Market analysis firm The Aberdeen Group, who gets paid by corporations to do the kind of benchmarking and analysis we do for free in public, recently released a report criticizing AMD's new "model number" rating system for its Athlon XP processors. The report, entitled "AMD's Gigahertz Equivalency: Inexperienced Buyers Accept Bad Science," isn't exactly what one would call even-handed. The Aberdeen analyst ignored a number of relevant facts in an apparent attempt to attack his imagined, bad opponent instead of AMD's actual, somewhat nuanced rating system.
For instance, the Aberdeen report criticizes the AMD rating system as a false equivalency for Pentium 4 clock speeds, but AMD was careful to state the ratings are based on the theoretical performance of Athlon T-birds at a given speed (since Palomino-based Athlon XPs are faster clock for clock). Splitting hairs? Maybe, but it's important to understand that which you're being paid to critici... err, analyze.
There are some legitimate reasons to gripe about AMD's rating systems, but the author of the Aberdeen paper seems unaware of them. Instead, he claims the "key flaw is that the equivalency rating is a snapshot in a moment in time," seemingly unaware that AMD has stated its current rating system is a "bridge metric" to something ostensibly better coming from its True Performance Initiative. In other words, AMD has itself stated the current model ratings are the product of a "snapshot in time."
I could go on, but all of that's beside the point. The real story here emerged shortly after publication of the report, when The Inquirer found out that Intel funded the Aberdeen "study." Which explains pretty much everything.
I won't editorialize too much on this bone-headed maneuver. Go read Mike Magee's take if you want to hear more of that. But I would like to point out that sites like The Tech Report are the perfect antidote this kind of corporate BS. Our in-depth, freely available CPU performance reviews show an Athlon XP 2100+ just edging out a Pentium 4 2.2GHz in our benchmark suite. If AMD is in fact practicing to deceive consumers, they're being overly modest.
|1. BIF - $340||2. chasp_0 - $251||3. mbutrovich - $250|
|4. Ryu Connor - $250||5. YetAnotherGeek2 - $200||6. aeassa - $175|
|7. End User - $150||8. Captain Ned - $100||9. Anonymous Gerbil - $100|
|10. Bill Door - $100|
|Thermaltake View 27 case offers a birds-eye view of builds||26|
|National Dog Day Shortbread||28|
|Corsair backlit keyboard lineup gets new Lux models||10|
|Nixxes turns out another Deus Ex: Mankind Divided patch||22|
|Upcoming Samsung CF791 is a high-contrast FreeSync ultrawide||50|
|Deals of the week: an unlocked Skylake CPU for cheap and more||19|
|PCIe 4.0 won't actually deliver 300 watts from the slot||57|
|iOS 9.3.5 fixes serious zero-day vulnerabilities||13|
|Intel 600P Series SSDs bring NVMe into the M.2 mainstream||42|
|Stupid physics getting in the way of all our fun.||+34|