and the speed of the IMC, and the 8th gen i7 will overclock to the speed of the 10th gen i5 and pull 3600MHz on it's IMC easily. Even considering the two additional cores/threads of the 10th gen i7, the difference was marginal at best - So from 8th gen > 10th gen was a sidegrade at best as they were all basically the exact same CPU.

When you consider the 11th gen CPU's: Not only do they have 16 + 4 pcie lanes and pcie 4.0 (10th gen are pcie 3.0 @ 16 lanes only), they also support AVX-512, and they have more L1 cache over the 10th gen CPU and double the L2 cache of the 10th gen CPU - a definite upgrade over the 10th gen CPU.

Personally, I've always taken tech reviews with a massive grain of salt.
 
housecat said:
I hate to necro this but I just read these two pages and what you said is something I've always struggled with. For example, I rebuilt my system when Intel 11th gen came out, and everyone told me I bought the worst Intel generation of all time. I'm still using it, bought it in May of 2021. I knew it wasn't a world beater. But the FPS differences between this platform and my four prior Ryzen chips (1700, 1800X, 2700X and 5700X) doesn't matter at all when your USB ports are disconnecting on any CPU load and other issues I fought for years on X470/X570. What is true is that 11th gen is the quirkiest of the Intel platforms released in the last 20 years. All of the issues I had on 11th gen were because I was too cheap and purposefully deluded myself to avoid replacing an old Corsair SF600 PSU until I moved to an ATX 3.1 SF1000.

There really is no place to go to get a balanced perspective or any sort of nuance. People back in the day, just looking at the FPS charts, do the same thing now.

But even articles like on Anandtech and HardOCP never tried to explain concepts that you're not buying a CPU, you're buying or investing in a platform. When you choose between say AMD and Intel. It always stayed within the lane of "trying to be fair" when that's not the whole story. So in a way, they led their audience to this thought pattern in the end.
Click to expand...
13th and 14th likely much worse than 11 due to all the chip failures

https://forums.anandtech.com/threads/degrading-raptor-lake-cpus.2608723/page-9
 
Mazzspeed said:
That's because what people were saying at the time was absolute rubbish. The only difference between an 8th gen i7 and a 10th gen i5 is clock speed and the speed of the IMC, and the 8th gen i7 will overclock to the speed of the 10th gen i5 and pull 3600MHz on it's IMC easily. Even considering the two additional cores/threads of the 10th gen i7, the difference was marginal at best - So from 8th gen > 10th gen was a sidegrade at best as they were all basically the exact same CPU.

When you consider the 11th gen CPU's: Not only do they have 16 + 4 pcie lanes and pcie 4.0 (10th gen are pcie 3.0 @ 16 lanes only), they also support AVX-512, and they have more L1 cache over the 10th gen CPU and double the L2 cache of the 10th gen CPU - a definite upgrade over the 10th gen CPU.

Personally, I've always taken tech reviews with a massive grain of salt.
Click to expand...
If I had a crystal ball I would have waited for 12th gen which I view as much lower stature but the closest thing Intel has had to a modern era Sandy Bridge.

I was having a lot of issues with my 5900X at the time and held out for Intel to launch 11th gen for the reasons you stated. Intel's PCIE 4.0, Thunderbolt 4 (also being a functional USB 4.0 port), and had read an interview with an Intel engineer that claimed they weren't going to be giving up on AVX512. So I assumed it was a good thing to have. I think looking back, having the last of the monolithic cores from Intel, with bugfixes for bugs left in chips from the Skylake era regarding L0 cache bugs and parity errors.

Which is general info which I got from this rather informative post. https://rog-forum.asus.com/t5/intel...ocket-lake-the-rules-have-changed/td-p/812999 It was really that post versus all the "but 10 cores" crowd. I was going from a 12C/24T chip to 8C so 10C with PCIE 3.0 was a hard sell. Reddit just defends whatever they happen to own at the time. Reddit definitely punished me for it. I no longer use Reddit. But yes, if only I had that crystal ball I would have waited less than 12 months for 12th gen. It has still been a good system for me since 2021.

I do think some, if not all of my 5900X issues would have been sorted out with a better PSU, but I'm also not so sure about that. I got to the point with my X470/X570 systems where I was doing a CMOS reset any time I changed a BIOS setting like a maniac. And C-States seemed to be off the table. While my recent issues that I started having with this aging Z590 started when I put in a 5070FE. An ATX 3.1 PSU seems to have sorted that out for me.

I am going to upgrade at a point, but I'm really enjoying things as they are. I'm not closed off to the idea of going back to AMD. I early adopted Ryzen (on miniITX on top of it) and paid the price, mostly with copious amounts of my time. But I'm hoping to hold out until at least Hammer Lake. I figure by then RAM prices might collapse again. And it will just be interesting to see how their unified AI-first design is. It definitely feels like the T800 Terminator will be arriving soon to kill us all.

El_Capitan said:
Lol, maybe if you had an i9 or i7. I've gone through seven 13th and 14th gen i5's with no issues. Cheap, too, since I got them before the BIOS microcode fixes.
Click to expand...
I wondered how widespread that was. It's hard to tell with how sensationalist most tech news can be. It would have definitely hit me since I upgrade once every 5-8 years, I grab the most expensive chip that makes sense for gaming. I would have been pissed. Similar situation, I did buy a Samsung 990 Pro that was hit by the write amplification bug, since I bought on day one. Got rid of it and got a good deal on a 9100 Pro but next build probably go with Sandisk due to being bitter about that.
 
housecat said:
Which is general info which I got from this rather informative post. https://rog-forum.asus.com/t5/intel...ocket-lake-the-rules-have-changed/td-p/812999 It was really that post versus all the "but 10 cores" crowd.
Click to expand...

I run an overclocked 8700k with an overclocked ring ratio, and when I could play Apex Legends under Linux (before they blocked Linux users via updated anticheat - it really wasn't a great game anyway) I never once experienced a crash. This article seems like mostly FUD to me.
 
You must log in or register to reply here.
Back
Top