Get the Guide See Current Sale

4 Trading Platforms Ranked by How Hard They Hit Your Computer

Key takeaways

  • TradingView is the lightest. It renders server side, so 16GB of RAM works if you're disciplined with browser tabs, 32GB if you're honest with yourself, and a benchmark score of 45,000.
  • TradeStation is third. 32GB and a 45,000 benchmark for a few hundred symbols in Radar Screen, 64GB if you're pushing 500 to 600 symbols with EasyLanguage in every column.
  • NinjaTrader is second. 32GB and 45,000 for live trading, but 60,000 or higher if you run tick replay or serious backtests.
  • Thinkorswim is the heavyweight. It runs on Java, claims memory and rarely gives it back. 32GB is the floor, 64GB for heavy options traders, benchmark 45,000 and up.
  • Run two platforms at once and you add the requirements. 32 plus 32 is not 32, and that's the number one reason a machine that looked fine on paper falls apart on a busy day.

Has this ever happened to you? Your machine is perfect at 9:30 right at the open, and then it's useless by lunch. You didn't change a thing. Same machine, same charts, same internet connection. It just got slower as the day went on.

Two traders called me with the exact same complaint in the same week. One of them had a real hardware problem. The other one's machine was completely fine. He was running the heaviest of the four platforms I'm about to rank, and it was quietly eating his RAM all day long.

Thinkorswim, NinjaTrader, TradeStation, TradingView. One of those four runs perfectly well on 16GB of RAM. Another one wants 64GB if you trade options. Same desk, same trader, same job, four times the machine. And nobody selling you a computer prices it that way.

Why judging a platform at 9:30 is the wrong test

Almost every trader has this backwards. Judging a platform by the resources it's using at 9:30 in the morning is like judging a marathon by the first few hundred yards. Everybody looks great at the starting line.

All four of these platforms are polite when you first open them. They're clean, they're reasonable, and Task Manager will tell you everything's fine. The difference isn't the starting number. It's how each one grows when you stack charts, studies, symbols, and hours of market data on top of it.

Some grow in a straight line. You add a chart, you pay for the chart. Fair deal. One of them grows like a teenager's grocery bill. That's the one that gets people.

So the real question isn't which platform is heaviest. It's whether your machine has anything left once the platform takes the resources it wants.

You need two numbers before any of this means anything

Every recommendation below is two numbers: how much RAM, and your processor's benchmark score. You probably know your RAM. Almost nobody knows their benchmark score, and without it the rest of this is just numbers floating past you.

So go get your score. I built a free CPU benchmark test made exclusively for traders. It takes about two minutes, it scores your processor, and it shows you exactly where you land against what these four platforms demand. Run it on the machine you actually trade on, not the laptop on the couch, and definitely not on your phone. If you want the background on where that number comes from, I wrote up the 45K trading benchmark rule separately.

Fourth place: TradingView, the lightest of the four

I know somebody just got mad at me. Relax. TradingView is genuinely the most efficient charting engine of the group. It renders server-side and hands your machine a finished picture per chart. Nothing here touches it.

Which raises an obvious question. If it's the most efficient one, why do I get calls from TradingView users whose machines are crawling?

Because TradingView doesn't run on your computer. It runs inside your browser. And your browser is a landlord with 14 other tenants living in the building. Every tab you've got open: your broker, your newsfeed, YouTube, that Discord you keep telling yourself you'll leave; all of it sits in the same memory pool as your charts. TradingView takes the blame for a mess it didn't make. It's the quiet roommate who gets yelled at because somebody else left dishes in the sink.

Honest minimum: benchmark 45,000, same as everything else here. The difference is memory. 16GB if you're really disciplined with your tabs, 32GB if you're not. And let's be adults; you're not.

But if somebody's selling you a $5,000 machine to run a browser, that person is not on your side.

Third place: TradeStation, and what Radar Screen really costs

TradeStation itself is a well-behaved desktop app. Efficient charts, stable platform. I've watched it run beautifully on modest hardware for years. I use it myself.

Then look at what's loaded into Radar Screen. Most traders picture Radar Screen wrong. It's not a spreadsheet. Every row is a live streaming data connection, and every column recalculates on every single tick. It's not a list of symbols; it's a payroll. Every symbol you add is another employee, and every column is another job you're paying that employee to do all day long.

A hundred symbols with a few basic indicator columns? Your machine shrugs. Six hundred symbols with options and a custom EasyLanguage indicator in every column? Now you've got a real job on your hands, and it's a job that never gets a lunch break.

One thing to know: EasyLanguage compiling is single-thread work. Core count doesn't rescue you here. Clock speed does.

Honest minimum: benchmark 45,000, strong single-core speed, and 32GB of RAM for a few hundred symbols with a normal column setup. Push to 500 or 600 symbols with EasyLanguage in every column and 64GB is the answer, not 32.

Second place: NinjaTrader has two personalities

NinjaTrader is different. It's not about what you load into it. It's that the platform has two completely separate personalities, and most traders only ever meet the friendly one.

NinjaTrader running live on a normal futures day with a handful of charts is surprisingly reasonable. It's a lean platform for what it does. Then you turn on tick replay or kick off a backtest, and you meet the other side.

Tick replay means you're asking your computer to relive every single trade that printed across your entire lookback period, in order, one at a time, and rebuild your indicators from scratch on every one of them. That's not charting. That's time travel. And time travel is expensive.

