Principled Tech (@PrincipledTech) header
Principled Tech (@PrincipledTech) profile picture

Facts matter. Partner with PT for all your marketing, learning, and testing needs.

Durham, NC
Joined August 2011
Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)Photo by Principled Tech (@PrincipledTech)
See how @surface Laptop Ultra and other Windows laptops with NVIDIA #RTXSpark compared to an Apple MacBook Pro with M5 Pro in AI, gaming, and 3D tests commissioned by @Microsoft facts.pt/WindowsRTXSpark
65
Get more database work from every watt. In SQL Server testing, @Dell #PowerEdge R4715 configurations with @AMD #EPYC processors delivered up to 87% more performance per watt than a legacy HPE server: facts.pt/VrPkcpi
18
No GPUs yet? Start on CPUs. With Intel AMX enabled, dual @intel #Xeon 6979P processors in a @Dell PowerEdge XE9780LAP supported 57 AI agents on a moderate financial #agenticAI workload within a 30-second SLA. facts.pt/88p1CEx
16
Scaling up matters. PT's hands-on tests found @AzureSQL Database Hyperscale matched or outperformed Amazon Aurora PostgreSQL at 32 vCores, with up to 12% better performance per dollar: facts.pt/QE8NpwH #managedDBservice #SQL #cloud
15
Help students stay productive all day. A @Lenovo #Chromebook w/ a @MediaTek Kompanio 540 processor stayed responsive while multitasking on battery and lasted 15 hours and 44 minutes while streaming video. Learn more: facts.pt/chromebooks-unplugg… #edtech
1
39
No throttling through long work sessions: After a 30-minute sustained load, the @HP #EliteBook X G2a delivered up to 25.2% better Cinebench 2026 performance than two premium 14-inch competitors, with a cooler chassis underside: facts.pt/wMfFrxz
15
Decode-heavy AI workloads stress inter-GPU communication. We saw wins for a @googlecloud G4 VM on our decode-heavy workload: 9.8% more output TPS, 26.8% faster TTFT, 7.3% lower E2E latency vs. AWS EC2 G7e. facts.pt/e7aXRMJ #AIInference #CloudPerformance
41
We compared the SQL Server performance of new @Dell #PowerEdge R4715 servers w/ @AMD EPYC 9005 Series processors to an older HPE ProLiant DL320 Gen11. Check out our report to see how a server refresh can help growing businesses: facts.pt/Dell-R4715-value #SQL
11
Spot risk and protect customers in real time: The @Dell PowerEdge XE9780LAP with @intel Xeon 6 CPUs supported up to 62 financial services AI agents for as little as $1.26 per million tokens. facts.pt/0StB9FX #AgenticAI
2
56
New infographic: Routine management and lifecycle updates: Same time, same effort: facts.pt/tfoxgzm @Dell #EndpointManagement
8
A @MediaTek Kompanio 540 processor-based Chromebook delivered up to 4 hours and 24 minutes longer battery life than Intel processor-based Chromebooks in PT hands-on testing: facts.pt/irc6uHY
18
In PT testing, the AMD Ryzen AI 7 PRO 450 processor-powered @HP #EliteBook X G2a finished first in every performance benchmark we ran against the Dell Pro 14 Premium and Lenovo ThinkPad X1 Carbon Gen 14. See the results: facts.pt/wMfFrxz
42
Catch problems before they cause downtime: In PT tests, the @Dell PowerEdge XE9780LAP with @intel Xeon 6 CPUs supported up to 205 manufacturing AI agents for as little as $0.43 per million tokens. facts.pt/0StB9FX #AgenticAI
12
New infographic: Out-of-band management: Same capabilities, single console: facts.pt/GJMjWYk @Dell #RemoteManagement
11
Running #agenticAI coding on device, an @HP ZBook Ultra G1a 14 inch mobile workstation w/ an @AMD #Ryzen AI Max+ PRO 395 can pay for itself in 15.7 mos. vs. a leading frontier cloud model at API list rates, saving $6,574 per system over 3 yrs. Read more: facts.pt/AMD-workstation-ROI
46
At 16 vCores, @AzureSQL Database Hyperscale beat Amazon Aurora PostgreSQL at nearly all concurrency loads PT tested, by as much as 52% on performance and 51% on performance per dollar: facts.pt/QE8NpwH #managedDBservice #SQL #cloud
19
Running agentic AI for manufacturing? In PT testing, a @Dell PowerEdge XE9780LAP with @intel Xeon 6979P CPUs and eight NVIDIA B300 GPUs supported 205 AI agents for as little as $0.43 per million tokens. facts.pt/88p1CEx #AgenticAI
29
A @MediaTek Kompanio 540 processor-based #Chromebook earned up to a 34% higher PassMark rating and up to 4 hours and 24 minutes more #batterylife vs. Intel N-series processor-based Chromebooks in our tests. Read more: facts.pt/chromebooks-unplugg…
1
21
Running multimodal AI in the cloud? A @googlecloud G4 VM was delivered a faster time to first token and lower latency vs. AWS EC2 G7e—with the same NVIDIA RTX PRO 6000 GPUs. Learn more: facts.pt/e7aXRMJ #AIInference #CloudPerformance
29
A @Dell #PowerEdge R4715 with a 32-core @AMD #EPYC 9335 processor delivered 2.4x the SQL Server NOPM of a legacy HPE ProLiant DL320 Gen11 while staying within the SQL Server Standard 32-core licensing limit: facts.pt/VrPkcpi
24