Interview with Anand Lal Shimpi of AnandTech.com
Benchmark Reviews has educated and entertained readers for almost four years already, which doesn't seem like very long if you're looking in from the outside. Operating from within this industry, I've already weathered a few harsh storms and learned several difficult lessons in only a short period of time. This helps to grow my respect for other webmasters and their experiences, as I'm certain they've each got interesting tales of their own to share. In this editorial, Benchmark Reviews is pleased to present our interview with the author and founder of Anandtech: Anand Lal Shimpi.
Over the years, what has been the single topic that continues to motivate your audience?
Five years ago we were still talking about K8s and Pentium 4s. Today it's Sandy Bridge, SSDs, Brazos, Boxee, smartphones. There isn't a single topic that motivates the audience, but the common thread is really the pace of innovation in this industry. The fact that you can spend $200 on a drive or CPU upgrade today and get something as much as an order of magnitude faster than what you bought a few years ago is what keeps people motivated. This is an industry that's easy to get excited about, it's a game of just waiting to see what people will do next.
Today we have more processing power and I/O power than we've ever had, and much of it goes underutilized. A high end desktop doesn't need 40K IOPS from an SSD, but we have it. Once the pieces start falling into place and we see some more innovative software begin to take advantage of all of this things will get really exciting. The fact that we're constantly in pursuit of the technology and user experiences you see in science fiction is what I believe motives the AnandTech audience. We all just want to be a part of it and learn about it as much as we can.
From the time you began testing hardware, how has the industry impacted your opinion of reporting real-world performance?
I view it as a challenge. Characterizing real world performance is incredibly difficult. You have manufacturers working against you by trying to optimize for the way you test and you have the applications working against you by not necessarily making it easy to repeatedly measure real world performance. Then, as an added bonus, you have to navigate around the issues presented by those two and paint a complete picture for the readers without burdening them with the struggles.
We have a number of systems in place to ensure that we can adequately report real world performance. We create a lot of benchmarks in house, we work with software and hardware vendors to put together new tests, we also influence the creation of industry standard benchmarks to make sure that the readership's needs are represented as best as possible. If there's something we need to test, we find a way to test it and present it to the readership. We also sometimes run tons of data that never even ends up on the site in order to make sure what we're writing is accurate. I have 22 excel sheets with thousands of numbers dedicated just to SSDs. I have a little less than that dedicated to CPUs. Only a portion of that data ever makes its way into a review, the rest is background information, theory testing to help me understand what the actual real world performance is of a product.
We've dealt with golden samples and application specific optimizations before and we continue to do so on a regular basis. We have no problems going to Newegg and buying something to verify its performance, and we have no issues switching up the way we test in order to stay one step ahead of benchmark optimizations. One thing we don't mention publicly but we've done in the past is consciously switch up our test applications in order to get vendors to optimize for applications that were being neglected. It's one thing to fill out a bug report, it's another to put an application or game front and center in a review and show poor performance. The latter usually results in a fix much, much quicker. We've done this several times over the years.
Anything we recommend on AnandTech, anything we suggest that someone part with their hard earned money for, ends up fully integrated into the daily lives of at least one of our editors. If you're going to recommend it, you had better well use it. Not just benchmark it, but actually use it. It doesn't matter whether it's a CPU, some memory or a smartphone, we owe it to the readership to stand by our recommendations.
Where do you see the desktop computer industry and its enthusiasts in five years?
A decade ago a PC was easy to understand - it was a beige box that sat on or under your desk. These days you might have half a dozen "PCs" in the room you're sitting in. You might have a desktop, a laptop, a netbook, a tablet, a smartphone or a media streaming box just within a 30 foot radius of you. Desktops are still sexy to me because of their TDP advantage. You get newer technology faster than you do in the mobile space and ultimately you get tons of performance. I don't see that industry going away, although it's very much a mature market. You'll continue to see innovation there, better performance, new features, new OSes, but the rate of growth in that industry isn't going to be as high as it was ten years ago.
AMD's Carrell Killebrew has been talking about a first generation Holodeck in the next 6 years. That experience will be powered by (at least in one form) a desktop PC. Right now, as a desktop enthusiast, I feel like I'm still waiting for a lot of the software to catch up. It'll happen, and when it does I expect we'll have some very happy enthusiasts.
The growth today is in mobile and what's exciting there is we get a repeat of the desktop PC industry over the past 20 years but in a much more condensed timeframe.
What would you like to see changed in the computer hardware industry?
I tend to have a broad definition of what the computer hardware industry is. Everything from the PCs that are being designed for cars in the next 3 - 5 years to the desktops we play CoD on fall under the umbrella of my definition. To be honest, I don't really have many complaints about this industry. I love it. I don't always like how some of the companies behave, but we generally have a good number of balancing forces in this industry. The users, the review sites and the competing manufacturers all generally work to make sure that the industry as a whole always moves forward and does what's best. It's more than I can say for any other industry. It's one of the reasons I've stuck around for nearly 14 years at this point.