Running a backtest across every parameter combination is even worse. It pegs every core you own at 100% and holds them there. This is the one place in trading where core count genuinely earns its money, and it's why I'll put a heavier processor in front of a systems trader that I'd never recommend to a discretionary options guy.

Honest minimum: 32GB of RAM and a benchmark of 45,000 or better. If you backtest seriously, you want 60,000 or higher. That's not an upsell; that's the one workload that actually uses it. If your charts are the problem rather than your back tests, start with the NinjaTrader settings that cut chart lag before you buy anything.

First place: Thinkorswim, the most resource-hungry of the four

This is the one I opened with, and it does something to your machine that the other three don't do at all.

I want to be careful here, because I'm not calling it a bad platform. It's an outstanding platform. It's also the single most resource-hungry piece of software the average trader will ever install.

Thinkorswim runs on Java. Java doesn't borrow memory; it claims it. It grabs a big chunk up front, and as your session runs, it takes more and rarely gives any of it back. It's the guy at the buffet who takes a plate, sets it down at his table, then goes and gets a second one and a third. And nobody's clearing the plates.

It's using 4GB of RAM before you've opened your tenth chart. That's before scanners. Before the options chains. Before level two. Before a single custom study.

This is why traders swear their machine gets slower as the day goes on. With Thinkorswim, it literally does. Your 9:30 machine and your 2:00 machine are not the same machine. That was the trader from the top of this article. Nothing wrong with his hardware. Everything wrong with the match between his platform and his RAM. I've written more about why Thinkorswim keeps getting slower through the session if you want the longer version.

Sixteen years running this company, more than 20,000 trading computers built and shipped, and Thinkorswim has generated more "it's flying in the morning and it's useless by lunch" calls than the other three platforms combined. It isn't even close.

Honest minimum: 32GB of RAM. That's the floor, not the goal. If you're a heavy options trader with the chain open, multiple scanners, and a big layout stretched across four monitors, 64GB is a legitimate recommendation, and I make it without flinching. Benchmark 45,000 and up.

All four platforms side by side

Platform RAM Benchmark What matters most
TradingView 16GB disciplined, 32GB realistic 45,000 Close some browser tabs
TradeStation 32GB, 64GB with EasyLanguage everywhere 45,000+ Clock speed over core count
NinjaTrader 32GB minimum 45,000, 60,000+ for backtesting Cores, for tick replay and back tests
Thinkorswim 32GB floor, 64GB if serious 45,000+ RAM, all day long


The one mistake that wrecks every number above

If you run two of these platforms at the same time, and plenty of you do, charts on one, execution on the other, you don't average the requirements. You add them together. 32 plus 32 is not 32.

That is the single most common reason a machine that looked perfectly specced on paper falls apart on a busy day. It's the same math behind why 16GB of RAM can't handle three platforms at once, and why I keep pointing traders to the minimum RAM a day trader actually needs before they spend a dime on a processor.

Your homework before the next open

Run the free benchmark test on your actual trading machine. Add up what your platforms really demand. Then match the hardware to the way you personally trade, not to whatever is stacked highest on the shelf.

And if this feels like a lot of numbers to keep straight, don't worry about memorizing any of it. I put all of it in writing in the free Complete Guide to Trading Computers: RAM, processors, drives, monitors, what matters and what's a waste of your money, in plain English with zero tech shame attached.

If you're curious what we've got on sale right now, take a look at the desktop trading computers at your leisure.

Frequently Asked Questions

Why does my trading computer get slower as the day goes on?

Most of the time it's the platform, not the hardware. Thinkorswim runs on Java, and Java claims memory instead of borrowing it; it grabs a chunk up front, takes more as your session runs, and rarely gives any of it back. Your 9:30 machine and your 2:00 machine really are not the same machine.

Is 16GB of RAM enough for day trading?

Only in one case out of these four: TradingView on its own, and only if you're genuinely disciplined about how many browser tabs you keep open. For TradeStation, NinjaTrader or Thinkorswim, 32GB is the honest minimum, and 64GB is the real answer for heavy options layouts or hundreds of Radar Screen symbols with EasyLanguage in every column.

What benchmark score do I need to trade?

45,000 is the minimum for all four of these platforms. The only place I raise it is serious NinjaTrader backtesting and tick replay, where I want to see 60,000 or higher, because that's the one workload that actually uses every core you own.

Do I need more cores or higher clock speed for trading?

It depends entirely on the platform. TradeStation's EasyLanguage compiling is single-thread work, so clock speed helps, and core count doesn't rescue you. NinjaTrader backtests peg every core at 100% and hold them there, so that's where cores earn their money.

Why does TradingView slow my computer down if it's the most efficient platform?

Because TradingView doesn't run on your computer; it runs inside your browser. Every other tab you have open- broker, newsfeed, YouTube, Discord- sits in the same memory pool as your charts. TradingView gets blamed for a mess it didn't make.

Do I need a $5,000 computer to run these platforms?

Not for every platform. If someone is selling you a $5,000 machine to run a browser-based charting tool, that person is not on your side. Match the machine to how you actually trade: light charting needs far less than a heavy options layout across four monitors or a systems trader running backtests all day.

Eddie Z
Eddie Z
Founder, EZ Trading Computers & EZBreakouts
Eddie Z is a full-time day trader who has spent 39 years on Wall Street, starting on the floor of the NYMEX in 1987. Since 2010 he has built more than 20,000 computers for traders — only for traders. Almost everyone else in this business came from the technology side and later discovered traders. Eddie came the other way round, and still has his own money on the line at 9:30 every morning.