What I'd love to see changed however is the direction the web has taken over the past few years. I've spoken about this at a number of events recently, but it's what I view as the Cable TV-ification of the web. We all love to complain about the poor quality of mainstream news organizations on Cable TV, but what's scary is that much of the web is following in those very same footsteps. The focus isn't on understanding what's being reported, it's not on educating the reader, it's first and foremost about entertaining the reader by any means necessary. I believe there are two approaches to any journalistic endeavor: entertainment through education and education through entertainment. As long as there's a balance between the two approaches we're fine, but I believe the latter is what much of the web has turned into over the past several years.
If there was one lesson you could teach new hardware enthusiasts, what would it be?
This is probably the most difficult question to answer. I'm not sure there's any grand lesson I can offer new hardware enthusiasts. If anything I'd say it's very important for people passionate about this industry to be involved in it. Learn as much as you can, read everything that's out there, comment, post in forums, give manufacturers direct/honest feedback and even try writing for the sites you like to follow. The power is entirely in your voice.
I always tell people that feedback is like free education. Even if harshly worded, the fact that someone is willing to give you their opinion without any sort of compensation is gold. Keep that in mind as whether you're talking about a website, another enthusiast or a manufacturer, your feedback and opinions are extremely valuable.
On behalf of the team here at Benchmark Reviews, we sincerely thank Anand Shimpi for taking time out of his schedule to freely answer our questions for this interview. Anand's been the driving force behind Anandtech for nearly fourteen years, and we look forward to sharing in their experience of this industry for many more.
Hopefully this interview has helped our readers to better understand the inner-workings of the computer industry, and gain perspective into our trade. We welcome your comments and suggestions below:
Related Articles:
|
Comments
"The focus isn't on understanding what's being reported, it's not on educating the reader, it's first and foremost about entertaining the reader by any means necessary. I believe there are two approaches to any journalistic endeavor: entertainment through education and education through entertainment. As long as there's a balance between the two approaches we're fine, but I believe the latter is what much of the web has turned into over the past several years."
Yes and that is how it will always go until changes are made at the code level--that is, at the system level. Capitalism is driven by the masses and statistically, we're mediocre. We're mediocre in just about every way. So the masses demand entertainment that is base, or fundamentally "mediocre" at best.
Until we as a society demand a change that demands we have available factual, content oriented information, garbage is what we will get because the masses demand it. So what is a "system level change?"
That's a good question, and I think it starts in grade school by society taking very seriously developing in people a very strong sense of culture, that being arts, sciences, literature and reward investigation into those areas. Note I said "ivestigation" not necessarily "becoming famous and being in the top 0000000.% of all humanity," or those "investigations" are "idle." That's how we think now.
Today, we teach survival, raw, brutish survival. We teach it at the grade school level, and we teach it every day and every hour to the masses everyday of the year. In other words, there are only two things worthy of our attention: money and escape. Thinking is a fools errand because no one is listening, and no one -- at the code level -- can appreciate it.
Hank
Let's not drift into defamation here.... People have relationships, companies have relationships, heck the Supreme Court just ruled that corporations are "persons", and we're ALL biased. It's how we humans have survived for thousands of years, by making quick judgements based on incomplete data. So, even with the best intentions and a clean heart, we all have our own personal favorites. That's whey we rely so much on objective measurements in this industry and on this site.
Even when I don't understand the numbers, I still publish them. I'm actually hoping someone can come on the site and offer a rational explanation why. I won't publish data I know is bad, but if the numbers seem artifically high or low, I still want to find out why, even if that means a large dose of humility when someone tells me in a public forum what an idiot I am.
Ultimately, I disagree with Anand. I don't think it's all about education, but I do see it as a communal learning process. The term education sounds too one-sided to me. Having said that, Ive probably learned infinitely more from Anand than he has from me...
While humans may have survived by making quick judgments based on incomplete data, sometimes the results have not been pretty. (Actually, I like your little statement, quite true.)
Regarding education, the more I learn about something, the more I realize there is so much more that I don't know.
An act of communication that causes someone to be shamed, ridiculed, held in contempt, lowered in the estimation of the community, or to lose employment status or earnings or otherwise suffer a damaged reputation. Such defamation is couched in 'defamatory language'. Libel and slander are defamation.
Although defamation is primarily governed by state law, the First Amendment safeguards for freedom of speech and press limit state law. New York Times v. Sullivan, 376 U.S. 254, 264 (1964); Masson, 501 U.S. at 510. The scope of constitutional protection extends to statements of opinion on matters of public concern that do not contain or imply a provable factual assertion. Milkovich, 497 U.S. at 20 (rejecting categorical exemption of all statements in form of opinion; statement that may imply verifiable assertion of fact is actionable). (##lectlaw.com/def/d021.htm)
The above slam is clearly the riffing of opinion and should be taken as such. It doesn't matter if he started the sentence with "It is true. . . ." Rush Limbaugh starts virtually every breath with "It's true . . . " and never get sued for defamation because what he is really saying is "It's my opinion that. . . ." but that phrase has less emotional appeal than starting a sentence with "It's true!"